Science | The Atlantic

Is There Anything BPC-157 Can’t Cure?

2026-10-27T07:00:00-04:00

The wellness industry loves a secret cure. I encountered dozens of them at Dave Asprey’s Beyond Biohacking Conference, held in May on several floors of a hotel in downtown Austin. There were supplements galore, of course: algae tablets for detoxification, ketones for concentration, nitric oxide for circulation. One company claimed that its capsules contained quantum energy. I saw people hooked to IV vitamin drips, sweating inside infrared saunas, and sealed in hyperbaric oxygen chambers. I watched a woman climb into a giant quartz bowl, which someone then struck with a mallet. I tried a pair of inflatable compression pants that rhythmically squeezed my legs to stimulate lymphatic drainage. I also stripped to my boxers and spent a thrillingly miserable two and a half minutes inside a cryotherapy booth.

The main attraction, though, was peptides, the short chains of amino acids now hyped as miracle drugs. In a ballroom packed with several hundred people, Asprey and a fellow health guru, Jay Campbell, extolled the many virtues of peptides, railed against restrictions on their promotion, and predicted a future in which they transform mainstream medicine. Both men are in their 50s, with shredded physiques and zero medical degrees. Asprey is best known for popularizing the trend of dumping butter and coconut oil into your coffee. Campbell made his name advocating for testosterone replacement and now bills himself as one of the world’s leading authorities on peptides. He spoke excitedly about one cutting-edge compound, called Klotho, that can “literally lower your biological age.” He told the crowd that he’s heard from women in their 70s who, after taking Klotho and the peptide FLGR242, started menstruating again. (Instead of imploring them to call their doctor, which is what the American Cancer Society recommends for postmenopausal bleeding, his advice was to “keep injecting.”) As peptides become more readily available over the next few years, Campbell anticipates that disease as we know it will come to an end, including “ontological” mutations. (I’m guessing these are mutations that afflict only philosophers.)

Campbell and Asprey have been promoting underground peptides in their books and on their podcasts for years. Lately, though, these gray-market drugs have started to gain traction with a broader audience. The most talked-about among them goes by the cryptic designation BPC‑157. It’s commonplace enough that asking questions about it immediately tags you as a tourist. BPC‑157 is a component in most of the popular peptide combos, ones with names such as Glow, Klow, and Wolverine (a reference to the comic-book superhero with self-healing powers, not the carcass-eating skunk-bear). It’s rumored to be good for stomach ailments including acid reflux, bloating, and ulcers. Also, sprained ankles. It will reportedly ameliorate arthritis, carpal tunnel syndrome, and sciatica, not to mention hernias and migraines. BPC‑157 has been said to loosen stiff shoulders and repair runner’s knee. Suffering from bursitis? Gingivitis? Road rash? Low libido? Its boosters promise that it can help with all of these conditions and more.

In recent months, BPC‑157’s profile has risen significantly, at least judging by Google searches and Instagram videos, some of which feature a creepy AI-generated character who magically restores torn ligaments. Praise from big-name podcasters such as Andrew Huberman and Joe Rogan has no doubt contributed to its popularity (Rogan declared BPC‑157 “super legit”). Reliable sales figures, though, are hard to come by. Conventional drugs that have been approved by the FDA—including GLP-1s like Ozempic, which are also peptides—are sold by multibillion-dollar global behemoths with legions of researchers and executives. Peptides like BPC‑157 are sold by … honestly, who knows? The companies pop up and disappear. The proprietor of your favorite American peptide supplier might be an enterprising undergraduate who buys in bulk from China, where many peptides are manufactured, and relabels them. A small glass vial of freeze-dried powder arrives on your doorstep; you dissolve the powder in sterile water, draw it into a syringe, and inject it into your abdomen. Peptides are usually advertised as “for research use only,” a wink-wink disclaimer that allows vendors to maintain that they’re not violating federal laws regulating drug sales. Maybe that gives you pause. Or maybe you think that these sketchy transactions are necessary because the establishment—that is, Big Pharma and the United States government—is suppressing the good stuff.

Amid the sellers of red-light panels and blue-light glasses at Beyond, I stumbled on a company marketing a drinkable version of BPC‑157. Its giveaway sticker read We Prefer Oral Peptides, with the word peptides in much smaller letters because haha sex. The representative behind the counter referred to BPC‑157 as “the soccer mom’s peptide” and nonchalantly squirted some from what was essentially a hand-soap dispenser into a tiny plastic cup, then held it out to me. I hesitated, and the peptide hung in the air, unclaimed. “Are you afraid of it?” he asked. The not-so-subtle subtext was that I shouldn’t be, that taking BPC‑157 was equivalent to trying the dissolvable herbal-tea strips a few tables down or giving the giant quartz bowl a whirl. For the record: We were in a hotel hallway and I was being encouraged—dared, really—to try a drug that isn’t FDA-approved, is prohibited by the World Anti-Doping Agency for athletes’ use, and allegedly works throughout the body via some mysterious mechanism that is not entirely predictable. I was being offered an unspecified dose, although I wouldn’t know what dose would be appropriate, because there are no agreed-upon dosing guidelines. I declined the peptide. The representative shrugged and drank it himself.

Perhaps I was being too cautious. Thousands of Americans have ingested or injected BPC‑157. That guy didn’t keel over after he swigged it. Whatever you think of their relevant credentials, Asprey and Campbell appear to be insufferably healthy middle-aged men. Robert F. Kennedy Jr., the secretary of health and human services, is on record as being “a big fan of peptides,” a casually sweeping endorsement of an entire class of drugs. Although Kennedy has stoked vaccine hesitancy for most of this century and argues that Americans are overmedicated, he has a soft spot for these particular shots. The “Make America Healthy Again” movement is officially on board too, which feels strange considering that MAHA is all about treating anything modern and artificial with suspicion. They’re the raw-milk people! They don’t even like Doritos and Coca-Cola! And yet they’ve decided that these highly processed powders, assembled by machines that link amino acids together, are somehow consistent with the nature-knows-best ethos. For MAHA, the renegade cachet of peptides is central to their appeal, and the fact that medical experts are wary of them is a plus. Under Kennedy, the FDA may soon allow six peptides, including BPC‑157, to be made in the United States, which would mean that doctors could prescribe them, and that telehealth companies such as Hims & Hers, best known for discreetly shipping men generic Viagra, could offer them for sale.

[Read: America is about to go over a peptide cliff]

Is there any evidence that BPC‑157 works? Its evangelists like to note that the peptide has several decades of peer-reviewed research behind it. Those papers, more than 200 of them, are packed with charts and citations and have been published in journals that are respectable enough, if not quite top-tier. BPC‑157 has been tested on mice, rabbits, guinea pigs, chickens, honeybees, and, above all else, rats. So many rats. These rats have endured all manner of miseries in the name of science. They have had their tendons severed, stomachs ulcerated, tails amputated, muscles crushed, bowels perforated, spinal cords compressed, blood vessels tied off. The ones that received the peptide consistently fared better, according to the studies: Their wounds closed more quickly, tendons knitted together more completely, ulcers shrank. Overall, they were more likely to survive this catalog of laboratory horrors than the unlucky controls. If the published literature is to be believed, BPC‑157 is a rodent panacea.

When you read the papers, one after another, until you’re bleary-eyed and bewildered, you start to notice some patterns. One is that very few of these studies have been conducted on humans, and humans, famously, are not rats. Another is that, as I’ve already suggested, BPC‑157 is said to work almost every time for almost everything. Behold these triumphant verbs: improves, enhances, promotes, stimulates, accelerates, restores, increases, prevents, abolishes, counteracts, heals, protects. And it seems to yield benefits regardless of the form, or method of application, or which orifice it is delivered into. Also, when you examine the list of authors, you see many of the same names crop up again and again. One name, however, appears at the top of nearly every paper: Predrag Sikirić.

Sikirić is a 72-year-old professor emeritus of pharmacology at the University of Zagreb School of Medicine, in Croatia. I emailed him. I wanted to know where this wondrous elixir, supposedly good for both heartburn and a ruptured Achilles, came from, and why we’re hearing about it only from podcasters. Precisely four minutes later, my phone buzzed. An insistent man with a thick accent wanted to know how quickly I could get on a plane.

The current global hub for BPC‑157 research is a cramped, cluttered room with a single window half-obscured by a tangle of plants. Sikirić’s laboratory is at the university’s medical school, roughly a mile north of Zagreb’s main square, near a small café where future doctors smoke cigarettes and grouse about their workload. These modest quarters are temporary; renovations are under way on a larger laboratory in the building across the street. For now, new experiments are on hold, and Sikirić and his longtime lab assistant are confined to a glorified storage closet with empty rat enclosures stacked next to boxes of small plastic tubes and bags of feed. Among the haphazard decorations—including a foam brain and a hat that reads HAPPY ANIMAL HAPPY HUMAN—hangs a collage of patent medicines from the heyday of cure-alls, back when you could make whatever claims you wished about your restorative tonic without fear of government regulators shutting you down. On a counter sits a large glass bottle of clear liquid, the letters BPC scrawled on the label in blue marker.

The peptide in that bottle is Sikirić’s life’s work. When he was still a student, inspired by the pioneering research of the endocrinologist Hans Selye, who studied the biological effects of stress, Sikirić came to believe that the body produced, in essence, some universal antidote to injury, a molecule that would cue damaged tissue to return to its normal state, regardless of what had caused the harm. He reasoned that the remedy was probably found in gastric juice, because the stomach is constantly adjusting to fresh insults. Sikirić spent the 1980s searching for it, and the decades since trying to convince the world that he found it. Along the way, his work has often drawn eye rolls from colleagues at the medical school, some of whom remain unmoved by the animal evidence he has amassed. Even those who admire Sikirić’s persistence and have known him for years, such as Miloš Judaš, the director of the university’s Croatian Institute for Brain Research, do not share his faith in the curative properties of BPC‑157 (“We are not sure whether it really works or not,” Judaš told me). Sikirić’s discussions with pharmaceutical companies have repeatedly stalled. The recognition he believed would be forthcoming from the scientific community has never materialized, despite his prolific scholarly output.

But none of that has deterred Sikirić in his effort to, as he told me, “bring BPC to mankind.” He’s just kept plugging away at this singular line of research, coaxing generations of graduate students to join him in the mission, and dreaming up new experiments to test his theory. Now the world is seeking him out. Along with reporters like me, he hears from peptide promoters eager to brag about being buddies with the discoverer of BPC‑157. How his peptide took center stage, Sikirić isn’t sure. For a long time, BPC‑157 wasn’t really known outside his lab except among readers of niche scientific journals such as Digestive Diseases and Sciences and the European Journal of Pharmacology. In the mid-2010s, it began appearing on bodybuilding forums and in biohacking guides, where it was touted as a way to speed healing and recovery. But it remained on the fringes until recently. Testimonials from Rogan and Huberman, whose health recommendations reach enormous audiences, certainly helped spread the word: Rogan swore it fixed his hurt elbow, while Huberman credited it with eliminating his back pain.

Mare Milin for The Atlantic

Predrag Sikirić has published hundreds of papers purporting to show the salutary effects of the peptide BPC⁠-157.

However it happened, Sikirić is pleased that his peptide is at last having its moment. He is gratified that BPC‑157 is flourishing on the gray market as well. “Officially, I have to be against any kind of usage without full approval,” he said. Unofficially, he’s all for it. He interprets BPC‑157’s word-of-mouth success as proof that his peptide is helping the masses. If it wasn’t, why would so many people recommend it? He is delighted, too, that the health secretary of the United States is a believer in peptides. I asked if he’s ever spoken with Kennedy, who has a habit of reaching out to researchers whose work he admires. “I would kindly reject to reply to the question,” he said. “This is good English?” (An HHS spokesperson did not respond to an inquiry about whether Kennedy had spoken with Sikirić.)

Sikirić can be at once engaging and cagey. Try to pin him down, and he launches into an unrelated anecdote. Broach a contentious topic, and he responds with a puzzling analogy. Ask about his past, and he grudgingly provides a few biographical details. Sikirić grew up in Zagreb, which was then part of Yugoslavia. His father, a lawyer, was jailed repeatedly for his opposition to the country’s ruling Communist party; his mother, an encyclopedia editor, encountered government censorship. As a child, Sikirić underwent multiple surgeries, and those hospital stays made him want to be a doctor. Medicine also offered an escape from politics: “The pain is the pain for anybody,” he said. After medical school, he worked for three years as an emergency physician in a small town, treating victims of car accidents, heart attacks, and stabbings.

But a country doctor can only do so much. The wonder drug he envisioned might help millions. Sikirić returned to the University of Zagreb to pursue this dream with a single-minded, near-manic intensity. In addition to churning out hundreds of papers, he has supervised numerous doctoral dissertations, all of them on BPC‑157. His students tell of late-night phone calls from Sikirić, eager to share an insight. He and his wife of more than 40 years don’t have children. He has no hobbies, doesn’t take vacations to the nearby coast, and works for long stretches without a break. Sikirić walks quickly, albeit at a slight angle, as if he might tip over. He drives quickly, too: When I rode with him, he sped through downtown Zagreb, braking a few feet from crossing pedestrians, half-grinning at my evident anxiety. You wouldn’t think it to look at him—baggy professorial sweater and corduroys, tightly cropped gray hair—but Sikirić is a septuagenarian in a hurry.

BPC‑157’s newfound fame has brought with it unwelcome scrutiny. Last year, a group of researchers in Poland raised concerns about BPC‑157’s safety, arguing that how it works is poorly understood and warning that it could cause neurological damage or promote tumor growth. These possibilities, they wrote, must be considered “before falling into complete admiration.” In response, Sikirić pointed to his animal studies suggesting that BPC‑157, on the contrary, inhibits tumor growth and protects against neurological damage. When I asked him about the Polish paper, he dismissed it as a bid for attention and emphasized how much it upset him. If BPC‑157 really increased the risk of cancer, Sikirić insists that he would tell the public to stop using it immediately.

One argument for BPC‑157’s safety rests on the assumption that if the body makes it, it must be safe. But does the body make it? Sikirić has asserted that his peptide is a fragment of a protein, but he’s never publicly shared key details about that protein. Several researchers have attempted to locate BPC-157’s amino-acid sequence in databases of protein and DNA sequences, to no avail. I asked William Pearson, a computational biologist who co-developed the first widely used search tool for such databases, to try to determine where BPC‑157 comes from. After hours of combing through repositories containing hundreds of millions of protein and DNA sequences, Pearson couldn’t find a match. If the peptide occurs in humans, those searches should have turned it up. He also identified no matches among proteins from other animals—including, it’s worth noting, rats.

Sikirić downplayed the failed searches. He doubts that those databases are comprehensive, though he acknowledged that he hasn’t used them himself. Sikirić told me he’s sure that BPC‑157 originates in humans because he isolated it from newborns in the ’90s. I didn’t remember seeing that study in any of his papers, but that’s because Sikirić says he never published it and doesn’t intend to (“Some things you have to keep for yourself,” he said). As for the gene behind BPC‑157, he maintains that he’s close to finding it, but won’t reveal how he is conducting that search. He likened providing such basic information to a country disclosing its nuclear secrets. The answers, he said, could be found only in his head. I asked what would happen if he got run over by a bus: “This could be huge problem.”

Sikirić believes that, like the cure-alls in his office collage, his peptide can be applied to any ailment, from minor aches to major diseases. During my stay in Zagreb, we played a little game: I named a malady and he told me whether BPC‑157 could treat it. Macular degeneration? “Yes, it could.” Gut inflammation? “This is the favorite site of activity.” Dementia? “Certainly.” COVID? “Naturally, yes.” Concussions? “We published data about it.” Anxiety? “The experiment was helpful in the mice.” Depression? “Exactly.” Heart attacks? “Any kind of condition.” Headaches? “Any kind of pain.” In Sikirić’s estimation, BPC‑157 can do no wrong. Its all-purpose goodness is fundamental to his theory: The peptide is consistently returning the body to its healthiest state. Other drugs have limitations and downsides, but not BPC‑157. He compares it to steroids, but believes it’s superior because BPC‑157 has no known toxic dose (“You could eat BPC without any harm—literally eat, like bread,” he said) and you can take it indefinitely.

Until I told him, Sikirić seemed unaware that some users report distressing side effects from BPC‑157. There’s an online support group called “Haters of BPC‑157,” in which people who believe that the peptide has harmed them commiserate and trade notes on recovery. The most common complaint appears to be anhedonia, which is a kind of emotional numbing. They say they struggle to feel pleasure after taking the peptide. Some blame themselves for using an unapproved drug. Others are angry at Sikirić (“Fuck him,” one member of the group messaged me). Sikirić told me he felt sorry that people were suffering but doesn’t believe it has anything to do with his peptide. (Predictably, he contends that BPC‑157 actually counteracts anhedonia.) Maybe the product they bought was contaminated, he argued, or maybe the cause was something else entirely. These are only anecdotes, he pointed out. And he’s right; that is all they are. But the human evidence in favor of BPC‑157 is also largely anecdotal.

Illustration by The Atlantic. Sources: Durham Peptides; Ultra Pure Bioscience; Amino Research; XL Peptides.

Peptides are usually listed as “for research use only,” a disclaimer to skirt federal laws regulating drug sales.

Sikirić takes BPC‑157 himself. Of course he does. I first suspected this when I noticed an empty plastic tube in the cup holder in his car. He was somewhat sheepish when I asked about it, and not forthcoming about how often, or for what purpose, he takes it. As a rule, it’s frowned upon for scientists to sample the merchandise, though there have been famous exceptions. Jonas Salk tested the polio vaccine on himself before it was given to millions of American children; Albert Hofmann dropped way too much acid once before biking home. I asked many of Sikirić’s colleagues and former students if they had taken BPC‑157. Several told me they would never try a drug that hadn’t been approved for human use. Others swore by it. They use it when they feel a cold coming on. They blend it into a cream and smear it on burns and diaper rash. They take it when they twist their ankle or strain their back. A bunch of them give it to their dogs. One former student, now a doctor, showed me a small tube of BPC‑157 that she carries in her purse, just in case.

This is not how drug development is supposed to work. After a promising candidate is tested on animals, it normally passes through three phases of human clinical trials, after which regulators review the evidence and make a determination. The story of why BPC‑157 hasn’t followed this path is convoluted. More than two decades ago, clinical trials began under the auspices of Pliva, a Croatian drug company, but didn’t reach the third phase. Pliva’s research institute, which was responsible for new drug development, was acquired by GlaxoSmithKline in 2006, and progress on BPC‑157 subsequently ceased. In Sikirić’s version, GSK, as it’s now known, decided to shelve the peptide because it was too effective and the company feared it would cannibalize sales of its other drugs. “Let’s just say major breakthroughs are not welcome,” he told me. Yet it’s hard to imagine a pharmaceutical company nixing an important discovery knowing that a competitor might snatch it up. A study conducted by Pliva researchers found that although BPC‑157 did hasten the healing of skin wounds in rats, the results were similar to those achieved with an already approved drug. The peptide did something, in other words, but it wasn’t a breakthrough.

Another problem may have been Sikirić himself. Two former Pliva employees involved in the BPC‑157 project told me Sikirić’s passion sometimes seemed to override his objectivity. He bristled at pushback, and his commitment bordered on obsession. As one of them put it, when it came to his peptide, “Predrag was in love.” Neither knew for sure why GSK didn’t move forward with BPC‑157 (a GSK spokesperson told me she couldn’t find anyone with knowledge about a decision made 20 years ago). But they noted that it’s far from unusual for experimental drugs to be abandoned during clinical trials; in fact, about 90 percent don’t make it all the way through the process, including ones that show some benefit in animal studies.

After GSK dropped the peptide, the rights were returned to Sikirić, allowing him to pursue BPC‑157 trials elsewhere. In 2015, a Phase 1 clinical trial was conducted in Tijuana, Mexico, but the results have never been published. The trial’s sponsor was PharmaCotherapia, a Croatian company co-owned by Sikirić that last year reported having no employees. Sikirić let me briefly review the safety findings on his laptop but declined to share the full report. The summary I read described mild adverse events among some subjects but concluded that BPC‑157 was “safe and well tolerated.” According to Sikirić, the report was never published because the trial’s principal investigator, Rufino Menchaca, a clinical neurologist, wouldn’t sign off once the trial was completed (Menchaca, through a representative, declined to speak with me). The reason for his refusal to sign isn’t entirely clear, although Sikirić told me it had to do with worry that others would steal the data if they were published. Sikirić hopes to restart clinical trials soon, again in Tijuana, looking specifically at whether BPC‑157 is a beneficial treatment for multiple sclerosis—though, in his mind, there is no question. He’s certain that the experiments will confirm his expectations.

Sikirić is so determined to vindicate BPC‑157 that he frequently sounds more like a salesman than a scientist. Money, however, doesn’t seem to be what drives him: As Sikirić will admit, he has no knack for business, and at the moment it’s shady online sellers who are getting rich from his peptide, not him. He just wants to be right, and he wouldn’t mind getting some credit, too. As I’m typing this, Sikirić is sending me yet more papers via WhatsApp even though it’s after midnight in Croatia. He needs me, and everyone else, to set aside any misgivings and agree with him that BPC‑157 is a marvel. Skepticism is taken as an insult. Too often Sikirić’s message is: Trust me. Trust me about where BPC‑157 comes from. Trust me that it works on humans, not just rats. Trust me that it’s safe.

FDA scientists don’t seem to trust him. In June, the FDA published a 68-page report that looked closely at BPC‑157 studies and found them unconvincing, arguing that they lack sufficient human-safety data and that too little is known about how or whether the peptide works. Agency researchers couldn’t tell from the literature how dosage affected outcomes. They flagged the possibility that it could provoke harmful immune-system reactions. Claims about BPC‑157 should be “interpreted with caution,” they wrote, because of troubling gaps in the literature.

But when an FDA advisory committee met in July, it more or less disregarded that report and recommended allowing BPC‑157, along with five other peptides, to be made by licensed compounding pharmacies in the United States. It probably wasn’t a coincidence that weeks before the vote, the agency had announced the appointments of eight new committee members, six of whom had ties to the wellness industry, which hopes to capitalize on a regulatory change. (An HHS spokesperson told me that “all committee members underwent the same ethics review and vetting process required of all FDA advisory committee members.”) One peptide lobbyist told me that the projected profits for BPC‑157 could be in the billions. Hims & Hers plans to start manufacturing BPC‑157 and allowing customers with a prescription to purchase it, assuming the FDA gives the green light (the company’s CEO, Andrew Dudum, told shareholders that he was “extremely excited about this category”). Although FDA officials still have to accept the committee’s recommendation, that seems like a formality, given that Kennedy oversees the agency.

[Read: The peptide boom is getting out of hand]

I have mixed feelings about the move. No, we don’t know enough about these peptides, and providing what amounts to backdoor government approval seems like a terrible idea. At the same time, if people are going to take BPC‑157, I’d rather the supply come from a licensed pharmacy than some guy with a storage unit and an overseas hookup. I kept thinking about the question the peptide representative at the Beyond conference asked me: “Are you afraid of it?” Sikirić sent me home with a small plastic tube containing a single milligram of BPC‑157. It looked like nothing, barely a few grains of fine white powder. I considered taking it, both for the sake of this story and because I have this weird twinge in my right shoulder sometimes. Instead, I sent it off to get tested. The report that came back was glowing: Unlike the gray-market peptides, many of which contain filler and God knows what else, the peptide from Sikirić’s laboratory was chemically pure, full-strength, 100 percent BPC‑157. It was the real thing—whatever that is.

This article appears in the December 2026 print edition with the headline “Mr. Peptide.”

Two Firefighters Died in Yosemite—For What?

2026-10-19T16:48:29-04:00

This year, the Trump administration gave federal firefighters a mandate: no more letting fires burn. Conventional wisdom among firefighters and public-land managers has held that allowing fires in remote areas to run their course can be beneficial. But now the administration’s new U.S. Wildland Fire Service was to pursue a policy of “full suppression” on every wildfire, Interior Secretary Doug Burgum instructed his workforce, to avoid the possibility of dangerous spread. “Accelerate initial attack,” Agriculture Secretary Brooke Rollins put it in her own directive to the Forest Service. “Suppress fires early and limit growth.” And when Yosemite National Park caught on fire last month, that is what firefighters did.

The hazy skies above the park filled with aircraft—planes dropping fire retardant on forested slopes; helicopters ferrying hotshot crews to the blaze. At one point, at least 17 helicopters were in rotation, along with more than 1,000 firefighters. “I’ve worked in Yosemite and with Yosemite for almost 40 years,” Tom Nichols, a former head of fire and aviation for the National Park Service (and no relation to the Atlantic writer Tom Nichols), told me. “I’ve never seen that kind of response.”

At the start, attacking the fire aggressively made sense: It had started near a popular hiking trail and was burning on a slope above Wawona, a small community in the park. It looked to be human caused. In such situations, federal policy, even before this year, called for firefighters to use “full suppression” to try to put it out as quickly as possible, former Yosemite fire officials told me.

But as it developed, the Dome Fire spread to the north and east, farther up the mountains, away from any homes or communities. The resources being poured into fighting it began to baffle some fire veterans. As a former Yosemite official put it to me, “There’s really no reason to be attacking a fire like that.” (I spoke with several sources for this story who requested anonymity because they were not authorized to speak publicly and hoped to avoid professional repercussions.) Such an aggressive approach can be dangerous. Five days into the effort to put it out, one of the helicopters was doing water-bucket drops from a nearby lake when it crashed. Both pilots died.

“For what?” one federal fire official told me. “There’s nothing out there. Just let nature do its thing.” Another official had to ask: “Did they die in vain?”

Under the Trump administration’s policy of full fire suppression, so far this year entrapments—when firefighters’ positions are surrounded by flames—are above the average for the previous two decades. At least 17 wildland firefighters, pilots, and contractors have been killed, surpassing the full-year average over the past decade, according to the Wildland Firefighter Foundation. That number includes four firefighters in Colorado who died this summer responding to the Knowles and Gore Fires as flames spread amid high winds and low visibility. After that incident, several Democratic senators called on the Interior Department to investigate whether the full-suppression policy put firefighters at undue risk.

Brian Fennessy, the Wildland Fire Service chief, doesn’t think that the new policy is putting firefighters in more danger, he told me. The two pilots’ deaths in the Dome Fire were a tragedy—one he hopes his agency can learn from—but “this was not the year to allow managed fire on the landscape,” he said. The winter was historically warm, and snowpacks were at record lows across much of the western United States. Federal authorities were not about to let fires burn unattended, even in the wild. And the administration’s policy of suppression does mean increasing the number of people and helicopters fighting a given fire, he said: “Instead of two engines and a helicopter, maybe you’ve got four engines, two helicopters, maybe an air tanker.” Given how hot and dry it’s been, had firefighters not focused on putting out fires, in Yosemite and elsewhere, many could have been worse, he said.

The question of whether to harness or extinguish fire is particularly relevant in Yosemite. This is where America’s formal debate over that policy began, after Abraham Lincoln signed a bill in 1864 to grant the Yosemite Valley and the Mariposa Grove of giant sequoias to California as a public trust. The commission to oversee the area disagreed about the question of annual burning by Native Americans, Stephen Pyne, a historian of fire, told me. Fires set by the Ahwahneechees were “deliberate and purposeful, even artful,” Pyne writes in his book about Yosemite’s fire history, Pyrocene Park: “They burned to keep grasslands open and attract game animals like deer to fresh forage; to keep trees from crowding out annual-grass prairies; to help harvest acorns; to stimulate willows useful for basketry; to promote berries.” Some of the Yosemite’s first commissioners saw the value in that. Others, such as Frederick Law Olmsted, considered such fires vandalism that undermined, as he put it, “the duty of preservation.”

For more than a century afterwards, the suppressionists prevailed until, in 1974, Yosemite leaders allowed a major lightning-caused fire to burn in the Illilouette Creek basin, whose creek runs behind Half Dome, the park’s famous bald mountain. This was the first big test of its new policy to manage certain fires for ecological benefit. Since then, more than 100 fires have come through that basin, and scientists have documented its renaissance. The proportion of the basin covered by dense conifer forest fell by 20 percent, and a mosaic of meadows and grasslands emerged. Biodiversity bloomed, as bees and bats and understory plants thrived. Even streams flowed a bit more abundantly, as fewer trees competed for water and snow had more open space to accumulate, Scott Stephens, a professor of wildfire science at UC Berkeley, told me. The regular burning also makes the basin more capable of handling future fire, he said, by minimizing how much of it burns at a destructive intensity. It’s now “one of the most resilient, functioning forests in all of California.”

Set under the right conditions, prescribed fires, too, can help prevent hotter-burning conflagrations. The evidence was plain just four years ago in Yosemite, when the Washburn Fire bore down on the Mariposa Grove and its revered giant sequoias. The park had conducted more than 20 prescribed fires in that area since the 1970s. And when the Washburn Fire approached the grove, the flames moved around the treated areas as if they were boulders in a stream, sparing the sequoias that have been burning up in great numbers elsewhere in California.

But this year, at least, under the Trump administration’s leadership, the park seems to have stepped back from prescribed burns as well. Although in many recent years Yosemite has done several prescribed fires, this year the park has done none and even canceled a long-planned prescribed fire near the Mariposa Grove scheduled for May, around Memorial Day, according to two people familiar with the decision. They added that Ray McPadden, the superintendent of the park, was concerned it would affect park visitation because it required closing parts of the grove temporarily. (McPadden did not respond to a request for comment; the Wildland Fire Service said Yosemite was looking for an opportunity to complete the burn this fall.)

Over the long term, avoiding fire puts the sequoias at risk, Cicely Muldoon, who retired as Yosemite’s superintendent last year, told me. And on top of that, jettisoning controlled burns, which ultimately keep fires modest, also means that “you’re really putting people in harm’s way.” A little fire—however it starts—can help prevent future huge fires.

Under the Trump administration’s reorganization of federal firefighting resources, superintendents of national parks no longer decide how to manage fires that start within their borders. That authority now lies with the Wildland Fire Service, which is housed within the Department of the Interior. After last summer’s Dragon Bravo fire at the Grand Canyon, a lightning-caused fire that destroyed more than 100 buildings, Trump-administration officials have shunned the notion of letting fires burn for the health of the landscape.

From the beginning, people fighting the Dome Fire—including those from the National Park Service and two different California interagency incident-management teams—did all that they could to put the fire out, for weeks. Because it was burning on remote slopes, these firefighters relied heavily on helicopters to reach the flames, haul in supplies, and drop water. (The teams also sent in mule trains and firefighters hiking on foot.) Crews camped in the woods as they cut fire lines by hand. Such a response is expensive too. Fighting a 6,700-acre fire that has burned no buildings has cost at least $55 million so far, according to the National Interagency Fire Center. The Wildland Fire Service said in a statement that fire managers were concerned about different assets as the fire progressed, including Yosemite West—another nearby community—and the Badger Pass ski area.

In the field, the work was logistically complex, and its tempo was high. Six firefighters got injured or became ill, according to the Wildland Fire Service. People were reporting falls, bee stings, smoke inhalation, one firefighter told me, and the frequency of incidents was “more than you’d expect” for what they were gaining from keeping the fire at bay, the firefighter said. “People were like, What are we doing? There’s nothing out here. We’re slamming this fire shut,” this person added. “People were questioning that.”

Congestion from so many helicopters caused safety concerns, too, including a close call on September 19 between a firefighting helicopter and skydivers near the Mariposa-Yosemite Airport, a small airfield outside the park. Although an employee from Skydive Yosemite had alerted firefighters about their plans in a morning briefing, when five people jumped from 14,000 feet later that day, they floated down as a helicopter carrying cargo hovered in the vicinity of their landing zone, Paul Wignall, the company’s owner, told me. The helicopter pilot did not appear to be monitoring radio messages from the skydiving plane, Wignall said, and one of the divers drifted close enough to make eye contact with the helicopter pilot. By the time others in the group came down, the helicopter had flown off. Wignall was particularly concerned about the powerful rush of air from the helicopter’s blade. “If he can see a parachute coming, he should get the hell out of there,” Wignall said.

During the incident, the helicopter pilot “took action in response” to the skydivers, the Wildland Fire Service said in a statement, and firefighters later met with the skydiving operator to discuss the incident. The FAA brought in a temporary control tower, and firefighters moved equipment to avoid potential problems, Adrienne Freeman, a spokesperson on one of the interagency incident-management teams for the Dome Fire, told me. Authorities would later close the airport for a week. The day after the skydiving incident, another helicopter went down fighting the fire.

The circumstances of the deadly crash on September 20 remain under investigation. The firefighting community has rallied support for the families of the two pilots who died—Greg King and Ryan Cutter—and held ceremonies to honor their service. The pilots worked for Precision, a Forest Service contractor based in Oregon. After the crash, the use of aircraft on the fire paused as Fire Service managers assessed the safety risk. The losses cast a shadow over the whole enterprise.

The danger of letting a fire burn might very well have been higher than normal because of the hot, dry conditions this year, as Brian Fennessy, the federal fire chief, said. But these types of years are more common now. The next one could be hot and dry too. In Yosemite, and across America, the choice to always put the fires out carries its own risks. The kindling keeps piling up, increasing the chance that the next one will be so big that no one can control it.

Nine-Inch Snails

2026-10-14T12:53:19-04:00

Since September, giant snails have put one Florida neighborhood on high alert. No one in a roughly 300-acre zone in the city of Hialeah, just outside of Miami, is allowed to move yard waste or even a single brick into or out of the neighborhood without explicit permission from state authorities. Teams of scientists patrol the area in search of giant African land snails, enormous mollusks with tiny teeth, a foot bigger than a 5-year-old’s, and the tendency to spread the rat-lungworm parasite, which is exactly as delightful as it sounds.

The snails, which actually grow to about eight inches in length, but can stretch—if nothing else, for headline-writing purposes—to nine inches, have worried Florida authorities since 1969. Since then, the state has spent millions of dollars fending off three infestations. One such operation took a decade to complete. Now, whenever a snail is found, inspectors from the state agriculture department’s Division of Plant Industry start patrolling. They set out bait and drench hatchling-harboring soil with anti-mollusk poison. Then they distribute mug shots to the public. (That step might seem unnecessary, given that we are talking about snails the size of chipmunks, but the experts I talked with all emphasized the importance of anti-snail awareness.) An area is considered clean only after it has been monitored and treated for three years.

This time, the problem “seems to be limited to just one city block,” Paul Skelley, a main architect of the agriculture department’s crackdown, told me. But the response is full-bore. “We are serious about this snail in Florida,” Skelley said. “When you find these cancers, you don’t want to let them fester. You need to cut them out.”

Skelley was part of the response to the last big infestation, which lasted from 2011 until 2021, and he sounds somewhat keyed up on the phone. Giant African land snails (also known as GALS) really are giant, and meaty and sweaty-looking, like a boxer’s fist just out of the glove. “When you see the whole body out moving around,” he said, “it’s unbelievable.”

Earlier this century, the International Union for Conservation of Nature declared these snails one of the worst invasive species on Earth. GALS are prevalent across a thick belt of the globe, centered on the equator, and their slime trail is widening: After expanding their range from East Africa through Madagascar, India, and the Pacific Islands, they have lately been spotted in Argentina, Beijing, and, somehow, the Italian Alps. And as climate change makes more of the world warmer and wetter, they could conquer even more territory. Skelley considers his state a firewall. If the snails spread through Florida, “they could get quite a distance up the East Coast and inland” by hitching rides with humans, he said.

Once they arrive somewhere new, GALS quickly multiply, and then make trouble just by being themselves. They love eating—beans, melons, ornamental flowers, shrubs—which is why they’re a hazard to agriculture. When cars run them over, the snails’ pointy shells can puncture tires. In search of minerals needed to grow their shells, they target stucco, eating our houses to build theirs.

All of this can happen at a scale difficult to imagine. In the past, in Hawaii, brigades of the snails have blocked roads after rainstorms, eventually creating “a sort of mush of snail bodies and slime,” Robert Cowie, a snail researcher at the University of Hawaii at Mānoa, told me. “Cars were skidding.” In Keralam, in India, where recent monsoons created the wet, humid conditions in which the snails thrive, teachers clear them daily from schools and playgrounds. In Florida, during the most recent big outbreak, Skelley’s team would purge a yard the size of a couple of parking spaces and end up pulling out maybe 2,000 snails.

On their own, the African snails move as one would expect, but Skelley jokes that they can go up to 70 miles an hour—“as fast as humans can move them.” And humans—whether they’re taken in by the snails’ charisma and impressive bulk, or just unaware of the threat—like to move them. Florida’s first brush with GALS began in 1966, when a young boy brought three home from a vacation in Hawaii. His grandmother found them, the story goes, and put them out in her North Miami garden. By 1973, those three snails had become 18,000, which the state located and destroyed.

Harboring the snails anywhere in the United States without a permit is illegal. Still, federal, state, and customs inspectors continue to confiscate GALS from pet owners and breeders, from schools, and from the occasional religious practitioner who (allegedly) drinks their mucus. So far, Florida’s giant-snail infestations have all been in cities and suburbs, far from farmland or protected ecosystems. But “we’re scared to death that it’s going to be found in the Everglades, some national park,” Skelley said. “That’s the worst case.”

Brazil is already there. An agronomist brought the snails into the country in 1988, most likely from Indonesia, in an ill-advised attempt to create a market for giant escargot. He sold snail-rearing kits to hundreds of hopefuls, who raised GALS in their backyards. “He had a very good intention,” Silvana Thiengo, the head of the Malacology Laboratory at the Oswaldo Cruz Institute in Rio de Janeiro, told me. But unfortunately, “Brazilians are not used to eating escargot.”

The snails were soon discarded on streets and in rivers. Now Thiengo receives snail specimens from around the country, and most of them are giants. They’re in urban gardens, tempting dogs to play with them; in croplands, eating bananas, beans, and soy; in the middle of the Amazon and in the country’s massive tropical wetland, the Pantanal. She and her colleagues worry about how large numbers of giant snails might compete with native species “for food, for space, for everything.”

A particular worry is the rat-lungworm parasite, which can infect humans, pets, livestock, and wildlife. In humans, the worms burrow into the brain and can cause coma and death. Although the disease is quite rare, Thiengo and her colleagues have found that human cases of it have been ticking up in Brazil. But at this point, the snails are so widespread in the country that little can be done about them. Although Brazil once ran national anti-snail campaigns, all have been phased out.

In some cases, trying to get rid of these snails has caused more problems than it has solved—and again, not through any fault of the snails. In Hawaii, where GALS were originally introduced by a gardener who thought they would make a fun yard ornament, the Hawaii Department of Agriculture tried to control the giant-snail population by introducing more than a dozen new carnivorous snails to the islands in the 1950s. One of the new species, the rosy wolf snail, developed a taste for Hawaii’s hundreds of species of endemic land snails, or kāhuli. It ate so many that more than half of those species are now extinct.

Still, once GALS have settled into the places they’ve invaded, something mysterious tends to happen, Cowie, the University of Hawaii snail researcher, told me. Their numbers die down. And the snails themselves get smaller. Hawaii’s snails are now at “a handleable level,” he said. “Maybe not acceptable, but you can deal with it.” The state has given up fighting them.

Cowie hasn’t. After it rains, he goes out into his garden with a pair of tongs and a garbage bag. He picks them up, giant out-of-place snail after giant out-of-place snail. He can’t stop the forces—globalization, ignorance, human folly—that ferry snails from place to place. But he can stop the snails in his own yard. He considers it his responsibility to move them one last time—into his freezer, and then into the trash.

Choose Your Overlord

2026-10-13T06:00:00-04:00

Updated at 7:00 a.m. ET on October 14, 2026

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

The Adriatic has been crossed by Odysseus and Caesar, by Venetian fleets and Napoleon’s navy, by countless men bent on dominating the world. It has seen its share of courage and folly. But one day in June 2025, off the island of Cres, it witnessed something genuinely new: a man worth $240 billion standing waist-deep in bubbles on the deck of a $500 million yacht, wearing a white bucket hat.

The man is Jeff Bezos, and the woman photographed in the string bikini next to him—hugging him, rubbing him with bubbles—is his fiancée, Lauren Sánchez. Bezos is shirtless, and it must be said: He is ripped. His hairless torso is so buff, his expression so amused, that he looks more like a louche cartoon character—a Club Med Ninja Turtle—than a 61-year-old corporate executive. In the foreground, an anonymous crew member, his face hidden from view and his thoughts forever unknown, mans a cannon blasting bubbles into the Croatian sky.

The foam party was a birthday celebration for Sánchez’s 19-year-old son, a prelude to the Bezos-Sánchez wedding five days later in Venice—with 200 VIP guests, hundreds of protesters, and an estimated cost of $50 million. The wedding consumed the internet for a week and was treated as a self-contained cultural event, a thing worth interrogating or gawking at on its own terms.

In fact it was the culmination of one of the weirder decades that any American billionaire has ever had. The whole republic, to be fair, has been pretty nuts, and Bezos has been both caught up in the insanity and an agent of its acceleration.

In 2013, Bezos, who founded Amazon and still owns roughly 8 percent of the company, purchased The Washington Post for $250 million. Two years later, Donald Trump accused him of using the paper as a tax shelter, to which Bezos jokingly responded that Trump should be launched into space. Trump then said at a rally, “If I become president, oh, do they have problems.” Trump became president, and the Pentagon steered a massive cloud-computing contract away from Amazon Web Services, while Trump repeatedly pressured the U.S. Postal Service to raise Amazon’s shipping rates.

In 2019, Bezos and his wife, MacKenzie, announced their decision to divorce. A month later, Bezos accused the National Enquirer’s parent company of attempted extortion after he said it threatened to release intimate photos, including what its chief content officer described in an email as a “below the belt selfie—otherwise colloquially known as a ‘d*ck pick.’ ” (The company said it had “acted lawfully” but would conduct its own investigation.) This is how much of the world learned about Lauren Sánchez, which was awkward because her brother Michael, a court later determined, had spoken with the tabloid about his sister’s relationship and sold—for $200,000—information, text messages, and a number of Bezos’s “ ‘racy,’ ‘explicit’ photos.” Bezos suggested, without evidence, that the plot had been instigated by the Trump administration and the Saudi government in retaliation for the Post’s reporting on the murder of its columnist Jamal Khashoggi by the Saudis.

That’s already a lot, even by billionaire standards.

Bezos then stepped down as CEO of Amazon in 2021; went to space aboard a rocket built by Blue Origin, the aerospace company he’d quietly founded two decades earlier; and began devoting the majority of his time to the company. Over roughly the same period, he commissioned the world’s largest sailing yacht and worked on building a 10,000-year clock inside a mountain. By the time of his second wedding, Bezos had transformed physically and politically—killing a Post endorsement of Kamala Harris for president, attending Donald Trump’s second inauguration in a seat of honor, reorienting the paper’s Opinion section around “personal liberties and free markets,” cheerleading Amazon’s reported $40 million deal for a documentary and future docuseries about Melania Trump.

People change. And if Bezos were just another middle-aged guy using his money to become unrecognizable to his former self—God bless. Enjoy the Met Gala.

But we live in a society where trivializing the actions of our wealthiest citizens is both unbeatable fun and a form of recklessness. For most of the past decade, the changes in Bezos that attracted attention were the meme-able ones: his body, his partner, his yacht. This made the transformation that mattered harder to see. We’ve been gawking at Bezos becoming a different kind of billionaire while missing his attempt to become a different kind of historical figure. The man who once rewired how the planet buys and sells everything has been chasing an even grander project—creating the infrastructure for an off-world economy upon which human civilization might someday depend.

If this sounds implausible and potentially menacing, you will not be surprised to learn that the person standing between Bezos and his ambition is the implausible and menacing Elon Musk, who has spent the past two decades building a commanding lead in space while pursuing a different vision of where humanity should go once it gets there.

Together, in fierce competition, these two billionaires are in the early stages of a race to vertically integrate the galaxy—a process that involves rockets, moon landings, enormous satellite constellations, and space-based data centers running AI models fueled by computing power that no terrestrial supplier can match. Both say they hope to preserve human life from the specter of climate change and expand mankind’s understanding of the universe. Both would also like to dominate a market that, unlike every other market they’ve previously conquered, has no ceiling and few established rules.

Whether any of this is wise is a question the rest of us are not really being asked. For most of human history, the power to decide where a civilization went next belonged to governments, monarchs, or despots. It’s a measure of how much has changed that a retail executive from Seattle and a South African–born engineer are well positioned to exercise that power. Nobody voted for either of them. And they may be the only ones who can check the other’s power. Jake Sullivan, Joe Biden’s national security adviser, described this to me as “a very weird situation.” You’ve got two men with unfathomable power and an unshakable faith that they’re destined to shape the world. But neither of them is “actually Lex Luthor,” Sullivan said.

“ … Right?”

Perhaps not. But any answer requires a deeper understanding of the two men involved. Musk, typically, has overshared. Bezos has been more secretive. To reveal his character and intent, you first have to get past the foam.

In the 1990s, when reporters began tracing the origin story of the first king of e-commerce, they made their way to Miami, where Bezos grew up, and to Ursula Werner, his high-school girlfriend. They had listened to ABBA together and he’d made her adoring flyers for her birthday, and she told those reporters something about him that, for years, seemed like just another cute bit of color: “The reason he’s earning so much money,” Werner said, “is to get to outer space.”

Chris Carroll / Getty

Jeff Bezos in 1999

In his high-school-valedictorian speech, Bezos quoted Star Trek—“Space, the final frontier”—and asked his audience to imagine vast orbital colonies where humanity could relocate Earth’s most polluting industries, an idea cooked up by the Princeton physicist Gerard K. O’Neill in the 1970s. “The earth is finite,” Bezos explained to a local reporter. “And if the world economy and population is to keep expanding, space is the only way to go.”

All of this nerd enthusiasm was out in the open until 2000, when Bezos created Blue Origin Enterprises. He based the company a short commute from Amazon’s Seattle headquarters and kept its existence a secret for three years, until Newsweek discovered it. This was not the only strange thing about Blue Origin.

For a while, the company’s sole employee was the science-fiction writer Neal Stephenson, best known for Snow Crash, a 1992 novel set in a near-future America where the government has collapsed and corporations run everything. In a blog post, Stephenson wrote that his title at Blue Origin was “Member, Technical Staff,” and that he “performed operations as prosaic as punching down Ethernet cable, operating a plate grinder, and passivating rocket parts.”

But that was just part of the job. “The bulk of my efforts,” Stephenson wrote, “were devoted to investigating possible alternatives to conventional rockets.”

Bezos wanted to know if almost a century of rocket science had gotten the basic question wrong—whether there was a smarter way to get to space than attaching a tube to massive amounts of fuel and setting it all on fire. Stephenson and the handful of other employees who were eventually hired ran software simulations that would either validate the conventional approach or, Bezos hoped, find its replacement. This is a pretty wild way to spend your money.

Three years of experiments proved that Wernher von Braun and Robert Goddard hadn’t missed anything. Only then did Bezos hire the former NASA engineer Rob Meyerson and buy 165,000 acres of Texas scrub brush where rocket prototypes might someday be launched without scrutiny.

Eventually Meyerson would be promoted to president—but of what, exactly? In the early years, Blue Origin had no customers or finished product, and little of what ordinarily makes a company a business. Bezos left the CEO job vacant and imposed no conventional timetable. It did have a tortoise mascot and a motto—“Gradatim ferociter” (“Step by step, ferociously”)—that reinforced its commitment to getting to space slowly, secretly, and stubbornly. Employees were focused on whiteboarding a tourism rocket, but Bezos told Meyerson not to worry about sales or pitching government agencies; he didn’t want to be held hostage by customers. It was “a science experiment for the first 15 years,” Tom Jones, Blue Origin’s vice president of human resources from 2016 to 2019, told me.

Contrast this with the hyper-logical way that Bezos had conceived Amazon. In an early interview, Bezos explained that the internet was growing at 2,300 percent a year; to capitalize on that growth, he needed a product category with more variety than any physical store could stock. There were more than 3 million active book titles worldwide. Therefore: books.

If Amazon was the company that Mr. Spock would create, Blue Origin was where Bezos got to lead with his heart and be Captain Kirk. He’d take days off from Amazon for field trips, packing Blue Origin employees onto his private plane and dropping in on companies with cultures or engineering accomplishments that he admired. Gradually, Bezos cleared all of his Wednesday afternoons and many of his Saturdays to work at the Blue Origin office. When Bezos wanted to learn about heat-transfer codes—software that models how heat moves through a rocket engine; the difference between an engine that survives its own combustion and one that doesn’t—Meyerson hired an employee to tutor him.

I asked Meyerson, who is now the CEO of Interlune, which aims to mine the moon for helium-3, if managing Bezos’s enthusiasm ever got in the way of accelerating Blue Origin’s progress. “My job was to do both,” he said. “The time that he put into Blue was a gift to us, because the more he understood, the more he was willing to support what was needed for the long run.”

And yet, when it came to money, the accountant overruled the dreamer. Bezos was rich—Forbes’s 100th-wealthiest American—but not send-my-money-to-the-moon rich. He didn’t want “to spend the kind of money he’s spending today,” Meyerson said. Blue Origin would work through simulations, models, and small prototypes—staying within its means until Bezos was ready to scale up.

This would have been fine if Blue Origin were operating in a vacuum. But two years after Bezos founded his dream company, someone else started a company with a dream just as large. And he was not interested in moving slowly.

Impatience is part of Elon Musk’s mystique. So is an eagerness to break and remake things: the economics of car buying (Tesla), the operations of the federal government (DOGE), the last remaining norms of social media (X—now a platform for disturbing racial theories).

But Musk didn’t set out to revolutionize aerospace and become the world’s most successful rocket builder. Originally he just wanted to be a rocket buyer.

Dan Tuffs / Getty

Elon Musk in 2008 with an early model of SpaceX’s Dragon capsule, designed to carry a crew or cargo into space atop a rocket

In 2001, Musk announced that he would send a robotic greenhouse into space, scoop up some Martian soil, and use it to grow a plant. This scientific publicity stunt may not go down as Musk’s best idea, but his attempt to buy refurbished intercontinental ballistic missiles for the Martian-soil gambit led him to Moscow, where, according to Ashlee Vance’s biography, Elon Musk, a Russian engineer spat on him (literally) and his PayPal millions. Flying back from a second unsuccessful trip, Musk opened a spreadsheet and realized that he could build a rocket for less than the $8 million the Russians wanted to charge him to buy one—and for less than anyone had ever built a rocket before. He founded SpaceX a month later.

It’s easy in hindsight, and with this year’s $1.8 trillion SpaceX IPO, to say that Musk had a better plan than Bezos to get to space. What’s inarguable is that Musk had a faster plan. He staffed up by cold-calling college aerospace students and handing them ludicrous deadlines—the first launch was scheduled for 15 months after the company was created. Speed gave SpaceX swagger, and that swagger gave its young employees faith. But the company’s most important virtue was thrift. Not as a means of preserving Musk’s wealth, but because lowering costs was the entire point of the business.

For decades, getting a person or a satellite into orbit was so expensive that only governments and the largest defense contractors could afford it. Musk believed that this had little to do with physics; unlike Bezos, he spent no time wondering whether there might be a fundamentally different way to get to space. Instead, he looked at launch costs the way a deregulator might have looked at long-distance telephone rates in the 1970s: not as a law of nature, but as the product of the people who had gotten there first and decided what to charge.

If SpaceX could create a reusable rocket capable of driving launch costs down far enough, it wouldn’t just swallow all the existing demand—it would transform space from a cartel into a free market. Satellite builders, scientists, start-ups, and, eventually, anyone with a payload and a business case could participate. Access to space would become as abundant as telephony. And then all sorts of innovation might happen.

Bezos had a similar vision, and both men believed that reusable rockets were the key to making space affordable. But while Blue Origin was moving step by step, SpaceX skipped every step it could. Employees were celebrated for outrageous hacks, such as figuring out that off-the-shelf car-wash valves worked just fine for rockets if you swapped their seals. All of that ingenuity led to Musk’s promise that SpaceX’s first rocket, Falcon 1, would carry a 1,400-pound payload into orbit for $6.9 million—75 percent less than any competitor.

He nearly went bankrupt doing it. The first three Falcon launches failed. By 2008, SpaceX and Tesla were burning cash. Musk was barely solvent enough to make one more launch attempt. On September 28, Falcon 1 reached orbit, making SpaceX the first privately funded company to launch a liquid-fueled rocket into space. The timing was exquisite. Three months later, as the global financial system teetered and Musk was days away from missing payroll, NASA awarded SpaceX a $1.6 billion contract to resupply the International Space Station. The company had gone from near death to indispensable in a single quarter.

A former senior NASA official who visited both companies in their early years described Blue Origin to me as an academic lab and SpaceX as a pirate ship. “SpaceX was scrappy,” the official said. “I went to the Blue factories in Kent, Washington, and they were these beautiful, massive buildings with the world’s best 3‑D printers. And I was like, Where are the people? Then you go to SpaceX in Hawthorne, and there’s these young people running around; everybody’s greasy.”

The differences were hardly lost on the people watching from Blue Origin. SpaceX had taken Amazon’s “get big fast” playbook and applied it to rockets⁠—staffing up, sprinting to deadlines, and treating its failures as content rather than embarrassment. The launch explosions and misfires it shared on social media didn’t make SpaceX look reckless; they made it look humble. At Blue Origin, the contrast was unmistakable: SpaceX seemed to be having more fun.

SpaceX soon proved that it could reuse rocket boosters—the bottom part where the flames shoot out. Its first reusable rocket, Falcon 9, launched and landed in 2015. By May 2016, SpaceX was on its way to becoming the UPS of low Earth orbit. Satellite customers could buy launches at published prices, and Falcon 9 was beginning to fly on something approaching a regular schedule. NASA was already relying on SpaceX to shuttle cargo to the International Space Station; the Air Force had certified Falcon 9 for national-security missions and had just awarded the company its first military-launch contract.

In West Texas, Blue Origin was completing the first takeoffs and landings of its New Shepard tourism rocket. Those missions were uncrewed and suborbital—which is to say, the rocket went to space the way a ball tossed upward in a backyard goes above the roofline: fleetingly. On Twitter, Musk pointed out how far ahead SpaceX was, beginning a series of exchanges in which two of the world’s richest men engaged in public rocket-measuring contests. Musk won most of these, which makes sense, because Musk was winning.

People who have worked closely with Bezos bring up two defining characteristics. The first is a comfort, bordering on delight, at being misunderstood. “I think he enjoys it,” a former Amazon senior executive told me. “You have no idea how stubborn he can be in defense of an idea only he believes. And of course a lot of times he’s right.” The second is that Bezos does not like multitasking. He’s told audiences that as a little kid in Montessori school, he was so resistant to leaving things undone that he had to be picked up in his chair and moved to the next activity. Bezos prefers to direct his focus, as intensely as possible, on one problem at a time. As long as he remained the CEO of Amazon, that problem was Amazon. For the first 15 years of Blue Origin’s existence, the company had to accommodate that reality.

Whether it was Musk’s taunts or the unignorable fact that SpaceX was racing ahead, Bezos finally changed course in 2016. During one of his regular Saturday meetings at Blue Origin, he asked a question that nobody in the room wanted to answer honestly: Why hadn’t Blue Origin directly bid on an Air Force contract for rocket-engine development that had just gone to SpaceX? The answer: Because you’ve always told us not to.

Bezos delivered a new mandate to Meyerson: “From now on, we go after everything that SpaceX bids on.”

When people in aerospace talk about SpaceX’s advantages, they nod to Musk’s maniacal expectations, his engineering instincts, and the particular quality of his attention. But Musk’s attention has at times been distributed across Tesla, X, the neurotechnology company Neuralink, the tunnel start-up the Boring Company, the federal government, and 14 children by four women.

Musk’s life is a browser with too many tabs. X is the tab that’s always open. It’s where he challenged Mark Zuckerberg to a cage match, offered to give Taylor Swift a child and guard her cats “with my life,” and feuded with (partial list) Kanye West, the British government, the Australian government, the Brazilian supreme court, and a cave diver in Thailand he called a “pedo guy”—leading to a defamation lawsuit that Musk won after his legal team argued that the phrase was a generic insult. It’s also where he regularly posts memes and has promoted anti-Semitic content that makes him, justifiably, a pariah among people who believe society is plenty coarse enough without him. (SpaceX did not respond to interview requests for this story.)

The point is, Musk is no more focused than Bezos. The difference is that he’s never needed to be. This is because, in 2002, he had a three-minute conversation with Gwynne Shotwell. Shotwell is 62, and remarkable in several ways—a woman in a male-dominated field, a proud former high-school cheerleader with degrees in mechanical engineering and applied mathematics. But her unique gift is the ability to read her boss’s mind while ignoring his id. “I like his in-person self better than his Twitter self,” Shotwell said in a 2022 talk at Stanford.

Shotwell is self-aware and funny. Not aerospace funny—actually funny. She grew up in the “boonies” north of Chicago and was a sophomore in high school when her mother dropped her off at a Society of Women Engineers event. She decided to become a mechanical engineer that day in part because one of the presenters “had the best suit on, and she had great shoes, and her bag matched her shoes,” Shotwell told a podcast interviewer with delight.

Shotwell was in the middle of a career in traditional aerospace when she went to lunch with a former co-worker who had been poached by SpaceX. She was introduced to Musk on a tour of the office, and received a job offer to be vice president of business development that afternoon. For weeks she hesitated before pulling over on the Los Angeles freeway, calling Musk, and telling him that she was an “effing idiot” for not accepting sooner. He laughed and welcomed her to the team.

Since then, Shotwell and Musk have built one of the most drama-free partnerships in 21st-century business. He oversees engineering and technology while she takes care of everything else—sales, operations, manufacturing, workforce vibes. While Musk was supervising the 2008 do-or-die Falcon-rocket launch, Shotwell was at an industry conference in Glasgow, watching the livestream at 2 a.m. in her hotel room while writing the company’s bid to resupply the International Space Station. When the rocket reached orbit, she ran down the hallway to rouse her colleagues and broke into the hotel bar to drink warm champagne.

In 2015, having proved that SpaceX could reliably deliver payloads to space, Musk announced plans for a constellation of thousands of satellites, now known as Starlink. The pitch had a noble dimension—SpaceX would bring high-speed internet to parts of the world that fiber-optic cables had left behind. But this disguised its real purpose. SpaceX’s first act—flying rockets for hire—was not profitable enough. Getting in on boring old telecommunications would give the company a way to monetize its own launch capability. Starlink was created to be the thing that would pay for everything else.

The following year, Musk revealed the full scope of his plan: SpaceX would build a fully reusable two-stage rocket and spacecraft standing just over 400 feet tall and powered by 42 engines, with a capsule—the top part—capable of carrying 100 people to Mars (or 60 Starlink satellites into orbit). “The objective is to become a spacefaring civilization and a multiplanet species,” Musk said at the time. “The main reason I’m personally accumulating assets is to fund this.”

Previous spacecraft carried only what they needed for a mission. This would be like launching a Costco.

Scale is the great paradox of rocket science. A rocket twice as large as its predecessor isn’t just twice as hard to build; as the rocket grows, so do the forces trying to tear it apart. At the size Musk envisioned, the fully fueled vehicle would weigh more than 10 million pounds, more than one and a half times the weight of the Saturn V, which carried astronauts to the moon. Lifting it would require engines more powerful than anything SpaceX had built before, with roughly three times the pressure of the engines powering its Falcon rockets—conditions so extreme that SpaceX would need to invent new manufacturing techniques simply to build the combustion chamber. And Musk intended to reuse everything: Both the giant booster and the spacecraft above it would return to Earth and fly again.

The design would change repeatedly—first landing legs, then enormous mechanical arms—but the economic premise remained the same. If full and rapid reuse worked, Musk believed that the cost of reaching orbit could eventually fall below $100 a kilogram, which would make the things that had always died on a spreadsheet—lunar bases, enormous satellite constellations, perhaps even human flights to Mars—possible. The architecture was elegant in its circularity. Starship, as the giant rocket and spaceship would come to be known, would make Starlink, the satellite array, far cheaper to deploy. Starlink would then generate revenue to pay for a fleet of Starships. And Starships, eventually, would get humans to Mars. Engineering, commerce, and Musk’s interplanetary ambition, all stacked inside the same rocket.

People at Blue Origin first noticed the woman around 2018. She was seen at the Cedarbrook Lodge near the Seattle airport, a few minutes before Bezos arrived for drinks with executives. In West Texas, she began showing up as a contractor on the helicopter video team that documented New Shepard launches.

Lauren Sánchez Bezos is impossible not to notice because her lifelong goal, exuberantly pursued, is to be noticed. Sánchez Bezos grew up in New Mexico, learned she was dyslexic in community college, hustled her way into the University of Southern California, and for several years proceeded through the carnival of gigs (local news, Extra, So You Think You Can Dance) available to attractive and relentless people who want to be on television. She had a child with the former NFL tight end Tony Gonzalez before marrying Patrick Whitesell, the Hollywood agent and power broker, with whom she had two more children. In 2016, she got her helicopter pilot’s license and started an aerial-production company.

The first public encounter between Bezos and Lauren Sánchez was at a December 2016 Amazon Studios party for the grim kids-die-in-a-fire movie Manchester by the Sea. It must have been like meeting Jayne Mansfield at a shiva. Both were married at the time, and what happened next seems clear enough, even without the benefit of any photos from the National Enquirer. “Wonderful madness” is how Barry Diller, the media mogul and Bezos’s close friend, describes the relationship.

Understanding the force and fury of the response to the couple requires some unpacking, and it starts with the woman Bezos left. MacKenzie Scott studied creative writing at Princeton under Toni Morrison, did Amazon’s early bookkeeping out of Bezos’s garage, and has spent the years since their 2019 divorce giving away her share of the proceeds—more than $26 billion to more than 2,000 organizations—while largely erasing any evidence of her generosity. In August, she released a complicated third novel on Substack.

“I have great respect for and fondness for MacKenzie,” Diller told me. “You can see how she’s conducting her charitable work, which is remarkable—but it’s all quiet and private. And that’s MacKenzie.”

J. Countess / Getty

Bezos with his first wife, MacKenzie Scott, in 2005

Cindy Ord / MG24 / Getty

Bezos with Lauren Sánchez at the Met Gala in 2024

Lauren Sánchez Bezos enjoys being photographed at parties and events, and her picture is prominent on the website for the Bezos Earth Fund, a $10 billion investment in protecting “the best planet in this solar system,” where she’s vice chair. She’s written three children’s books—The Fly Who Flew to Space, The Fly Who Flew Under the Sea, and The Fly Who Flew to the North Pole—and if a brick wall were the only thing standing between her and the best-seller list, you wouldn’t want to be the wall.

“People assume she’s going to be Cruella, and she’s not,” a woman who’s been friends with Sánchez Bezos for a decade told me. “She’s smart and kind, and she has real accomplishments. But it’s confusing because snobby people look at her big, vulgar boobs, and she’s like, Yeah! Look at ’em! She likes being sexy. She just wants everyone to have a good time, and she doesn’t mind being the good time.”

Bezos and Sánchez Bezos make each other happy. This is not nothing, but the wonderful madness has had its costs. There was an alleged blackmail attempt to manage, a 25-year marriage to unwind, and a company that still depended on Bezos’s attention. The scandal “had a huge impact at Blue Origin,” Alexandra Abrams, the former head of employee communications, told me. (Abrams, who was fired in 2019, later went public with complaints about flight safety and criticized leadership, Bezos included, for what she described as a “toxic” culture; in response, the company said it had no tolerance for discrimination or harassment and stood by its safety record.) Abrams said that the news of Bezos’s relationship with Sánchez “started to break the goodwill of a lot of the senior employees at the company. They were shocked that Jeff would comport himself in that way.”

Only Bezos can say how much of his transformation has been driven or enabled by his new wife, and how much was always there, waiting for permission. And Bezos, who declined multiple requests for an interview, isn’t talking.

A few weeks after the Manchester by the Sea party, Trump’s NASA transition team invited Blue Origin to a meeting. With Bezos’s decree to start competing, Bretton Alexander, the director of business development and strategy, scrambled to prepare.

Mike Pence was passionate about space. Trump understood the MAGA value of rockets and astronauts. So, the transition team asked: What big idea could Blue Origin offer that would be ready by the electorally convenient date of 2020?

“I didn’t see the question coming,” Alexander told me. But he blurted out, “We’ve landed New Shepard in the desert five times in the last year. We can land something on the moon, because we have the engines and we have the technology.”

The last person to walk on the moon was the Apollo 17 commander, Gene Cernan, who let a crewmate climb the ladder before him so that he could utter the following: “We leave as we came and, God willing, as we shall return, with peace and hope for all mankind.” That was in 1972.

There’s no single explanation for why America hasn’t returned, but there are several timely reasons to go back. The moon is the most convenient place in the cosmos: three days away, without the communication lags that make deep-space operations so complicated. It has water-ice deposits, which theoretically could sustain human life and be split into hydrogen and oxygen, also known as rocket fuel. And because the moon’s gravity is one-sixth of Earth’s, less fuel is required to launch from its surface and explore the rest of the solar system. For 4.5 billion years, celestial objects have used the moon as target practice, enriching the lunar soil with most every rare mineral the solar system has produced. Half its face is shielded from Earth’s electromagnetic noise, which makes it valuable for radio telescopes and, as national-security officials will warn in varying degrees of specificity, potentially useful to America’s adversaries. The Chinese have committed to landing astronauts on the moon by 2030.

For all of these reasons, Alexander’s pitch appealed to the transition team. That it nudged federal space policy toward the moon—elevating Bezos’s high-school dreams while usurping Musk, who hoped the Trump administration would back a Mars mission—was, for Blue Origin, the chef’s kiss. Alexander called Rob Meyerson right away. “I said, ‘Oh shit, I just created a program.’ ”

Greg Autry, who now leads international space policy at the State Department, was on the transition team. “We knew they didn’t really have this thing in their garage,” Autry told me. The company had announced New Glenn, its first orbital rocket, earlier that year, but the new rocket had not even carried a commercial payload, let alone human beings. But Blue Origin had “what appeared to be a simple and relatively quick way to get there, with a reusable system.” NASA, perhaps, was also beginning to realize that having a single dominant space-launch provider might in itself be a risk.

Animosity continued between Trump and Bezos, who remained publicly critical of the president’s immigration policies. The Pentagon awarded a cloud-computing contract worth up to $10 billion to Microsoft when Amazon Web Services was the undisputed market leader. Amazon sued—and asked for Trump to be deposed, arguing that the president’s dislike of Bezos had corrupted the bidding process. (The Pentagon’s Office of Inspector General found no direct evidence of White House interference but noted that it could not complete its assessment, because senior officials had asserted “presidential communications privilege.” The suit was thrown out in 2021 when a new contract was awarded to both Amazon and Microsoft.)

But meaningful friction appears to have stopped at Blue Origin’s door. A few months after Alexander’s meeting, Bezos flew to Washington to personally present his company’s proposal to the transition team. “That didn’t necessarily happen with the other companies,” Autry said. “You didn’t see their CEO or funder take the time to do that.” So began a drip of white papers and study contracts that, over several years, led to $592 million in government funding. Other than a 2019 meeting where Pence’s chief of staff told a Blue Origin executive that the company had a “Washington Post problem”—a detail later reported in The Washington Post—no one I spoke with could recall an instance when the Trump administration’s hostility toward Bezos translated into prejudice against Blue Origin.

There was plenty of turbulence at Blue Origin during this period, but it was self-inflicted, and well overdue.

At a Saturday leadership-team meeting in 2017, Bezos was presented with a memo outlining Blue Origin’s annual budget request. Senior leadership wanted more than $1 billion, double the current budget. All of which would come out of Bezos’s pocket.

When Bezos finished reading the memo, “he walks out, doesn’t say a word,” Tom Jones, Blue Origin’s former vice president of human resources, told me. “About 15 minutes later, he walks in and he says, ‘Why did I not know that you were going to be making a request this large until today?’ ” Jones described Bezos as calm, even affectionate; Bezos said that he loved the people in the room but that the request was “craziness,” and wildly disproportionate to their achievements. The company had convinced Washington to take its lunar ambitions seriously, but it had yet to put anything into orbit. “It’s probably one of the things that might have led to him thinking, like, Okay, I’ve got to make a change here.”

Bezos responded by bringing in Bob Smith, a veteran of the aerospace company Honeywell, as Blue Origin’s first CEO in September 2017. Smith told me he arrived at a company with no steady revenue stream and no supply chain: “Engineers would design a part, put the order out, and then it’d show up by mail back on their desk.” It was essentially “a prototype shop.”

Blue Origin might not have had a supply chain, but it did have a cosmology. Bezos “is God at Blue Origin,” Alexandra Abrams said. “Jeff’s big message was that it’s too late for climate-change efforts on Earth, and we need to be able to leave. We need to sustainably go into space and build resources, have a base on the moon, which would act as a gas station to the rest of the solar system. At the company, everyone felt like we were saving humanity.” (After publication of this story, Bezos said, through a spokesman, that he does not believe that it is too late for Earth.)

But God remained an intermittent presence. Bezos had tried to extract himself from Amazon for years—installing deputies, restructuring leadership. COVID wrecked that project; Amazon’s logistics network buckled in the spring of 2020, and Bezos had no choice but to dive back in. Meanwhile, Blue Origin was in the middle of a significant expansion. Bezos had authorized Smith to make thousands of new hires and build new factories, a transformation that required exactly the kind of consistent attention he couldn’t provide. Bezos had made himself indispensable by being the only person who knew exactly what he wanted. He had never found his Gwynne Shotwell—or perhaps he’d never believed the job could be done by someone who wasn’t him. In his absence, Blue Origin was growing fast and bidding on everything, but it was still chronically late.

New Glenn, Blue Origin’s version of the Falcon, still wasn’t flying by the time SpaceX began developing Starship. SpaceX announced its Starlink constellation in 2015, put prototypes into orbit in 2018, and began deploying at scale in 2019—at which time Amazon had not yet launched a single prototype for Project Kuiper, its we-can-do-satellites-too Starlink competitor. Bezos wasn’t chasing Musk so much as arriving at places Musk had already left.

The pattern held as SpaceX started flying astronauts to the International Space Station. Blue Origin’s answer, when it finally came, was personal.

On July 5, 2021, Bezos vacated the CEO seat at Amazon. Two weeks later, he flew to space on New Shepard, the tourism rocket his company had spent years building, alongside his brother, Mark; the aviation pioneer Wally Funk; and an 18-year-old whose father, a Dutch financier, had bought him a ticket. Bezos stepped out of the capsule in the West Texas heat wearing a cowboy hat and declared it the best day ever.

Joe Raedle / Getty

Blue Origin’s New Shepard rocket lifting off in 2021, with Bezos aboard

The Bezos who returned to Blue Origin was different. He seemed determined to run the company the way he had once run Amazon—present, demanding, involved in every significant decision.

Smith’s 2023 resignation took no one by surprise. He had grown Blue Origin from roughly 850 employees to more than 10,000, accumulated billions of dollars in customer orders, and built enormous manufacturing facilities in Huntsville, Alabama, and Cape Canaveral, Florida. But New Glenn continued not to launch. And when NASA, under the Biden administration, finally awarded the contract for its Human Landing System—the moon lander that Bretton Alexander had improvised into existence—it did not go to Blue Origin. SpaceX, proposing a lunar lander based on its Starship architecture, undercut Blue Origin’s $6 billion bid by more than half.

What followed was also personal—and a little embarrassing.

Blue Origin filed a protest, prompting Musk to tweet, “Can’t get it up (to orbit) lol,” and to offer Bezos some unsolicited management advice. “I think he needs to run BO full-time for it to be successful,” Musk told The Washington Post. “Frankly, I hope he does.” Bezos offered to cover $2 billion of the lunar-lander costs, then $3 billion. When NASA ignored Bezos, Blue Origin sued. Musk, again: “You cannot sue your way to the moon.” The courts agreed. The saga ended when NASA, nudged by Senators Patty Murray and Maria Cantwell of Washington, awarded Blue Origin a $3.4 billion contract to build a competing lander.

Bezos decided that the next person to lead Blue Origin didn’t need to understand rockets or politics; they needed to understand him. Dave Limp had run Amazon’s consumer-devices business—Kindle, Echo, Alexa—for nearly 14 years. Limp had also overseen Project Kuiper, where he pushed to spread the risk among multiple launch providers rather than depend on Blue Origin alone. He even said Amazon would be “crazy” not to consider SpaceX, given its track record.

In his final interview before accepting the job, Limp asked a single question: Was Blue Origin a hobby or a business? Because if it was a hobby, he wasn’t interested. Bezos told him, “It’s obviously a business. It has to be a business or it’s not going to work.”

During the decade when Bezos was reorganizing his life around space, he walked away from a promise he’d made on the day he bought the Post : “The paper’s duty will remain to its readers and not to the private interests of its owners.”

The killing of the Harris endorsement, the subscriber cancellations, and the cascade of staff reductions—all of this has been well chronicled, because there’s no beat that journalists swarm quite like a competitor’s newsroom. But the core questions about Bezos and the Post remain: Why does he appear to have inverted his own promise? Was this an act of betrayal or of neglect? And, in the end, is there a difference?

When Bezos first showed up in Washington, the paper’s circulation was plummeting. One former Post executive described Bezos’s 2013 arrival to me as a kind of rapture. He was a forceful presence, respectful of the wall between the editorial and business sides, but relentless about improving the customer experience. Those first two years were “absolutely fucking magical,” the former executive told me. The Post went on a sustained run of profitability, and subscriptions surged throughout the first Trump administration.

But traffic and urgency declined in the Biden years, while COVID and Bezos’s other interests made him a phantom in Washington. He surfaced to participate in hiring Sally Buzbee to succeed Marty Baron as executive editor in 2021. In 2023, the Post lost $77 million. Bezos surfaced again, to replace the paper’s longtime publisher with Will Lewis, a British-tabloid veteran.

There were other candidates—a person close to the hiring process told me that one of them found their conversation with Bezos baffling. The Post’s owner couldn’t answer basic questions about the newspaper. Bezos seemed inattentive, and said that his priorities would not change: He was focused on Blue Origin and Lauren. Bezos appeared to think that his lack of involvement would be attractive; the next publisher would have free rein.

I was a columnist at the Post for a year, and had a few conversations with Lewis early in his tenure. He’d been a journalist before entering the wolf pack of Rupert Murdoch’s executive program, and among his talents was an unusual ability to throw back pints with reporters while keeping the secrets of powerful men. I found him to be a delightful, charismatic scoundrel.

Many colleagues were charmed by Lewis. But within a few months, his fury over the newsroom’s coverage of accusations that he’d helped conceal or destroy evidence while managing a scandal at Murdoch’s News International⁠—accusations he has denied—had soured the dynamic. Equally concerning was that Lewis barely demonstrated any interest in his job. The Post lost $100 million in his first year—and an astonishing 60 million monthly visitors from November 2020 to November 2024. But Bezos wasn’t paying much attention to Lewis or the newsroom; he regularly canceled meetings with top editors.

Bezos seemed more interested in the Opinion section. He never forbade the editorial board from writing what it wanted, and seldom voiced strong feelings about politicians or parties. But his interests and ideology were evolving, and as the owner, he hoped the editorial board would move in step. Bezos seemed to enjoy speaking with the section’s editor, David Shipley, about creating a forum for pragmatic ideas and policy solutions. Wonky centrist stuff.

That hands-off approach changed with the spiking of the Harris endorsement 11 days before the 2024 election—abruptly ending a half-century-long tradition and provoking 250,000 subscription cancellations. Those sympathetic to Bezos say that the last-minute timing wasn’t his fault. Lewis had told the editorial board that he would handle getting Bezos’s approval. But as the election drew closer, it clearly had not been handled—or was not handle-able. “The timing of their refusing to endorse was too close to the election,” Barry Diller told me. “If the Post had been his only job, he would have done that much earlier.”

Lewis, who stepped aside in February 2026, did not respond to requests for comment. He certainly deserves blame for some of the destruction. But the most consequential decisions were not his. It was not Lewis who withheld the endorsement. It was not Lewis who complained to Trump about the “terrible” businesspeople at the Post shortly after Trump’s reelection, according to Maggie Haberman and Jonathan Swan’s Regime Change. It was not Lewis who dined with the president on the very day he announced that the Opinion section would focus exclusively on “personal liberties and free markets” (a decision that prompted Shipley’s resignation).

Bezos has framed the changes at the Post as a matter of necessity and opportunity, writing in an op-ed that he wanted to reposition the paper for Middle America. Journalism isn’t a charity, and Bezos’s own politics have clearly shifted. He is, after all, the owner.

Kent Nishimura / Bloomberg / Getty

A rally outside The Washington Post’s headquarters in February after the company announced layoffs

But Bezos plainly appears to have made a Faustian bargain. Just as Blue Origin was getting its act together, a second Trump presidency dawned, with Bezos’s chief antagonist, Elon Musk, in a position of extraordinary influence. Musk had spent $288 million to help elect Trump and other Republicans. He was a senior adviser to the president. Amazon had fought the first Trump administration for more than a year over a cloud-computing contract. The administration had been lobbied hard by Oracle, Amazon’s cloud competitor, whose co-founder had reportedly given about $45 million to a nonprofit supporting Trump’s reelection. Oracle had also produced a docket of baseless conspiracies about Amazon that found its way to the president’s desk.

Marty Baron was the executive editor of the Post from 2013 to 2021, a gilded era that produced 11 Pulitzer Prizes and vital reporting on Edward Snowden, the Trump Access Hollywood tape, and the Robert Mueller investigation. “Everything I saw in my years there was that he truly believed what he was saying about democracy and the press,” Baron told me. “But I don’t think he thinks these other businesses could survive unless he finds a way to successfully navigate through the Trump era. And if the Post is the thing that has to pay the price, then, well, that’ll have to be the thing that pays the price.”

Donald Trump is not a simple transactional figure. He gets what he wants, but he also warps reality in the process. Countless executives have learned that successfully dealing with him requires both capitulation and an insistence that no capitulation has occurred. If Bezos and Trump did come to an arrangement, it would make Bezos a hypocrite twice: once for breaking his promise to Post readers, and once for pretending he hadn’t.

Jeff Bezos also tends to get what he wants—and unlike Trump, he’s willing to wait years for it. He’s comfortable with being misunderstood. Which is another way of saying that there’s more than one way to look at a Faustian bargain: not in terms of what must be sacrificed, but in terms of what might one day be gained.

“NG‑3, baby!” whooped a man in a Titusville, Florida, 7-Eleven in April, shortly before flashing me his admittedly glorious NASA-logo calf tattoo. People who live along Florida’s 72-mile Space Coast rarely end up there by accident. Some work in aerospace; the rest seem to wish they did. They buy condos and arrange their schedules for optimal viewing of the rockets rising from the Kennedy Space Center. As subcultures go, they’re somewhere between Deadheads and bird-watchers: obsessive, migratory, and willing to wait a very long time for the thing they came to see.

I was in town at the invitation of Blue Origin to witness the third launch of New Glenn—NG‑3. This would be the first time Blue Origin reused a booster from a previous launch, a crucial step. The nonreusable upper part of NG‑3 would be carrying a large telecom satellite for AST SpaceMobile.

Dave Limp was in from Washington State to watch the launch and schmooze the satellite guys. Limp is 60, and in his sensible jeans, glasses, and Blue Origin three-quarter zip, he looks like the most reasonable man in every room he enters. When he started the job, he’d wondered if a consumer-electronics guy was the right fit for a rocket company, but he’d come around to Bezos’s diagnosis—that Blue Origin needed more disciplined decision making, more manufacturing expertise, and more Bezos. “Jeff can be annoying at times, but he’s right a lot,” Limp told me. “And he was right on this.”

The standard that Limp is being held to has a name: Gwynne Shotwell. Limp is an admirer, and has arranged his duties accordingly. “My job is to have Jeff live in the future,” Limp said. “So if he has to hear about an HR issue or a financial issue or a schedule issue, I have not done my job.”

His other job is to get rockets up, fast. The Federal Communications Commission requires satellite operators such as AST to put a certain percentage of their constellations in space by specific dates, or they risk losing authorization for the satellites they failed to deploy. Amazon was supposed to deploy half of its 3,236 satellites by July 2026. The FCC granted a waiver in June after Amazon explained that it was producing satellites faster than any of its launch providers—including Blue Origin—could get them into space.

A year earlier, the New Shepard tourism rocket had carried a team of six women, including Lauren Sánchez and Katy Perry, into space. Now Limp put it on indefinite hiatus. All focus, he told me, was on the lunar lander and getting New Glenn rockets produced with assembly-line speed: “The ‘step by step’ component of our motto—let’s fast-walk it, you know?”

At 7:25 a.m. the next morning, NG‑3 lifted off from Blue Origin’s launchpad at Cape Canaveral Space Force Station. It did not blast off like the rockets of old. Accelerometers and gyroscopes fed its position to the guidance system as computers continuously adjusted thrust and steering, making millions of pounds of rocket seem almost to levitate. When the rocket disappeared, it left behind a long, knobby cloud—a strangely beautiful albino Cheeto. Six minutes later, the booster landed a few hundred miles away in the Atlantic Ocean on the drone ship Jacklyn, named for Bezos’s late mother. Everyone went nuts.

I didn’t find out for a few hours that the mission had been a dud. One of NG‑3’s engines had failed to produce enough thrust and dropped AST’s satellite in the wrong orbit, too low to be useful. AST had insurance, but the customer doesn’t care if you save money by reusing your booster; it cares that you get the job done. Blue Origin had not.

At 12:12 p.m., the day got a little worse: SpaceX announced that it had just landed its Falcon 9 booster for the 600th time.

A month later, a New Glenn rocket undergoing a static-fire test caused the largest explosion in the history of Cape Canaveral, badly damaging Blue Origin’s only operational launchpad and temporarily grounding the program.

Only two people have the means and ambition to vertically integrate the galaxy, and one of them has the upper hand. In 2025, SpaceX launched 83 percent of the mass sent into orbit—about 10 times more than its nearest competitor, the Chinese government. Starlink’s 11,000-plus satellites account for roughly two-thirds of all active spacecraft orbiting Earth.

Limp, at least, knows all of this. He acknowledged that SpaceX is “by far the leader.” But, he said, “there’s so much demand for space right now that if we get to the scale and the execution of SpaceX, the world will be a happy place. And that does not mean Gwynne and Elon have to fail.”

The exact size of that demand is debatable. SpaceX and Blue Origin, along with a handful of smaller competitors, provide launch services to other companies and the federal government, but most of the demand comes from their own or affiliated satellite fleets. In finance, this is called captive demand, and it’s generally regarded with suspicion. In commercial space, it’s called vertical integration, and it’s the whole point.

This is why the SpaceX prospectus reads like both a filing with the Securities and Exchange Commission and a transcript of an investment banker’s weekend on mushrooms. “Our mission,” it states, “is to build the systems and technologies necessary to make life multiplanetary, to understand the true nature of the universe, and to extend the light of consciousness to the stars.” It then arrives at a number for a total addressable market that includes the light of consciousness: $28.5 trillion. It must have been fun to sit in the room as that number got calculated.

Musk’s mesmeric power over investors is well-known. Tesla earned just $3.8 billion in 2025 and is somehow one of the world’s dozen most valuable companies. (Caterpillar, which makes backhoes, earned approximately $8.9 billion.) It’s best to read the prospectus not as a ledger but as the blueprint for a future empire: the East India Company of space.

The business is a ladder that starts with rocket-launch capability. The next rung up is satellite arrays. This is a much more profitable business—Starlink really does pay for most everything at SpaceX—but it depends on launch capability. The third rung requires a leap of engineering faith: data centers in space. The fourth is the reason to build them—control over an off-world computing economy capable of training and running AI models at a scale that’s hard to achieve on Earth. In January, SpaceX filed a proposal with the FCC to operate up to 1 million data-center satellites in orbit—describing the project, in the filing itself, as the “first step towards becoming a Kardashev II–level civilization.” A Kardashev Type II civilization is one capable of harnessing the entire energy output of its star.

Does this sound crazy? If so, technology’s biggest players are eager to break into the asylum. Space offers AI companies abundant solar energy available 24 hours a day, and freedom from the NIMBYism that makes earthbound data centers and power plants so contentious. “Should I make a rocket company?” OpenAI’s CEO, Sam Altman, asked on a podcast in 2025. OpenAI subsequently tried, unsuccessfully, to take over Stoke Space, a reusable-rocket start-up founded by former Blue Origin employees. Alphabet’s Google and the satellite operator Planet intend to launch two prototype satellites in early 2027 to test whether AI chips can operate effectively in orbit.

Google and OpenAI do not have their own ladders. Musk does, and he’s already climbed the first two rungs; in August, he posted that SpaceX would put its first satellite data center in orbit by the end of next year. Bezos is in between—he has a ladder; it’s just shorter and shakier. The question is how much the gap matters as the race moves upward, into new areas of engineering and imagination where no one—not even SpaceX—has a lead.

In late 2027, Blue Origin plans to deploy TeraWave, a constellation of 5,408 new satellites custom designed to provide data centers and governments with internet up to 6,000 times faster than Amazon’s satellites. New Glenn is expected to place TeraWave into orbit.

John Couluris leads Blue Origin’s lunar-permanence team, charged with creating the infrastructure for colonization. “Maybe seven years ago, people wouldn’t have said, ‘Well, we could do data centers on the moon,’ ” Couluris told me. Now he considers it complicated but more and more plausible. The challenges are real—chips fail, cosmic radiation is weird, and maintenance is essentially impossible. But if launch costs continue to drop, they’ll offset the expense. Couluris said Blue Origin is already working on manufacturing solar cells directly from lunar soil. The chips would come from Earth; almost everything else could be made on-site.

The destination for all of this potential computing power has only recently become clear. In 2024, I heard rumors about Prometheus, a company whose plans were inscrutable even by the standards of AI start-ups. No one associated with Prometheus would confirm its existence. Given the history of Blue Origin, this should have been a tell.

What’s now known is that Prometheus is trying to build something quite different from the generative-AI models that spit out text or code. The most complicated things that humans create—jet engines, drug molecules, semiconductors—live at the intersection of invention and engineering, where physical forces interact in ways that are expensive to model and even more expensive to test. An engineer can sketch a thousand designs; each one still has to survive contact with physics. Prometheus is intended to compress that process—to create an artificial general engineer capable not merely of proposing designs, but, it seems, of simulating, testing, refining, and, eventually, helping manufacture them. The grandest possible outcome is a machine that makes machines.

There’s an echo here for Bezos. Neal Stephenson and Blue Origin’s other early employees spent years exploring engineering possibilities in slow motion while the competition raced ahead. An artificial general engineer could compress all of that painstaking discovery into rapid computation: a Neal Stephenson bot.

This could be useful to all humankind—and particularly useful to someone trying to build dexterous warehouse robots, cheaper rockets, and data centers on the moon. Whether any of this is possible, and on what timeline, is unknown. Large language models were jump-started by the availability (or theft) of trillions of words; there’s less training data on the physical world. The models also have to understand multiple kinds of physics at once, account for uncertainty, and produce results trustworthy enough to bet hardware and human lives on. Those problems are very expensive to solve. At least for most people.

In November 2025, Prometheus’s existence became public, along with the news that it had raised $6.2 billion, the largest chunk of which came from Bezos. He was also named co-CEO, his first operational role since Amazon. Prometheus has since hired 150 people, many from OpenAI and Google DeepMind. Dave told me that Bezos spends his days bouncing between Prometheus and Blue Origin. “But he multitasks very well. Well, he single-tasks in a multitask world very well.”

Musk already has his own AI company, xAI, which he merged with SpaceX. His chatbot, Grok, has mostly distinguished itself through creepiness and anti-Semitism, but Musk used $60 billion in stock from the IPO to buy Cursor, an AI-coding assistant that’s one of the fastest-growing software companies. That leaves tens of billions more to spend on the most precious commodity in tech: margin for error.

It’s not a coincidence that, a few weeks after the SpaceX IPO, Blue Origin said it had secured $10 billion in outside funding, the first non-Bezos money in the company’s 26-year history. Limp called it a sign of confidence. It was also an acknowledgment of reality: Blue Origin is going to need more money. It has less margin, and more errors.

Gary Hershorn / Getty

A SpaceX Falcon 9 rocket passes over Lower Manhattan in August.

SpaceX is winding down Falcon 9, the workhorse rocket that’s served as the West’s space-transportation system, in order to concentrate on Starship (as of this writing, the company was preparing for its 14th test flight—the first to attempt orbit). Other satellite companies have reported trouble buying Falcon launches beyond 2028. Starship will offer significantly more capacity—but SpaceX may use much of that capacity itself: Starlink and the orbital data centers it hopes to build are potentially far more lucrative than carrying other people’s satellites. In other words, the company that solved America’s launch shortage may now be creating another one.

This is not just a business problem. It also matters to a sovereign nation that has recently discovered that one of its citizens possesses powers normally reserved for states. All of humanity may learn what it means to have its future built atop someone else’s rocket.

In 2022, Ukraine asked Musk to extend Starlink coverage over Crimea so it could carry out military operations. Musk refused. General Mark Milley, then the chairman of the Joint Chiefs of Staff, and National Security Adviser Sullivan spoke with him about the decision. “And there wasn’t much we could do about that, absent taking some very coercive measures,” Sullivan told me—measures no one wants to take.

Three years later, during an X fight with Trump, Musk threatened to immediately decommission SpaceX’s Dragon capsule, the only American vehicle capable of transporting astronauts to and from the International Space Station. Without it, NASA would have had to pay Russia $80 million a seat on Soyuz capsules. A few hours later, Musk backed down. “We absolutely need diversification on this,” Sullivan said. “And what’s challenging is that the most likely pathway to diversification lies through another mercurial billionaire.”

That the republic finds itself needing Bezos’s extraordinary concentration of private power to mitigate Musk’s extraordinary concentration of private power is probably not what the Founders had in mind. I asked Senator Mark Kelly, a former astronaut, if he trusted either man completely. He replied, “Kind of a loaded question.”

Kelly has met Bezos. His history with Musk dates back to at least 2012, when he joined SpaceX’s human-spaceflight safety-advisory panel, consulting on the Dragon capsule that would eventually make SpaceX indispensable to NASA. After Kelly, who serves on the Armed Services and Intelligence Committees, visited Ukraine in 2025, Musk called him a “traitor” on X. So it goes. But Kelly doesn’t begrudge SpaceX its dominance. “It’s not SpaceX’s fault, right?” Kelly said. “They got ahead. They do it well.” What he wants is for SpaceX to “continue to be a successful company because it is in our national-security interest that that’s the case. I also want to see Blue Origin be successful.”

The pragmatist’s position, shared by most of the national-security community, is that one company capable of getting you to space is better than none, and two is better than one. But even if Bezos is the less mercurial billionaire, Blue Origin is not yet a reliable substitute for SpaceX.

Andrew Harnik / Getty

At the start of Trump’s second term, Elon Musk served as a senior adviser to the president and wielded extraordinary influence.

So the government can’t use Blue Origin as leverage. It can threaten to withhold contracts from SpaceX, but, as Trump discovered, Musk will respond by threatening to decommission the only American vehicle capable of reaching the space station. It can nationalize the companies, which would require a political consensus that doesn’t exist and a legal argument that hasn’t been made. Kelly thinks campaign-finance reform would make it harder for the next Musk to buy his way into the center of government, or the next Bezos to buy his way back into its good graces. When I asked which is more likely to happen first—campaign-finance reform or data centers on the moon—Kelly laughed. “I hope it’s campaign-finance reform.”

The government is not negotiating from a position of strength, but Bezos and Musk have a shared vulnerability: Both say they would like to go to the moon—for their empires, for glory, and because their youthful love of space has never settled into something more rational. To put it in terms they and the president would all understand, the moon is leverage. If the people trying to bargain with Charles Foster Kane had known about Rosebud, they would have dangled a sled.

For all of the stated urgency of the 2016 Trump transition meeting, what’s really happening with America’s return to the moon is only just now coming into view. In February, Musk posted on X that SpaceX had shifted focus from Mars to building a “self-growing city on the Moon.” The SpaceX prospectus has a section called “Establish the lunar economy,” which includes a base supplied by a continuous flow of people and cargo, eventually capable of producing its own water, fuel, power, and manufactured goods. Moontown sounds great, but the engineering details are a kind of truth serum—revealing the distance between what Bezos and Musk say they want and where their priorities actually lie.

SpaceX’s Starship cannot get to the moon without stopping for gas. The plan it sold NASA works roughly like this: SpaceX launches a Starship to serve as a fuel depot; then it launches a succession of tanker Starships to fill the depot; then it launches the lunar lander, which gasses up at the depot and departs for the moon. NASA’s Office of Inspector General estimates that at least 15 Starship launches will be needed to deliver all the fuel. No spacecraft has ever transferred ultracold methane and oxygen between vehicles in orbit.

In August, SpaceX committed $100 billion to build what Musk called “the biggest launch site on Earth” in Louisiana. This, along with its other facilities, should allow it, eventually, to launch multiple rockets a day. But even if SpaceX can get refueler rockets to the depot before the other rockets run out of gas, Doug Loverro, NASA’s former associate administrator for human exploration and operations, told me he’s skeptical of all the other things Starship has to prove: that it can reach orbit, keep astronauts alive, transfer fuel repeatedly in space, fly to the moon, land, and—crucially, if you’re an astronaut—take off again. His point is not that any single step is impossible. It’s the sheer number of steps that have to work in sequence. “Hurdle No. 1 is let’s get Starship to orbit,” Loverro said. “Which still hasn’t happened yet.”

Blue Origin’s route to the moon is complicated too: New Glenn first sends a transporter into low Earth orbit to be refueled. The lander, Blue Moon, launches separately, takes on fuel, and heads toward the moon, while the transporter climbs to a higher orbit for another fill. Near the moon, the two dock for one final transfer. Only then do the astronauts arrive and move into Blue Moon, which they use to descend to the surface and return to their spacecraft.

It’s an elaborate chain of rendezvous and refuelings, but the jobs are divided among purpose-built vehicles: New Glenn launches, the transporter carries fuel across deep space, and Blue Moon lands. SpaceX assigns nearly all of those jobs to variants of the same enormous Starship.

Saul Loeb / AFP / Getty

Jeff Bezos unveils Blue Moon, a lunar lander, in Washington, D.C., in 2019.

Also, as NASA’s Office of Inspector General has pointed out, Starship’s lander is 171 feet tall—roughly the equivalent of a 14-story building—and its astronauts descend in an elevator that begins 115 feet up. There’s no other way back into the vehicle if the elevator fails. Blue Moon is 52 feet, and its crew walks down a set of stairs that starts six feet off the ground.

If you’re curious why, 60 years after Apollo, getting to the moon seems harder, part of the answer is that NASA has 60 years of pent-up expectations. The original astronauts flew in tin cans atop disposable rockets. A sustained presence on the moon will require more mass and more propellant.

Most people I spoke with have reached the same conclusion: On this, Blue Origin is ahead. It has completed thermal-vacuum testing on Blue Moon, and, as of August, its structure, propulsion, and avionics were fully assembled; it’s already communicating with NASA’s satellites and deep-space network. Even with the launchpad explosion, Blue Origin still plans to deliver an uncrewed test mission to the moon by early next year. That first flight will be a chance to work out many of Blue Origin’s hardest lunar-specific tasks on a single trip.

The crux of the difference is that Blue Origin—the smaller, slower company in need of a win—seems to have built a product to fit the brief, while Musk has built a giant rocket designed for his satellite business and his personal goal of deep-space travel to Mars. Making Starship work for the moon is like adapting a cargo ship to win a yacht race.

None of this guarantees that Blue Origin gets to the moon first. “I’m the first to say never count Elon out,” Loverro told me. “He’s a clever guy.” It also doesn’t really solve the mercurial-billionaire problem as much as shift it. But Loverro had a suggestion so simple and clever that it made me laugh out loud.

The Apollo missions also sprang from private industry, but NASA divided the work among contractors—Boeing, North American Aviation, IBM, and more. Different companies, different roles, forced into collaboration by NASA. Why not make Starship the refueling station for Blue Moon? SpaceX would have to reconfigure some of its designs, but “it only takes one Starship to fill up a Blue lander. It may take 10 or 15 Starships to fill up a Starship lander. If they work together, they could both get there sooner.

“We’re individualistic in the United States,” Loverro continued. “Every man for themselves. But that’s not how we do great things. We do great things because we find the best of everything that we can bring to the table, and then we do it together. That’s what NASA’s job truly is—to organize that team. And they’re failing to do so.”

The person responsible for organizing that team is NASA’s administrator, Jared Isaacman, a self-made billionaire who twice flew to space on SpaceX’s Dragon capsule and performed the first private spacewalk. His relationship with Musk complicated his confirmation. On his first day in the job, Isaacman said that whichever lander was ready first would get the mission. He declined an interview request for this story.

Isaacman may hold to his winner-take-all position. He may avoid angering two of America’s most powerful men and hope their engineering successes or failures make the decision for him. But the Solomonic solution has a certain appeal: Force Musk and Bezos to work together, and test their willingness to share a win.

One small step toward reminding them that the nation is still king.

This article appears in the November 2026 print edition with the headline “Choose Your Overlord.” When you buy a book using a link on this page, we receive a commission. Thank you for supporting The Atlantic.

This Is the Hurricane Season El Niño Foretold

2026-10-09T18:40:18-04:00

Updated at 1:36 p.m. on October 10, 2026

For more than four months, since this hurricane season began, the Atlantic has been eerily placid—a remarkable stretch without a single hurricane. That prolonged calm has been rudely interrupted by Hurricane Isaias, a Category 2 fastball that slammed into the Gulf Coast last night along the Florida panhandle. “I think we knew at some point we would likely have a hurricane,” Daniel Brown, the operations-division chief at the National Hurricane Center, told me.

Not to be outdone, the Pacific Ocean has two major storms brewing, including Hurricane Simon, a dangerous Category 4 storm off the coast of southwestern Mexico. A smaller one, Tropical Storm Rachel, is forecast to bring potentially more storm surge and flooding as far north as Southern California. “It’s a busy day here,” Brown said on Friday.

This season’s dynamic—an unruffled Atlantic and a frothing Pacific—has been right out of the El Niño playbook and was predicted months ago. El Niños tend to generate opposing winds—low-altitude trade winds out of the east and higher-up westerlies—that tear apart Atlantic storms as they try to form and also keep them from gaining strength. This type of wind shear is affecting Hurricane Isaias now, Brown said, and weakened the storm a little before it made landfall.

Isaias landed along the Gulf Coast with sustained winds of more than 100 mph; in places, it dropped more than a dozen inches of rain. Coastal counties had ordered evacuations; the governors of Alabama and Florida had declared emergencies; residents had scrambled through hurricane prep. During the storm, the wind ripped off roofs and toppled trees. Two people, one in Florida and one in Alabama, died after trees fell on their house, authorities said. Nearly 800,000 people were without power across the Southeast as of Saturday morning, according to poweroutage.com.

In the Pacific, by contrast, a Category 4 hurricane is almost par for the course this season. By this time of year, the Pacific has usually generated 13 storms potent enough to warrant their own name, Brown said. This year, we’re on number 19. This onrush of hurricanes has a great deal to do with the buildup of abnormally warm waters in the equatorial Pacific—which, at this point, are in certain areas more than 4 degrees Celsius warmer than usual. This warming gins up storm activity in that ocean; it’s been spinning up hurricanes and hurling them toward Hawaii for months. The storm swells have also eroded Southern California beaches and swamped homes, and they keep coming. On Friday, seawater flooded the streets of Balboa Peninsula, California, inundating cars and homes, well before Tropical Storm Rachel was anywhere close to arriving.

The coast of Mexico is at particular risk from Hurricane Simon, which is forecast to bring potentially dangerous flooding and landslides to coastal states such as Jalisco and Nayarit. Further north, on the Baja California peninsula, communities are still recovering from the effects of Hurricane Polo, which caused widespread flooding last month and displaced hundreds of people.

The storm scorecard in these oceans has only reinforced the notion that the world is in the midst of an unusually powerful El Niño. As I reported in August, key areas of the Pacific Ocean had already gotten hot enough to make this a “Super El Niño”; the experts I spoke with then were anticipating that the event could end up peaking at twice that amount of warming. If those trends hold, this could be the strongest El Niño ever recorded, and right now, “everything seems to be on track with this being a remarkable event,” Nick Novella, a National Weather Service meteorologist, told me.

This season has set a record already—this is the latest in the year that a hurricane has formed in the Atlantic at any time during the satellite era, since 1966. And it isn’t even close. “This record was broken by nearly a month,” Novella said. Normally, five Atlantic hurricanes form by the end of September. For residents of the Gulf Coast, the beleaguered hours after a storm makes landfall are all too familiar. As they assess the wreckage, hopefully the year’s first Atlantic storm will not have changed too dramatically how we remember this season of relative calm.

When Anesthesia Goes Wrong

2026-10-05T07:30:00-04:00

Photographs by Julia Sellmann

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

When Freddy Marquez was 12, he was hit by a motorcycle while biking and smashed his face into the pavement. He healed fast, but as adolescence reshaped his boyish features, the scar started to tug on his upper lip, leaving his mouth ajar. “Mom,” he said, “I’m going to high school. I don’t want to look like this.” So in September 2024, his parents found a surgeon in central Florida, where they live, to do a scar revision.

On the morning of his surgery, Freddy went into the operating room around 9. His mom, Joana Jardin, settled into a waiting-room chair, expecting to drive him home in a few hours. Noon came, then 1 o’clock, then 2, then 3. “Where is my son?” Joana began to ask. “Where is my son?” The staff couldn’t tell her anything, except that she had to wait. She watched as other parents picked up their kids from surgery and left. At 4 o’clock, she realized she was the only one in the waiting room.

When Joana finally got to see Freddy, around 5, he was barely conscious. His eyes were closed, his breaths shallow. He was just sleepy, a nurse reassured her, because he’d gotten Benadryl after a hive-like rash broke out—but not to worry, the monitors tracking his vitals showed nothing else awry. Eventually, Freddy managed to slump into a wheelchair, his eyes still closed. He was sent home.

The next day, Freddy woke up. Then he threw up. He was still drowsy and now irritable. His head hurt. Joana took him to urgent care and later to the ER, where doctors who evaluated him figured he was just recovering from surgery. But in the coming days, his symptoms became stranger and more disturbing. Freddy started losing control of the muscles in his face. He developed tremors. He choked on his saliva. He lost his balance. He felt what he could only describe as electricity coursing through his brain. “Please, Dad, pray for me,” he would cry when the “electricity” struck.

At the hospital where Freddy was finally admitted, doctors ordered two separate MRIs of his brain, and, to keep him still during the scans, they put him under anesthesia. Both times, he took hours to regain consciousness. Both times, he emerged worse than before.

Freddy lost the ability to walk. His body spasmed and twitched as his muscles took on a life of their own. When Joana saw her son in the hospital bed—eyes staring at the ceiling, limbs flailing wildly—she remembers thinking this is how people once believed you could be “possessed by a demon.”

The scans found spots of white in Freddy’s brain—signs of dying and dead neurons—but no explanation as to what had caused them. In the coming weeks, he got better and then worse, worse enough that doctors ordered yet another MRI. His parents begged the hospital not to use anesthesia again. Don’t worry, Joana remembers being told; the doctors can do it safely. They do it all the time. Joana felt that she had no choice but to trust them. As Freddy slid into unconsciousness, she once again started ticking off the hours waiting for him to wake up: 1 p.m., 2, 3, 4. “Where is my son?”

Freddy never woke up. His body went into cardiac arrest, then respiratory failure. He died on December 29, 2024, at age 14.

Julia Sellmann for The Atlantic

Photographs of Freddy Marquez, who died at 14, shortly after routine surgery under anesthesia

Julia Sellmann for The Atlantic

Freddy’s mother, Joana Jardin

Before that final MRI, in one of their last conversations, Freddy had blurted in a half plea, half accusation, “You don’t understand what is happening to me, Mommy.” He was old enough, by then, to have lost faith in his parents’ omniscience but young enough to still yearn for it. His ordeal had revealed to him that the adult world is riddled with unknowns; that the assurances of even doctors in white coats can mean very little. “I promise that I’m going to understand. Until my last breath,” his mother told him. “I’m going to understand.”

And she would, eventually. Joana did not know it then, but Freddy’s death would be linked to dozens of cases of unexplained catastrophic brain damage after routine anesthesia, all in children or young adults. Several, like Freddy, died. Others became permanently disabled.

The mystery of why Freddy died would ultimately collide with one of the greatest unsolved mysteries of medicine: How does anesthesia work in the first place? How is it that anesthesiologists can induce a state that some have called “functionally equivalent to brainstem death”—only to reverse it? This process happens 60,000 times a day in operating rooms across the United States. Yet it remains poorly understood. What happened inside Freddy’s brain, some researchers now think, may be an extreme version of what happens inside any of our brains when we slip into unconsciousness under anesthesia.

This past May, in Germany, I watched a doctor give a talk to her colleagues about two eerily similar cases almost two decades apart. One was Freddy’s. The other was that of a boy in Germany who went under anesthesia for a broken arm in 2006. Like Freddy, Ulrich “Uli” Gabriel Gobin Hernández was 14. Like Freddy, he took hours to wake up and then gradually lost control of his muscles. But unlike Freddy, Uli survived. The woman giving the talk was not Uli’s doctor. She was his mother.

Yaremy Hernández Castillo, who is soft-spoken and modest to a fault, wore a white coat with sensible sneakers as she flicked through slides of her son’s brain scans. Most of her colleagues at the cardiac-rehab hospital where she worked had no idea that she’d spent the past 20 years—outside her job as an internist—trying to decipher his condition. Early on, she had intuited that what happened to her son could be explained only by something still unknown to medical science. “Uli Gabriel, patient zero,” she came to think of him. But in the initial lonely years of her investigation, his doctors doubted her. Her own medical colleagues did too. “Everyone told me it was absurd,” she recalled to me, even as they could muster no better explanation.

These were the facts of Uli’s case: On Saturday, January 14, 2006—a beautiful day to hit the slopes—he went snowboarding near his family’s home in the Bavarian Alps. Uli took a jump way too fast and landed on his right arm; the humerus bone did not withstand the force of impact. He ended up at the local clinic, where Yaremy worked at the time, and where she trusted he was in capable hands. On weekends, the alpine-village clinic treated an influx of skiers with mangled bones.

But as the hours passed and Uli did not wake up from surgery, Yaremy began to worry. When he finally did rouse—confused, vomiting, and talking “like a drunk man”—her instincts both as a mother and as a doctor screamed that something was wrong.

The trauma surgeon came to check on Uli’s arm and declared it perfect—his own job was done. No one else seemed to share Yaremy’s growing alarm at her son’s mysterious neurological symptoms. The word she kept using to describe this period was alone. To get an MRI of Uli’s brain, Yaremy personally arranged for an ambulance to take him to a larger hospital. On the MRI, two ghostly white spots appeared in his basal ganglia.

The death of brain cells in the basal ganglia, which coordinates movements of the body, explained Uli’s symptoms, which progressed to erratic twisting, painful muscle spasms, and involuntary protrusion of the tongue. But what had caused the damage? No one knew. Uli could still walk and talk, albeit with some difficulty, and he remembered everything: the confusion, the pain, the seven dwarfs from Snow White plastered all over his pediatric hospital room, which, as a 14-year-old, he found tragically uncool. He was impatient to get better and to get out; he had tickets to a Foo Fighters show less than two weeks later. But the day of the concert came and went, and Uli remained stuck in the hospital, his symptoms unresolved, their underlying cause still unknown. The tests that doctors ran all came back negative.

Had his anesthesiologist made a mistake? Yaremy couldn’t dismiss the possibility. How does a perfectly healthy kid become so sick from routine anesthesia, other than because of a terrible mistake? Malpractice could give grief a direction and a shape, suggesting a person to blame (a doctor) and a means of redress (a lawsuit). Friends kept telling Yaremy to sue. In the end, she did, filing a lawsuit against Uli’s anesthesiologist, Ludwig Düthorn, a senior colleague whom she regularly passed in the hallways of their clinic. At the very least, she figured, the lawsuit could help uncover what had really happened.

At the clinic, Yaremy began to avoid the ICU, where Düthorn worked, unsure of what to say about the tragedy that had rattled both of their lives. Düthorn, too, was haunted by the case. He replayed the surgery in his mind, searching for any hint of the debacle to come. Uli’s blood pressure had dipped briefly—but that wasn’t uncommon during anesthesia, and would not have been enough to cause this level of brain damage. Düthorn had counteracted the drop with a routine dose of noradrenaline, and that had worked as intended. All of Uli’s other vitals were unremarkable. All of the dosages had been double-checked by a nurse.

Afterward, whenever Düthorn had to anesthetize a healthy young boy, he would think of Uli. “I was personally devastated,” he told me. “I thought, I did the anesthesia. It is my fault. But on the other hand, I could find no mistake.”

Julia Sellmann for The Atlantic

Uli was an active kid, skiing at 3 years old and playing soccer in elementary school.

The death or serious injury of a healthy child from anesthesia is so unusual as to be universally shocking to doctors these days. Anesthesia is known to pose risks to the elderly or ill, but even then it is far safer than it once was.

In the 19th century, as many as one out of 2,500 patients died under chloroform, then a popular anesthetic. But patients were willing to take the risk, because the alternative was pure, visceral horror: During the pre-anesthesia era, hospitals would put their operating rooms in towers, so as to spare other patients from all the screaming.

Like many medical breakthroughs, the discovery of anesthesia was something of an accident. Doctors became aware that ether, a popular party drug at the time, made giddy revelers insensitive to pain. They decided to give it a try in the operating theater: The first public demonstration was carried out on October 16, 1846, when a 21-year-old man under ether had a malformation removed from his neck at Massachusetts General Hospital. (The name anaesthesia, from the Greek for “without sensation,” was suggested by Oliver Wendell Holmes Sr., the eminent doctor and essayist who also gave this magazine its name, in 1857.) At the time, doctors did not understand why ether worked; that it did was good enough.

General anesthesia today is often likened, colloquially, to sleep. But what it does to the brain is stranger and frankly spookier: A patient who is merely asleep would not lie still and take a bone saw to the chest. To render someone so deeply unconscious, anesthesia shuts down multiple distinct functions in the nervous system, making a patient insensate, paralyzed, and unable to form memories. The brain under anesthesia is suppressed to such an extent that Emery Brown, an anesthesiologist at Harvard and MIT, has compared it to coma or brain-stem death.

“The key thing is that it’s reversible,” Brown told me. When anesthetics stop being administered, the patient starts to rouse within minutes. Their consciousness flips back on. The spell breaks. This total reversal has happened millions of times in millions of people, the magic of it now mundane. But Uli had needed hours to find his way back to consciousness, his brain adrift someplace too deep and too dark to return from unscathed.

In the months after his accident, Uli’s symptoms became manageable enough for him to return to school. Still, he could be unsteady on his feet. On a walk with friends and their dog one day in 2007, he took a tumble and fractured his collarbone. His mother hoped the bone would heal without surgery, but his erratic limb movements from the damage to his brain kept jostling the broken bone. Surgery became inevitable.

Wary of anesthesia, Yaremy asked his doctor to avoid the type Uli had received for his arm surgery. She had no scientific evidence to cite, but she hoped he would listen to her, as a worried mother and as a fellow doctor. He did not.

This time, Uli did not wake up for days. Yaremy thought he was going to die. She would open his eyelids to check his reflexes. Nothing. When his eyes finally stayed open on their own, Uli was delirious. He screamed whenever an unfamiliar nurse entered his room. He had trouble following conversations. His skin became hypersensitive to touch. He lost more control over his limbs. He would never again be able to play guitar in his band or impress girls with snowboarding tricks. He would never even walk or talk on his own.

Yaremy struggled to care for him at home, and she began to contemplate her life in small-town Germany. She had grown up an ocean away, in the boisterous city of Caracas. At 21, she met a German man at a disco in Venezuela, who became her husband and then Uli’s father, and who persuaded her to move to this tiny ski town, where her new in-laws ran a hotel. When she arrived—with 1-year-old Uli in her arms—she spoke little German. For a long time, she was the only foreigner in town. She missed the warmth of the Venezuelan sun and she missed her large family, who she knew would rally to help her back home.

And so, soon after the disastrous second surgery, Yaremy boarded a plane with Uli, now 17, to Caracas. Her younger child decided to stay in Germany with their father. They all planned to reunite in Venezuela later, but Yaremy knew her marriage was already over, its existing cracks widened by the new challenges of caring for their son.

The move proved, in one way, fortuitous. A few years into living in Caracas, Yaremy encountered the first hint that what had happened to her son had also happened to someone else. When picking up medication for Uli—Botox to relax his clenched muscles—she met another mother doing the same for her own son. The facts of his case sounded awfully familiar: anesthesia, delayed awakening, brain damage, loss of muscle control. The mother told Yaremy about a spate of anesthesia complications in the Venezuelan state of Carabobo in the early 2000s, which were blamed on a contaminated batch of sevoflurane, then a fairly new anesthesia drug in the country. Her son, this mother said, was one such case.

Uli, too, had gotten sevoflurane during both surgeries. The colorless and faintly sweet gas has since become one of the most widely used anesthetics worldwide. But Uli’s surgeries were two years apart and 5,000 miles away in Germany, which had had no reports of contaminated sevoflurane. Yaremy couldn’t see how the incidents in Carabobo could be connected to her son’s. Still, she filed them away in her mind.

Not long after, Yaremy began chasing down a second clue, which emerged as Uli’s malpractice case wound its way through the German legal system. The lawsuit itself was unlikely to succeed, now that Uli had suffered the same complications under a different anesthesiologist, but an expert examining the facts of the case made a crucial observation: The two symmetrical white spots in Uli’s basal ganglia, he said, looked much like the brain damage found in people with inherited mitochondrial disease.

Mitochondria are the tiny structures inside our cells that act as power generators. When they become dysfunctional, cells are first starved of energy; then they die. The most vulnerable cells are the most energy-hungry ones; neurons are particularly susceptible, especially those in the basal ganglia. In children with mitochondrial disease, neuron death can be triggered by physiological stress of any kind: fasting, illness, dehydration, and, yes, anesthesia. Perhaps the fault lay with Uli’s mitochondria too.

In 2013, Yaremy managed to secure an appointment at a leading genetic-disease lab in Houston. She boarded the plane with Uli, convinced that the answer awaited them at their destination. There, specialists ran a series of blood tests on Uli and searched his DNA for mutations that cause mitochondrial disease. “I felt like I was going to see the light,” Yaremy told me over breakfast in May. “But nothing happened.” She sighed, her shoulders tense with disappointment.

The tests found no mutations known to cause mitochondrial disease. Moreover, Uli had never exhibited other symptoms; mitochondrial dysfunction affects numerous parts of the body, including the heart and muscles, and symptoms frequently manifest in childhood. Uli had been a healthy and energetic boy: walking before age 1, skiing at 3. Yaremy wondered if her son’s reaction to sevoflurane could be caused by a novel mitochondrial mutation. Uli remembers hearing the doctor, who did not know he understood English, mention her belief dismissively to his medical students.

Yaremy and Uli returned home to Caracas, where they had been building a new life. Uli enrolled at Venezuela’s largest university; using an eye-tracking device to communicate, he graduated with a degree in modern languages. Yaremy worked her way up to professor of medicine at the same university.

Then, in the mid-2010s, Venezuela’s petroleum-dependent economy cratered. Public hospitals lacked equipment as basic as disposable gloves; inflation soared to more than 1 million percent. Even with her connections, Yaremy began to struggle to obtain Uli’s medications. She decided in 2018 to uproot their lives once again, taking Uli back to Germany, where he could receive more reliable care.

By then, Yaremy was closer to finding the explanation for Uli’s condition than she knew. She already held key pieces of the puzzle. But it would take a second tragedy in her family for her to realize how they fit together.

In June 2020, Yaremy got a call from her younger cousin—and Uli’s godmother—María Fernanda Pozo Castillo, or Mafer. They are close, having grown up together in Caracas. Because Yaremy is the doctor in the family, Mafer told me, everyone bugs her with their ailments. (“Yaremy, I have a headache,” she said, slipping into an imitation.) But she was calling about something much more serious. Her 13-year-old son, Matías, had just had knee surgery after a bike accident in Austin, Texas, and he’d struggled to wake up. He had gotten sevoflurane.

“Mafer, no!” Yaremy said.

Even as she began to relive the cold dread of waiting for Uli to wake up, Yaremy pulled herself into problem-solving mode. “Don’t leave the hospital without an MRI, okay?” she told her cousin. “If the doctor says it’s not necessary, it doesn’t matter. You demand an MRI.”

Mafer remembers that the doctors were reluctant to order a brain scan for her son—he’ll be kicking a ball again in three weeks, they told her. But she kept asking and asking until they acceded. Sure enough, the MRI—which was done without further anesthesia—showed two white lesions in the basal ganglia. “Same places,” Yaremy immediately said when she heard this. “Same places as Uli Gabriel.”

In the coming days, Matías started experiencing involuntary limb movement. He developed muscle contractions, even more severe than his older cousin’s, that left him screaming in pain. The contractions have persisted, even with high doses of muscle relaxant delivered straight to his spinal cord. This past summer, Matías’s pain became so excruciating that he had to be hospitalized in the ICU. His mother cares for him full-time. Before we spoke on the phone, she kept apologizing that she might need to leave at any minute to tend to him.

Yaremy was devastated for her family; so was Uli. “I wouldn’t wish it on anyone,” he told me when we met in the spring, tapping the sentence out letter by letter on a board that he uses to communicate.

Julia Sellmann for The Atlantic

Uli has been unable to walk or talk on his own since his second surgery. He uses eye-tracking devices and letter boards to communicate.

Yet amid her grief, Yaremy’s scientific mind got to work: One kid in her family? Okay, a fluke. But two? It had to be genetic. She thought of Uli’s mitochondrial-DNA test from seven years earlier. The sequencing had found no mutations known to cause disease. But the report did note some “variants of unknown significance,” including one called m.11232T>C, meaning that the 11,232nd letter of his mitochondrial genome had been altered from a T to a C. No one made much of the finding; all of us have variants of unknown significance in our genomes, most of which are of little consequence, like minor typos that do not obscure the meaning of a sentence. But maybe this mutation was, in fact, significant.

Mitochondria carry their own loop of DNA, separate from the chromosomes that make up the rest of our genome, and are inherited through the maternal line. Yaremy’s mother and Mafer’s mother are sisters, so they and their children share the same mitochondrial line.

Could a single typo in their mitochondrial DNA be responsible for all of the catastrophic brain damage in the family? And what was the connection to sevoflurane?

Yaremy now began searching for a scientist to study her family. She suspected that Uli and Matías carried the same mutation, as might other relatives. Two of her mother’s sisters, she realized, had also suffered complications after surgery that now seemed to trace back to anesthesia. Were others at risk too?

She found Julio Montoya, an expert in mitochondrial disease at the University of Zaragoza, in Spain, who saw her email in a waiting room as he was about to undergo emergency surgery for a detached retina. With his one good eye, he read it and replied immediately. He was keen to investigate.

Montoya and his colleague Eduardo Ruiz-Pesini had, it turned out, seen the m.11232T>C mutation before. In 2018, they had found it in a Venezuelan boy at a Barcelona hospital. This patient had become severely ill after surgery to remove his tonsils and adenoids—and deteriorated even further with subsequent rounds of anesthesia in later years. The Spanish team decided, along with Yaremy, to put together a formal research proposal. The boy from the Barcelona hospital would become patient one in their study; Uli became patient two, and Matías patient three.

Over the years, Yaremy had talked with colleagues in Venezuela about Uli so relentlessly that word sometimes came back to her of similar cases: nearly always kids, nearly always after routine surgery. An introvert by nature, Yaremy had to learn to badger anyone and everyone for a useful crumb of information.

She began running down some of these rumors: The wife of the son of a friend knew a neurologist who knew an anesthesiologist who knew of a boy who had fallen ill after surgery on a broken forearm. He became patient four. Patient five was the boy whose mother she’d met picking up Botox in Caracas. A physical therapist told her about a boy who’d had cleft-lip surgery—patient six. An old friend from medical school, an anesthesiologist, had examined a boy who developed symptoms after surgery for a cochlear implant. The friend had once been skeptical of Yaremy’s theory, but now he helped her gather data. This boy became the seventh and final patient in the study.

All of the patients who ended up in the study were boys, but Yaremy heard of girls affected, too. For every patient she managed to track down, she followed other trails that ran cold. Families who’d filed lawsuits were reluctant to talk; others were lacking complete medical records. The allegedly contaminated sevoflurane in Carabobo, Yaremy learned, had been widely discussed among doctors at the time, but news coverage of the events 20 years ago had disappeared from the internet as a result of the regime’s media crackdowns and poor digital recordkeeping. She managed to get hold of one newspaper article—a printout saved by the mother picking up Botox in Caracas. But she was not able to find any other families whose children had died or become disabled in Carabobo.

A haunting pattern appeared in the cases she was able to pull together. All seven patients in the study had received sevoflurane. Most of them could trace their maternal relatives to Carabobo; Yaremy’s own maternal grandmother was from the state. And all seven had the same T-to-C mutation. This mutation is extremely rare, even in Venezuela, but it is an estimated 15 times more prevalent in Carabobo, where it is believed to have originated many, many generations ago. The Spanish researchers hypothesized that this old mutation had taken on new, potentially fatal importance in the era of modern anesthesia.

General anesthesia today is usually achieved through a cocktail of drugs, both inhaled and intravenous. Of these, sevoflurane is one of the most popular, particularly for children. It is easy to administer, quick to work, and not irritating to the airways. But sevoflurane seemed to behave differently—and dangerously—in the presence of the T-to-C mutation.

To prove this, the Spanish researchers would need to collect cells with the mutation and subject them to sevoflurane in a lab. Yaremy volunteered her own. In the fall of 2024, she found a colleague to draw her blood and a lab to biopsy a ribbon of her skin. Her cells soon made it to Spain, where the team got to work.

As the lab experiments were ongoing last summer, tragic news came out of Chile. Five children there had suffered terrible outcomes after routine anesthesia; four were dead, according to an alert from the Chilean Society of Anesthesiology. All five children were Venezuelan.

The news spread within the Venezuelan diaspora. In the past decade, more than half a million Venezuelans have moved to Chile, where they have not always found themselves welcome. Now some of the parents alleged that poor medical care and discrimination had killed their children. They filed lawsuits and demanded justice on social media. In videos the parents shared of their children on Instagram and on TV, they are all bouncy and bright: a 4-year-old girl swinging in gymnastics class, a 2-year-old boy in a Messi jersey kicking a soccer ball.

Chilean investigators were able to collect genetic samples from only three cases—out of what had grown to a total of seven in the country—but all three patients had the same T-to-C mutation. One of the children who died in Chile, Yaremy later learned, was the cousin of a patient in her study. As word of the mutation spread around the world, doctors started reporting cases in the United States, Canada, Colombia, Peru, and Guyana.

“I suffered with each kid, each mother,” Yaremy told me. Even when a child survived, she saw ahead to their “lifetime of complications.” She often thinks about the boy with the cleft lip from her study, who, abandoned by his mother, lives in a poor village in Venezuela with his grandmother. His family can’t afford Botox or other treatments. In videos she has seen, his muscles are always clenched, his limbs unusable.

Back in Spain, the researchers raced to finish the experiments. They ultimately found that cells carrying the mutation were indeed hypersensitive to sevoflurane. This mutation could explain why Uli—and Matías and all the other children—had stayed unconscious for so long after surgery. The study, with Yaremy as a co-author, was the cover story of the June 2026 issue of Anesthesiology, the official journal of the American Society of Anesthesiologists. Accompanying it was an editorial written by two physician-scientists in the U.S., Richard Levy, at Stanford, and Philip Morgan, at Seattle Children’s Hospital.

I had first spoken with Levy a few months earlier, after the ASA issued an alert—following up on the one in Chile—about terrible complications from anesthesia, primarily sevoflurane, in people with maternal Venezuelan ancestry. In our conversation, I asked what seemed like a basic question: How does sevoflurane work? “If you can figure that out, you have a Nobel Prize,” Levy told me. “The reality is, we don’t know.” We don’t know how any inhaled anesthetics work. But this novel mutation out of Venezuela, he and Morgan thought, might hold a clue.

Julia Sellmann for The Atlantic

More than a decade after his injury, Uli became a case study in a scientific paper, co-authored by his mother, about how anesthesia interacts with a rare mitochondrial mutation.

Anesthesia likely taps into something evolutionarily ancient in the body. A one-cell protozoan, a Venus flytrap, and a human are all susceptible to anesthesia—and two of those things don’t even have a brain.

For most of the 20th century, scientists clung to the idea that these drugs perturb the fatty membrane of living cells—a hypothesis based on how the gases behave in olive oil. But by the 1980s, a new dogma began to take hold: Anesthetics bind to proteins on the surface of brain cells, which interrupts their signaling. This is best understood in a drug called propofol, which is a common anesthetic but not an inhaled one; the milky-white liquid is delivered via IV. In an elegant set of experiments in the 2000s, scientists identified propofol’s primary target as a protein called the GABA receptor. Yet the same experiment, when repeated with a common inhaled anesthetic, failed to find an analogous target. Decades later, it remains elusive. Maybe it doesn’t exist.

Today, the general consensus regarding the mechanism of inhaled anesthesia is that the gases bind weakly to many proteins in the brain, which together add up to complete loss of consciousness. Most practicing anesthesiologists don’t think much about this. Sevoflurane works. The proof is in the operating room every day. What more do you need?

But a group of researchers dissatisfied with the conventional wisdom on proteins, including Levy and Morgan, has been slowly building the case for an alternative hypothesis: Actually, it’s the mitochondria.

When these researchers heard about children with a mitochondrial mutation dying from anesthesia, they immediately saw the implications. Scientists have long relied on rare mutations to probe the mysteries of biology. In another Venezuelan state called Zulia, for example, families with a shared mutation for Huntington’s disease have been crucial to our understanding of the condition. Rare mutations are a kind of natural experiment.

Morgan, along with his wife and scientific collaborator, Margaret Sedensky, had once run lab experiments looking for mutations that altered sensitivity to anesthesia—though only in a tiny, transparent worm called Caenorhabditis elegans. When they induced random mutations in the worm, they found that the relevant ones were clustered in the mitochondria. Their findings didn’t make much of an impact among their fellow anesthesiologists, though. “We’re like the comic relief at our big national meeting of 15,000 anesthesiologists,” Sedensky told me. “ ‘Here come the worm doctors.’ ”

But Morgan and Sedensky kept working out the details of their theory. Inhaled anesthetics such as sevoflurane, they now argue, work by inhibiting the mitochondria, inducing a kind of temporary brownout in the brain. The bubbles used to transport brain chemicals become frozen in place, like cars that have run out of gas. They restart as soon as anesthesia is discontinued. But in some hypersensitive patients, the mitochondria could become so impaired that their brain cells die. Morgan had found some of these hypersensitive patients before, but he believes Uli’s mutation could help pinpoint more precisely how anesthesia works on mitochondria. This past summer, cells with the mutation—derived from Yaremy’s blood sample—arrived in the couple’s lab in Seattle.

Understanding how anesthesia works, Levy told me, could enable scientists to design better drugs that precisely target the brain, without lingering side effects. Though modern anesthesia drugs are generally quite safe, they are imperfect. They affect the cardiovascular system; they affect ostensibly unrelated areas of the brain, such as the region that controls nausea. And they have been linked to cognitive decline in the elderly after surgery—though the degree to which the drugs are responsible, as opposed to other factors such as the physical effects of surgery, is still being debated. How much of that can be avoided?

Before her study was published, Yaremy had someone she needed to call: Ludwig Düthorn, Uli’s anesthesiologist after the snowboarding accident. She had not seen Düthorn since testifying against him at the malpractice trial years earlier. She’d lost the suit; legally, he had already been absolved. Personally, she felt she had to apologize.

Düthorn was gratified to know what Yaremy had found, but even now, he told me, he feels a doubt gnawing at him in the operating room. Uli had a silent and novel mutation in his DNA. Could his next 14-year-old patient have a different one?

Cases like Uli’s unsettle everyone they touch. In the 2000s, Rafael Legórburu, then an anesthesiologist in Caracas, treated a young boy who would not wake up after sevoflurane—a case he now suspects was connected to this mutation. The boy suffered profound brain damage. Legórburu still finds it difficult to talk about 20 years later. “I have a scar,” he told me, “for the rest of my life.” Today he lives in Spain, where he practices aesthetic medicine, administering Botox and fillers; he says he’ll never touch sevoflurane again.

Doctors are now revisiting a number of old malpractice cases—some of which resulted in large settlements—that might be explained by the T-to-C mutation. An anesthesiologist whose hospital had settled one such case involving their patient told me they felt relieved, and vindicated. (They spoke with me on the condition of anonymity because they were not authorized by the hospital to speak publicly.) At the same time, they could not entirely shake their sense of responsibility. “It was the worst experience of my life,” they kept telling me.

A group of Venezuelan anesthesiologists—largely based in the U.S.—recently formed the Venezuelan Anesthesia Perioperative Risk society, or VAPOR, to raise awareness of the mutation among health-care workers, and to promote guidelines for screening and treating patients with maternal Venezuelan ancestry. A quarter of Venezuela’s population—8 million people—has left the country in recent years, meaning the mutation is scattered more widely than ever. Yet within Venezuela, no lab is testing for it, and anesthesiologists do not routinely have access to the specialized equipment needed to administer anesthesia without inhaled gases.

Yaremy is working with scientists to continue studying the mutation, about which many questions remain: Do other anesthetic drugs, inhaled or administered via IV, pose a danger? Are children more vulnerable, or have we simply not looked hard enough yet in adults? The oldest known patient to be affected was 31 years old.

In her own life, Yaremy has finally found a measure of stability. She and Uli live in Radolfzell, a bucolic lakeside town in southern Germany, where she works at the cardiac-rehab hospital. It’s calmer than the ICUs where she’s worked in the past. “B-O-R-I-N-G,” Uli joked, tapping out the letters in English. But the calm is welcome after years of upheaval. Uli has his own apartment and a rotating cast of full-time assistants; he works as a translator for medical records. (He is fluent in English, Spanish, and German.) He is a doting uncle to his niece and nephew, the children of his younger sibling, Levi.

The mutation still lurks in their lives. Both times after giving birth, Levi, who uses they/them pronouns, developed an infection in the uterus and needed surgery. Yaremy made sure that they did not get sevoflurane. But what about Levi’s kids, who also have the mutation? What if they’re rushed into emergency surgery on their own? Levi had closely followed the news of the children who died in Chile. “They could have been my kids,” Levi remarked as we parted ways after a family lunch by the lake. The water was a brilliant clear blue that day; so was the sky.

A few days after I left Germany, Yaremy learned about a new suspected case: a 12-year-old boy who’d had trouble waking up after surgery in Caracas. “How can this still be happening?” Uli wrote to me. “I mean what the ****!?”

Julia Sellmann for The Atlantic

Uli and Yaremy now live in the lakeside town of Radolfzell, Germany. Uli works as a translator for medical records.

Joana, Freddy Marquez’s mom, first caught wind of the suspected mutation in the summer of 2025, when she heard on social media about the children who’d died in Chile. She, too, is originally from Venezuela. And she made the connection instantly. Months earlier, when Freddy had been hospitalized, his doctors had ordered a genetic test that turned up a variant of unknown significance. Freddy had the m.11232T>C mutation as well. This information never made it into his chart for the same reason Uli’s mutation wasn’t initially taken seriously: No one understood its importance at the time.

Joana began commenting about Freddy on every related post she could find. One by one, she linked up with others who told similar stories: another mother in Florida, the family of a girl who died in Houston, Yaremy’s cousin Mafer in Austin, and so on. Joana got added to a WhatsApp group for the parents in Chile, which has since evolved into a de facto global support group for parents of children with this rare mutation. Through the group, Joana also learned of Yaremy. They spoke for the first time in February, and recognized in each other a maternal instinct that is both steely and tender. “She was in tears,” Joana told me, “and I was in tears.”

In February, Joana appeared on Instagram Live in a zebra-print shirt to talk about Freddy, alongside other affected mothers. The zebra print was a reference to the old medical adage “When you hear hoofbeats, think horses, not zebras”—meaning: think of the common explanation, not the exotic one. But sometimes, it is zebras. Freddy died because what happened to him lay just outside medical knowledge. His particular mutation is rare, but encounters with the unknown in medicine are not. Joana was asking for some humility, some openness to the possibility of zebras. Since Freddy’s death, she has made advocating for people with rare mitochondrial conditions her life’s mission.

Joana can still picture him: swimming, playing basketball, riding “like crazy” on his bike as Gordito, his Alaskan malamute, raced alongside him. “His flipping dog is now chasing me wherever I go,” she said. And Joana continues to propel herself forward. She’s cold-calling researchers. She’s going to medical conferences. She’s offering up her own cells to study. She is keeping the promise she made to Freddy before he died.

This article appears in the November 2026 print edition with the headline “Going Under.”

AI’s Real Gift to Science

2026-10-04T11:56:28-04:00

Dario Amodei, the chief executive of Anthropic, has often said that AI will cure most diseases in the next 10 years. Late last month, he claimed that his company had taken a first step toward that future: Claude had, in less than 24 hours, discovered a new “enzyme system” that the company says resembles CRISPR, a gene-editing tool currently curing people of diseases. While announcing the discovery, Anthropic hinted that this was the type of finding that revolutionizes medicine and quoted a gene-editing pioneer who called it “genuinely intriguing.”

The company was open about not knowing what exactly it had spotted or whether the finding—the first released from a biology lab that the company established this spring—was really that significant. Still, many academics, along with the CEO of Eli Lilly, pointed out that the company had dressed up a very incremental finding as a major discovery. Fyodor Urnov, a scientist who coined the term genome editing and is working on a CRISPR project with Anthropic, told me that “the jury is out” on whether what Claude found is even an enzyme at all. Multiple researchers told The New York Times that, in fact, they had been aware of the supposedly new enzyme system for years, including a team at the University of Copenhagen that alleged Anthropic’s new results might have been stolen from their lab’s interactions with Claude. The company, in response, said that Claude isn’t trained using user transcripts. (Anthropic did not respond to a comment request for this article.)

In other words, Anthropic made a big deal about a finding that—although achieved in a novel, Claude-led way—may not lead to any biomedical advances, and that the company knew may not lead to any biomedical advances. Ironically, in making such grandiose statements, the company and other AI boosters obscured what Claude does appear to have accomplished: It sorted through reams of biological data that would have taken scientists much longer to analyze, if they could have at all. This capacity might not be a miracle, but for scientists facing mountains of data that have surpassed their capacity for analysis, it’s still meaningful.

Around the turn of the 21st century, researchers sequenced for the first time all 3 billion letters that make up human DNA. This effort, called the Human Genome Project, took 13 years and cost $2.7 billion. Today, sequencing a person’s full genome takes just over a day and costs as little as $200. Accordingly, the total amount of publicly available genomic data has become so large as to be unmanageable. In other fields, such as protein science, technological advances have led to a similar boom in the sheer amount of data being collected.

Trying to pull anything useful from all that data has become its own long-term project. In 2012, a committee of bioinformatics experts found that the “bottleneck in scientific productivity” had moved from data production to data analysis, but early efforts at the National Institutes of Health to deal with that bottleneck quickly sputtered out. If researchers could efficiently store, send, and examine all of the information they generated, they would likely be able to, for example, develop new drugs by identifying better genes or cellular parts to target. But it’s simply too massive for them to comprehensively attack.

The biological corpus is not, however, too large for AI agents. Bo Wang, an assistant professor at the University of Toronto and the chief AI scientist at a drug-discovery company, recently wrote that AI’s biggest impact in biology will come from its help “scaling scientific attention”: Setting agents loose on the enormous data sets can surface the outliers that warrant further human inspection. Agents can propose and test hypotheses for what might cause the unusual patterns they find, and predictive modeling can help scientists choose what experiments to carry out on their actual lab benches. Eventually, automated labs might even carry out those chosen experiments—a few already exist in the United States, and the Trump administration is spending hundreds of millions to establish more. Meanwhile, scientists remain divided over whether AI models actually steer them toward potentially paradigm-shifting experiments, or safer, less exploratory ones.

AI is already changing medicine: Models that use people’s medical history to predict their health risks, including their likelihood of developing certain kinds of cancer, are now used in most U.S. hospitals. At the same time, not every field of biology generates enough data for this approach to work. AI might be less useful in some domains that rely on less abundant microscopy data, such as the study of how cells and tissues communicate; such data tend to be harder to come by than genetic sequences, Stirling Churchman, a professor of genetics at Harvard Medical School, told me. Organelle biology and the science of cellular transport similarly require painstaking, time-consuming experiments.

Still, Churchman said, “there will be magic where there’s data.” She spent the summer at OpenAI working on a specially trained model for life-science researchers. She anticipates that by allowing scientists who can’t code to examine their data themselves, AI will enable researchers to “analyze data at least 10 times faster, freeing us to spend more time on the biological questions and experiments.”

The recent Anthropic finding is an example of how AI can identify outliers that might otherwise have gone unnoticed. Existing tools can comb through heaps of genomic data (and have already been used to identify bona fide new gene-editing systems). What AI can do is run those kinds of analyses at a much larger scale, while also evaluating intermediary results, Kendell Clement, who runs a computational-biology lab at the University of Utah, told me. Of course, that scale requires dumping in correspondingly huge resources: Anthropic directed about 950 Claude bots to simultaneously pore over a database that contained nearly 2 billion DNA entries.

AI companies have every incentive to keep evangelizing about the medical miracles they say their models will unlock, especially as backlash against them mounts. Amodei has an idea for how to quiet AI skeptics: “The thing that will work is actually curing cancer,” he recently wrote on X. That would certainly be a coup. But helping scientists work more quickly is, too, even if it doesn’t sound quite so revolutionary.

AI Could Worsen a Common Cognitive Trap

2026-10-03T09:00:00-04:00

One of humanity’s most remarkable cultural achievements is the division of explanatory labor. When my daughter came down with a fever, I solved the problem by consulting a pediatrician, who shared a likely diagnosis of the flu. Unlike the doctor, I wasn’t able to tell my daughter much about the mechanism by which the flu leads to fever, or exactly why it was so important for her to keep drinking fluids. But the explanation I acquired—roughly “fever because flu”—was what was needed to treat her.

Society runs on this efficient type of information sharing. Just as coordinating specialized skills can streamline the fabrication of products—be it pins or cars or newspapers—coordinating specialized knowledge can facilitate the answering of questions. You can drive a car even if you don’t have the expertise to manufacture one yourself, and I benefited from the pediatrician’s explanation even though his understanding far exceeded my own.

We can thank this divide-and-conquer method for many advancements in medicine and science, engineering and commerce; in fact, it has helped drive the accumulation of human knowledge across communities and over time. But it also leaves us susceptible to illusions of understanding: cases in which we mistake the understanding of others for our own.

Humans’ sense of whom to ask for what is fine-tuned by evolution, culture, and personal experience. Even 3-to-5-year-olds are more likely to say that you should ask a doctor rather than a mechanic about fixing a broken arm, and a mechanic rather than a doctor about fixing a flat tire. As kids mature, they develop an even deeper understanding of expertise. In a 2004 study published in Child Development, for example, 11- and 12-year-olds tasked with finding out why tennis balls bounce better on the sidewalk than on grass were more likely to suggest asking someone who knows why bubble wrap keeps glass things from breaking (which is also relevant to physics) than someone who knows why tennis balls come in cans of three (which just happens to also involve tennis balls). Their choice of whom to ask was based on fundamental principles—not just superficial similarities.

With plenty of time and care, years of education, and a healthy appetite for learning, most of us transition from children who know little about science or other expert domains to … adults who know little about science or other expert domains. Even those of us who acquire expertise in one area are unlikely to do so in many others. So the ability to track outside expertise is no less important for adults than it is for children. We need to know who can explain a child’s fever, a wilting plant, a clause in a will, or an imperfect soufflé—and research finds that adults seem to do this pretty well. (As a population, of course, we do not do this perfectly: Plenty of people trust in bad advice from their friends, influencers, and self-proclaimed experts.) A study from Germany found that even college students who aren’t science majors do a good job of guessing which kinds of scientific expertise are relevant to different scientific topics.

So far, so good. The trouble arises when our connections to other minds become a little too seamless, and we start to lose track of where our own understanding ends and another’s begins. When we receive an explanation from an expert—such as “fever because flu”—we might mistakenly conclude that we’ve learned more than we really have.

In fact, research finds that merely having access to expert answers can lead to illusions of understanding. In one set of studies, participants read about various fictional scientific discoveries and reported how well they understood the phenomenon described in each one. For instance, some participants received the following short passage:

A May 19, 2014, study in the journal Geology reported the discovery of a new rock that scientists have thoroughly explained. The rock is similar to calcite, yet it glows in the absence of a light source. The authors of the study, Rittenour, Clark, and Xu, fully understand how it works; they provided a description of the remarkable appearance of the mineral and outlined future experiments.

The passage didn’t actually include any explanations about why the rock glows. And yet, after reading the passage, participants thought they had some comprehension: what the authors call “a contagious sense of understanding,” because they seem to have caught it from the scientists described in the passage. When participants instead read that scientists “have not yet explained” the glowing rock, and that they “do not yet understand” how it works, there was no contagious sense of understanding; participant ratings of their own understanding nearly hit the bottom of the scale.

More recent work has shown that a contagious sense of understanding can even infect our interactions with technology. In one study, participants learned about a bank’s “robo-adviser,” which could provide investment recommendations. Participants also learned that an explanation of how the robo-adviser works had been published in a scientific journal. But half were told that the explanation was available to them on the journal’s website, and half were told that the journal had a subscription paywall. Those who thought the explanation was available to them reported a better understanding of how the robo-adviser works. They also reported a greater likelihood of using the robo-adviser themselves.

These results raise a pressing question. As information becomes more accessible through AI assistants, chatbots, and other technology, will these illusions of understanding become even more pronounced? Some early evidence suggests that the answer is yes. Several studies have now found that when people have information available to them through the internet, they overestimate how much they themselves know.

For example, in one study, participants were asked to explain how something works, such as a zipper, and half were given access to the internet to “confirm the details of the explanation.” All participants were then asked how well they could offer detailed answers to unrelated questions, such as why cloudy nights are warm. Those who had access to the internet in the first part of the study rated themselves significantly more capable of explaining in the second part—suggesting that the experience of having more knowledge at their fingertips supported the illusion that the knowledge was in their own heads.

Interactions with fast and informative chatbots will only heighten such illusions, but with a worrisome twist: The evidence so far suggests that humans aren’t great at tracking who knows what when it comes to large language models such as ChatGPT and Claude. For instance, when people in a study learned about an LLM that could correctly answer a question about economics, they expected it to correctly solve a basic algebra problem—but they were wrong. And when they learned that an LLM could answer a question about moral philosophy, they thought it could identify which character in a scenario had done something immoral—but they were wrong again. Participants almost certainly applied assumptions about how knowledge clusters in human minds to the AI system, and in both cases, this led them astray.

The kinds of errors that AI systems make are likely to change as AI develops, and human intuition about AI will probably evolve alongside these changes. As we expand our community of explainers beyond human experts to chatbots and other forms of AI, we create new opportunities for human understanding. But we also introduce an increased potential for illusion and error.

This article was adapted from Tania Lombrozo’s new book, Why We Ask Why: The Science of Explanation and the Human Drive to Understand.

Bait and Switch

2026-10-03T09:00:00-04:00

Earlier this year, one of my fishing partners sent a 1 a.m. message to our group chat. It was a link to an Instagram post of an angler holding a sizable striped bass. Dappled gold and silver with a dozen distinct black lines that stretched from gills to tail, it would have been a gorgeous fish if not for two problems: Striped bass usually have seven stripes, not 12, and men usually have five fingers, not six. The photo, it seemed, was AI-generated. To say we were dismayed would be an understatement.

From May to November each year, I spend most of my nights doing an extreme version of surf fishing for striped bass called “wetsuiting,” which involves wading or swimming out to distant ocean rocks and sandbars after dark. This kind of surf casting is a niche activity, and like many niche activities, the community for it is small. Wetsuiters can be found in a smattering of local fishing clubs throughout the Northeast—the typical range of the striped bass—but are primarily connected through Instagram, where some people, mostly for bragging rights, post images of fish they’ve caught.

[From the October 2024 issue: Inside the dangerous, secretive world of extreme fishing]

In our group chat, my fishing partners and I spend too much time scrutinizing photos of bass that other anglers have posted. People who fish are habitual liars (I say this fondly), and most of our conversations take the form of debates about whether a fish is really as big as someone claims. Lately, though, we’ve been asking a different question: Is that fish even real?

In recent months, the small corner of Fishing Instagram devoted to surf casting has been plagued by a slow drip of AI-generated or AI-enhanced fish. Sometimes, the fraud is egregious—one of my friends, a fishing photographer, discovered that another angler had stolen one of his photos and crudely replaced my friend’s face with her own. More often, the distortions are subtle. Typically, a fish or fisherman looks ever so slightly off, usually in ways that a casual observer might not notice: A bass’s fins aren’t quite right, or the lure in its mouth looks oddly distorted, or the angler’s arms are a bit too long. In my favorite example, a man lifts a large striped bass just as he is getting pummeled by a wave. The scene looks daunting—until you realize that the ocean behind him is peaceful. Where did this rogue wave come from? One suspects that the enterprising angler, wanting to look more hard-core to his Instagram followers, prompted a bot to dial up the intensity. ChatGPT, make me look badass.

Fishing magazines report growing concern about AI-generated fish in freelance photo submissions. I have written occasionally for the small online magazine Surfcaster’s Journal. Over the summer, the Journal announced that it would now run all manuscripts and photographs through AI-detection software. “For some reason, I naively assumed that fishermen wouldn’t stoop to the ‘core of the earth’ lows required to use AI to prop oneself up as a fisherman worthy of clicking the follow icon,” Dave Anderson, a well-respected wetsuiter and an editor at The Fisherman, wrote in an article lamenting the scourge of artificially generated fish on social media. “I was very wrong about this.”

Angling, even extreme angling, is just a hobby, but although the stakes may be low, the anger about artificial fish is real. Partly, I think this is because of the work that being even remotely good at the sport demands. Striped bass move and feed in locally specific patterns that can take months or years to crack. There is a risk of drowning, and the mounting reports of bull sharks, tiger sharks, and great whites in the Northeast call attention to another layer of danger: We smell like fish, swim in dark water, and look like clumsy seals in our black neoprene. Wetsuiting does not change the world, but it is an activity that is hard and honest and, in that way, human.

[Read: The year great white sharks took over Maine]

Even taking a picture of a fish is demanding, requiring salt-water-proof cameras and often wading or swimming a bass back to shallow water, where it can safely be photographed and then released. The Photoshopping of fish pictures—and of the natural world in general—has been an issue for years, but AI, with its ability to conjure a simulacrum of reality out of nothing at all, is another matter. It automates away the need not only to photograph a fish, but to even catch one in the first place.

The problem extends to other sports and hobbies, many of which are now defined by what Stuart Franklin—who took one of the celebrated “Tank Man” photos at Tiananmen Square—calls the “documentary impulse.” Social media has turbo-charged that impulse into an imperative. A good local skateboarder’s or surfer’s reputation would once have been built through word of mouth. Today, posted photographs and short-form videos may constitute the primary proof of ability. Concerns about AI slop are already being raised in cycling, skiing, and other activities. Climbing Magazine recently called out an AI-saturated YouTube channel purporting to chronicle climbing disasters that in fact never happened.

As AI-generated fish have appeared more frequently in my social-media feed, I’ve been reminded of a scene in the bird-watching documentary Listers, where a highly competitive birder is asked whether he would keep bird-watching if the photo-logging website eBird disappeared. His answer—“No”—was unsettling, not least because one suspects that this sentiment is widespread in hobbies besides birding. I have often wondered if, for some of the wetsuiters who populate surf-casting Instagram, the dopamine hit from the likes and fire emoji is sweeter than the hit from catching the fish itself. Now that posted images are being looked upon with an ever greater degree of suspicion, I also wonder how much longer the dopamine hits will come for them at all.

This is hardly the greatest challenge posed to humanity by the advent of AI. But hobbies, particularly those that require some measure of skill and endurance, have an influence that extends beyond those who directly participate in them. Free Solo, the Oscar-winning 2018 documentary about the death-defying rock climber Alex Honnold, found a wide audience outside the narrow world of climbing because people are invested in seeing other people do difficult things. I love videos of big-wave surfing even though I never plan to climb onto a surfboard. If AI makes it impossible to trust what we see, even something as simple as a surfing video, then we have lost more than the integrity of any one sport. We have lost the ability to believe in the accomplishments of others.

This article appears in the November 2026 print edition with the headline “The Fish Are Getting Bigger.”

The Edge of California Is Falling Into the Ocean

2026-09-28T07:00:00-04:00

This season, El Niño is offering a glimpse of what the future has in store for California. Engorged with warm water, its waves powered by a frenzy of hurricanes, the Pacific Ocean has been clawing away huge chunks of the coast. Waves overtopped defensive sand berms and tore up the boardwalk on the Long Beach Peninsula, in Los Angeles County. Beaches in places such as Carlsbad, to the south, got wiped clean of sand. Train service to San Diego was interrupted. In Malibu, an access road to Point Dume, a promontory popular with hikers and surfers, partially collapsed from erosion. And on Labor Day weekend, in the driveway of Nicolas Cage’s beachfront home, a small sinkhole opened up.

Francois Navarre, who once ran a paparazzi agency, had moved into the Malibu neighborhood just a few weeks before and discovered the hole while strolling down Sea Level Drive, the private street that runs behind a row of $8 million–to–$10 million homes, perched on pilings, that have an unimpeded view of the Pacific. The Palisades Fire had damaged Navarre’s home in Los Angeles last year, so he’d been living in Colorado. Now he could peer down through the pavement of his new neighborhood to the seawater coursing more than 20 feet below.

The hole grew fast, expanding more every hour, Navarre told me. Gas and water lines soon ruptured. The ocean dislodged the home’s septic tank, which slammed into a neighbor’s home, a security guard for the community told me. Nearly 40 years ago, a coastal engineer had sized up this exact lot and written that the Pacific Ocean posed no threat of washing away the building or its septic system.

Over the next week, the chasm beneath the street stretched below at least four waterfront houses and impeded emergency access to the area, forcing the city of Malibu to red-tag some 30 homes as uninhabitable. Cage did not respond to a request for comment. As an actor, he has infiltrated Alcatraz, navigated underground tunnels, and been pinned under rubble. During this real-life disaster, he was away. Not all of the homeowners in this gated community—a former mayor; the proprietor of a local Ford dealership; the president of the Miami Heat, Pat Riley—live there year-round. The security guard, Cesar Quintanilla, said that among the houses deemed unsafe, only four have full-time residents. When the Red Cross opened an emergency shelter at a nearby high school, nobody showed up.

California’s shoreline faced considerable risk even before this year’s El Niño intensified, when a mass of abnormally warm water in the Pacific Ocean began moving east along the equator. When it hit South America, it cleaved in two, heading south and north. This pulse of warmer water, known as a Kelvin wave, at times this past summer added nearly a foot to tides in Southern California, Jonathan Warrick, a coastal scientist at the U.S. Geological Survey, told me. These conditions also tend to drum up storm activity in the Pacific, and earlier this month, three hurricanes were spinning simultaneously in the ocean. The one closest to California, Hurricane Marie, pushed the swell that hammered the coast over Labor Day weekend.

Sea-level rise from global warming, too, is slowly adding inches of water that are raising the whole system, Warrick said. Malibu’s coastal-vulnerability report, published earlier this year, estimated that 1,410 of the city’s residential buildings—20 percent—plus more than half of the city’s commercial buildings could be affected this century by flooding, seawater intrusion, erosion, or other manifestations of rising seas. Malibu Mayor Bruce Silverstein expects that almost all waterfront structures will be in danger in the decades to come. “Nobody should be buying a home like that unless they can afford to lose it,” he told me.

The victims in this situation don’t always get much sympathy, partly because many people assume that they can afford to swallow the costs. South of Los Angeles, in the city of Dana Point, nearly 20 waterfront homes got damaged or destroyed by high seas, and “we’re getting a lot of hate,” Natalie Driessen, a plastic surgeon whose home in a gated community along Beach Road has been damaged by the waves, told me. “A lot of people are saying, ‘Oh, let the rich homes fall in.’” Buying a home here costs $5 million or much more, but many people live in these houses year-round. Driessen, the daughter of a mechanic and a stay-at-home mom, said that she’d gotten to this neighborhood only after putting herself through medical school and working more than 100 hours a week while she had young kids at home. She said that dismissing the community’s plight is unfair. Even more galling to her and other homeowners is the feeling that authorities have abandoned them to the onrushing waves.

When I visited Driessen’s home earlier this month, several neighbors and their contractors huddled among sandbags on what remained of her mangled patio. Every so often, a wave would thunder against a newly built pile of boulders in front of her house, dousing the group with mist and spray. For years, residents in this community of more than 180 homes on Beach Road have sought permission, in and out of court, from the California Coastal Commission to erect such defenses; many have built barriers without permits. The commission, a powerful state agency, rarely approves such projects: Seawalls can accelerate erosion and prevent the “natural inland migration of the beach,” the agency wrote in a letter to residents two years ago. It warned of potential fines, up to $11,250 a day, for unpermitted barriers. When the waves began ripping into homes in earnest earlier this month, the commission approved some emergency permits for 18-foot-high seawalls. The permits themselves cost $1,600 each, and homeowners rushed to pay the fee, then scrambled to find anyone who could throw down beach boulders in the narrow window between tides.

The coastal commission has warned residents for years about the extreme risk they face living in a FEMA flood zone, one where the beach is eroding by about five feet a year. It “provided immediate and ongoing assistance” to homeowners, including issuing 24 emergency permits for barriers, an agency spokesperson, Joshua Smith, told me in a statement. “While seawalls can provide temporary relief for homes, they can also increase erosion and destroy the public beaches that help protect those same structures,” he noted.

In front of Driessen’s home, contractors had started some of that work before nailing down the price, which, they told the group, could amount to more than $100,000 a home. Driessen didn’t see any alternative. “Just build the wall,” she told the contractors. “Build the wall.” In the longer term, Driessen and her neighbors want regular injections of fresh sand and a revetment—a sturdier, continuous barrier down the coast—but they have little guarantee that anything will change. In fact, the commission told homeowners that they’d have to remove within 120 days the boulders they’d already put down—again at their own expense—if the commission did not approve permits for a more permanent solution. The deck, kitchen, living room, and game room of Chris Rogers’s beachfront home got consumed by the waves before the contractors he’d hired could pour out concrete to defend the house. To save the rest of it, he told me, “we need immediate help.” Among the last homes to fall into the sea earlier this month, he said, was one whose owners had been in years-long talks with the coastal commission for protection.

The most intense months are likely still to come. El Nino’s supercharged storms could soak parts of the state with heavy rains. Los Angeles is clearing storm drains and mountain reservoirs to try to prevent flooding. Up and down the coast, bulldozers and heavy machinery are moving rocks and building berms. In areas burned by wildfires, heavy rains can trigger landslides. The extreme weather is already clearing away the coastline. Driving down the Pacific Coast Highway in Malibu, I passed mile after mile of empty lots and the ruins of waterfront properties that burned in last year’s Palisades Fire. Of the roughly 300 homes along the beach that burned, none has been rebuilt. (The mayor told me that only 31 of them were owned by people who used them as their primary residence.)

The people with multimillion-dollar homes on the beach may be making the most obvious sacrifices to El Niño, but all Californians are paying for damages. The residents on El Niño’s front lines have been peppering their elected officials with pleas for help; a bipartisan group of state legislators wrote to Governor Gavin Newsom asking him to declare a state emergency, which he did last week. Disaster in even the wealthiest enclaves can demand government resources. Taxpayers bear the brunt of repairs to infrastructure and public roads. Although many people want regular infusions of new sand and barriers, such solutions can overwhelm city and county budgets and cause unintended consequences elsewhere. Yet if the coast erodes away, everyone loses. Sandy beaches are an economic engine for Southern California, and denuding them drives away visitors.

Few Californians want to face the prospect that beachfront homes may simply no longer be viable along stretches of the coast. Of the more than 180 homes on Beach Road in Dana Point, “I wouldn’t say all of them are vulnerable, but most of them would be,” John Gabbard, the city’s mayor, told me. “There’s going to be a very difficult long-term discussion about how do we protect—if we protect at all—homes that are built on the beach.”

On two occasions in recent years, State Senator Ben Allen, a Democrat who represents Los Angeles County, has put forth legislation to create a loan program for local governments that want to buy beachfront homes and then rent them out to cover the cost of the program until the time comes to demolish them. Newsom vetoed it both times. “We have these beautiful coastal homes that are a part of California life. And we’ve all gotten used to this idea that people can live right up against the water and have these spectacular sunsets and have dinner with your toes in the sand,” Allen told me. But to him, how much society should pay to battle the sea is an open question. At some point, surely the price of building doomed houses is too high.

For the moment, the reality of California’s coasts is that Francois Navarre, in one of the most desirable seaside enclaves in the world, is schlepping five-gallon jugs of water under the Malibu sun; he has no running water amid ongoing sinkhole repairs. He and his neighbors cannot rest easy. Another Kelvin wave, the largest and warmest one of the year, is making its way up the coast of Mexico and is expected to hit Los Angeles early next month. The king tides, the highest of the year, are landing around Thanksgiving and Christmas, and they happen to be the highest in an 18-year cycle. If more El Niño–juiced winter storms arrive, the ocean will almost certainly claim more homes. “They say that this is barely starting,” Quintanilla, the security guard, said. His job is to protect Sea Level Drive, but he cannot protect it from everything.

The Pacific Is a Mess of Hurricanes

2026-09-26T16:37:35-04:00

If you could look down at the Pacific Ocean right now, you’d see three hurricanes spinning in a row, between 15 and 20 degrees north. Furthest east is Polo, an unusually strong hurricane that has been churning off the southern coast of Mexico for nearly a week. To its west, Odalys, once a Category 4 hurricane, is slowly weakening in the open ocean. In the central Pacific, Hurricane Nolo has reached Category 2 status and has crept about 145 miles off the coast of Hawaii’s Big Island, where it’s expected to bring up to 25 inches of rain, potentially causing flooding and landslides. The ocean is a mess of hurricanes, and it’s not the first time this season.

Just last month, four storms formed across the Pacific over a 12-day period. Two of them, Hurricane Lala and Tropical Storm Moke, came dangerously close to Hawaii. Driven by an exceptional El Niño event, these tropical storms have increased in frequency compared with an average year. Typically, at this point in the season, 12 tropical storms would have formed in the eastern Pacific, Michael Brennan, the director of the National Hurricane Center, told me. This year, there have already been 17.

Of the storms in the Pacific now, Polo, which has wobbled between Category 4 and 5, is the monster, and it will likely bring catastrophic damage when it makes landfall on Monday. The warm water temperatures associated with El Niño increase the likelihood of stronger storms, and this year’s event is shaping up to be one of the most dramatic on record. “El Niño is the ‘great nudger’ of weather systems around the planet,” Michelle L’Heureux, who leads the El Niño–Southern Oscillation team at the National Oceanic and Atmospheric Administration’s Climate Prediction Center, wrote in an email. “Very strong El Niños are often able to nudge more.”

Warmer waters also mean that the area where a storm can form expands. Storms tend to show up in the region where Polo is lingering, Brennan said, and generally, by the time they reach the central part of the ocean, near Hawaii, they weaken. But this year, waters in the central Pacific are more than 3 degrees Celsius warmer than usual, allowing hurricanes such as Nolo to form south of the islands and much closer to them. When I asked Brennan how abnormal this was—I’d seen others calling this a historic hurricane season—he said that a season like this comes around every decade or so. Still, it’s a big one.

Nolo is only the latest in a string of disasters that the island of Hawaii has endured this year. In January, the Kilauea volcano erupted for eight hours, and blanketed communities on the island with rocks and ash. Then, back-to-back Kona lows, a type of seasonal cyclone, brought historic flooding in March. Two months later, a 6.0-magnitude earthquake shook the island and damaged nearly 150 homes. In mid-August, Hurricane Lala, a Category 1 storm, dumped some 30 inches of rain on parts of the Big Island without making landfall.

David Ho, a professor of oceanography at the University of Hawaii at Manoa, was at home in Honolulu when Lala hit. His neighborhood flooded, and he was one of 200,000 electricity customers across the state who lost power. More than 200 homes and an estimated $67 million worth of public infrastructure were damaged, according to the Hawaii Emergency Management Agency. “We have a lot of houses that were built before the building codes were modernized,” Ho told me. “So once in a while, you have roofs that just fly away.”

Weeks later, many people, especially those in rural areas with limited access to groceries, are still feeling Lala’s lingering effects. “We always talk about the importance of being prepared, and you want to have three weeks of food supply and everything,” Janice Ikeda, the CEO of the nonprofit Vibrant Hawaii, told me. “Well, we don’t have that anymore, because it got wiped out.” (The state reissued Supplemental Nutrition Assistance Program benefits to more than 65,000 households after the storm.) The damage and losses the island suffered from Lala only make lives and infrastructure more vulnerable in the face of Nolo.

For even one storm to skirt that close to Hawaii would have been an event: The Pacific Ocean is big enough, and the Hawaiian Islands are small enough, that most hurricanes roll by at a much greater distance. But earlier this month, another storm, Lowell, approached the western part of the state, near Kauai and Niihau, pushing hurricane-strength winds and waves ashore. More than 90 percent of utility customers in Kauai lost power.

Nolo likely won’t be a repeat of Lala or Lowell, because it is farther out and should remain 100 miles from the coast, Brennan said. Even so, the island could see more than two feet of rain, coastal flooding, and high surf. The hurricane season doesn’t end for another two months, and the effect of El Niño is likely to only get stronger, according to NOAA’s current predictions. Before the year is through, the “great nudger” will likely hurl more hurricanes Hawaii’s way.

In the Atlantic, the nudge is working in the opposite direction: El Niño generates strong wind shears, which make it hard for hurricanes to form. September should mark the peak of the Atlantic hurricane season, Allison Wing, an associate professor of atmospheric science at Florida State University, told me. Nearly four months into hurricane season, no hurricanes have appeared, and none are on the horizon.

What Does the Climate Movement Do Now?

2026-09-25T10:04:15-04:00

The day the Biden administration celebrated the first major piece of climate legislation in U.S. history, the weather happened to be sublime. The White House’s South Lawn, a billiard-table green, stretched out beneath a cerulean sky. Ben Beachy, who served as a special assistant to President Biden and would help oversee the law’s implementation, told me that “it felt like the dawn of a new era.” The Inflation Reduction Act of 2022 was set to deliver hundreds of billions of dollars in federal investments to clean-energy programs, and to those partying in Washington, the law was a major step toward a long-sought goal: a plan for America’s future that addressed the realities of climate change.

Four years on, the Republican Congress has mostly dismantled the IRA, electric-vehicle sales are in a slump, and the Trump administration is accelerating oil and gas development while axing environmental protections. The U.S. climate movement, Beachy said, “is currently experiencing a bout of dysphoria.” That’s putting it lightly. Longtime allies in the Democratic Party seem to be losing interest. Corporate America has all but ghosted climate advocates after making big pledges to go greener. Philanthropic support has gone squishy. Meanwhile, the planet is on track to speed past a 1.5-degree-Celsius increase in average global temperatures, the threshold that has been considered the bright line of danger since the Paris Agreement.

Climate leaders have sought to project a keep-calm-and-carry-on attitude. “I would characterize this moment in time as just a moment in time,” Michael Brune, the former head of the Sierra Club and the CEO of the Clean Break Fund, told me. This is just an ebb tide, advocates insist, not all that different from past periods when climate slipped to the margins of the political conversation.

There’s no doubt, however, that the dysphoria is real. I spoke with more than a dozen climate leaders spanning the academic, advocacy, and philanthropic sectors who told me, in different ways, that the American climate movement is unsure where to go from here. Having sunk to such lows so soon after climbing to such great heights, climate activists are scrambling to find the movement’s next big idea—some path that will lead them back out of the political wilderness. But right now they’re lost.

If the Inflation Reduction Act managed to be one of the most tepidly—and confusingly—titled laws in American history, it also represented a genuine breakthrough for climate action, an opportunity to begin reorganizing the U.S. economy around climate goals.

Previously, the climate movement had two strategic strands, broadly speaking. The first focused on policy: pushing for international agreements (such as the Kyoto Protocol, which was never ratified) and national laws that would limit greenhouse-gas emissions (such as the Obama-era cap-and-trade bill that died in the Senate). Advocates scored their clearest win in 2015 with the signing of the Paris Agreement—though the signatories have failed to meet the accord’s lofty goals.

The second strategy was to attack fossil fuels directly, along with the financial institutions underwriting them. The movement had some victories: The Keystone XL pipeline died; hundreds of U.S. colleges and churches divested from fossil-fuel companies; coal plants across the country were scheduled for closure.

Then came the Green New Deal and its vision of an American manufacturing renaissance. After the neoliberal consensus that markets could manage carbon emissions fell apart, advocates and their allies in the Democratic Party got behind the idea that the best way to ratchet down emissions would be to build things: solar farms, electric-vehicle factories, high-speed rail. It was technocratic, sure, but also bold, a callback to the FDR era. Under industrial policy, the invisible hand of the market would be enclosed in a tight glove of federal priority-setting. The now-deflated Inflation Reduction Act was the wonky and somewhat less bold version of this ambition.

Today, the overriding theory of change remains the same as ever: Back fossil fuels into the grave and drive a transition to renewable energy marked by speed and scale. But within the movement, there’s little consensus on how to do that. The climate movement does not lack for clever proposals—state-level cap-and-invest programs, grid modernization, make-polluters-pay laws, greater investments in climate-change adaptation. What it lacks is support that’s as resolute as it is broad.

This has always been the climate movement’s Achilles’ heel, even at the high-water mark of the Greta Thunberg–inspired school strikes and mass marches in the years leading up to the pandemic. Climate change doesn’t score well on what public-opinion pollsters call “salience.” A majority of Americans are worried about climate change; in fact, the number of people who say they are concerned is near an all-time high. But when asked to rank climate change among other issues such as the cost of living, immigration, and health care, it turns out that most Americans aren’t very worried.

This is a structural hurdle. For decades, climate advocates have anchored their efforts in grand ideals of generational responsibility, international solidarity, and care for other species. Although inspiring, those ideals might also sound—how to put this?—abstract. So, desperate for a way to connect with Americans and their immediate concerns, climate-movement leaders are searching for some political wave to surf.

The most talked-about notion for a climate-action banner is the same issue everyone seems to be talking about: affordability. In this, the climate movement has a new ally—economics. As recently as a decade ago, “wind and solar were not cheaper. Period. Full stop,” Pete Maysmith, the president of the League of Conservation Voters, told me. “There were a bunch of reasons to do it, but it wasn’t affordability.” Now solar-plus-battery storage has become the cheapest form of new energy generation, as well as the quickest to build. Despite President Trump’s attacks on wind power, more than 90 percent of new electricity capacity that came online in the U.S. in 2025 was from renewables.

That technological and market reality means that advocates can argue, for really the first time, that a clean-energy transition is good, not just for the planet, but also for pocketbooks. This is especially true as Trump’s war against Iran spikes household energy costs. The hope among advocates is that solar, batteries, and wind can go from being coded as liberal to being coded simply as cheap—the Costco of energy as opposed to the Whole Foods of energy, as Bill McKibben likes to say.

This emphasis on cheap, clean energy is mostly a messaging strategy, but it does have real potential as policy. Environmental groups have used the clean-as-cheap argument to push public utilities commissions to approve renewable-energy projects, and Democrats in Congress have floated legislation, such as the Energy Bills Relief Act, that seeks to use the affordability argument to revive portions of the doomed IRA.

But some climate advocates caution, usually off the record, that crowing about how cheap clean energy is risks overpromising. Even a full utility-sector embrace of renewables is unlikely to cut most Americans’ monthly bills in the short run, and will merely slow the projected increase in costs. “Cost plateau” is a tough political sell, and recent polling suggests that the clean-as-cheap message isn’t breaking through.

So many climate advocates are looking to hitch the clean-energy future to populist concerns about AI, which is driving U.S. energy-demand growth and, in case you missed it, is strikingly unpopular.

Some climate leaders think the AI boom could be a way to underwrite the transition to renewables. “We’re going to make the hyperscalers pay for it, and that’s what I think is going to likely catch on with Democratic presidential candidates,” John Podesta, a consigliere to the last three Democratic presidents and a longtime climate-action champion, told me. BeYONCE is the cringeworthy acronym for this idea—bring your own clean energy. A parallel notion is to force the AI giants to offset their electricity guzzling by paying for household energy upgrades—putting heat pumps in every home in exchange for being allowed to build data centers

But that would ultimately ally the climate movement to one of the least popular industries in America. A community that is organizing to stop an AI data center because of its impact on local landscapes or the area’s rural character is unlikely to change positions because the AI facility comes with a giant solar farm. And for more radical climate activists, the notion of getting cozy with the AI bros raises major red flags.

Aru Shiney-Ajay, the executive director of the Sunrise Movement, the youth-led climate group that helped popularize the Green New Deal, floated the idea that challenging AI build-out could be a political bank shot. She’s less interested in using the AI boom to also build renewable energy—à la Podesta—than she is in exploring whether antipathy to the tech giants can be married to antipathy to oil giants. “It’s basically a question of whether or not corporations get to control the land and the air and water around us, or whether ordinary people get to,” she told me.

Leah Stokes, a professor of political science at UC Santa Barbara and the author of the new book The Carbon Wave, also proposed a political bank shot—connecting high energy costs and the data-center blowback to anger with Trump-administration corruption. Frustration with elites and “throw the bums out” populism could, in theory, lead some people to focus on the profiteering, and the massive political influence, of oil and gas companies.

“People don’t like fossil-fuel companies,” Stokes told me. “It’s quite intuitive for a normal person to understand that those people are profiting and making billions while the rest of America can’t even pay their energy bills.”

The climate leaders I talked with believe that Americans are thinking about climate change differently these days. They have a clearer understanding that it is contributing to painful increases in grocery costs and beginning to directly weigh on some families’ most valuable asset—their home; as homeowners in disaster-prone areas of California, Florida, and Louisiana are finding it more difficult to get insurance. It is time, some climate advocates believe, for a wholesale pivot from a morally centered movement to one centered on people’s material conditions—to reframe climate change as a kitchen-table issue for the first time. “You’ve never had more people who are experiencing climate change as a reality, not as a future problem,” Abigail Dillen, the president of Earthjustice, an environmental-law nonprofit, told me. “Real people at actual kitchen tables are talking about extreme heat, extreme weather, the insurance crisis.”

Real talk among real people is, no doubt, essential for any political movement. It’s also insufficient. If a climate message begins with energy affordability or the simmering insurance crisis or the health impacts of wildfire smoke, it then needs to build to the ultimate goal of limiting carbon pollution, whether through the supply-side strategy of stopping future oil and gas projects or through the demand-side tactic of installing more solar panels that are hooked up to EVs. Perhaps that’s a bridge too far?

The climate movement’s best ideas these days are little more than communication plays. The movement has a pile of white papers big enough to practically fuel a power plant’s boiler. But nearly 40 years after the NASA scientist James Hansen warned Congress of the dangers of global warming, a brutal mismatch exists between the number of Americans for whom climate change is priority No. 1 and the scale of the task at hand.

All political movements go through times of boom and bust, and even the bust moments can be a fertile time for experimentation and visioning. On that point, Beachy, the former Biden-administration official, likes to quote (with no small amount of ironic self-knowing) the advice of the arch free-market fundamentalist Milton Friedman, to keep developing alternatives to favorite policies, “until the politically impossible becomes the politically inevitable.”

Inevitable, though, can take an awfully long while to arrive, and addressing climate change is, infamously, a timed test. Despite its many bright ideas, the one thing the climate movement doesn’t have is time.

An Existential Threat to the Oil Industry

2026-09-23T07:00:00-04:00

In late June 2021, the ghost of summertime future visited the Pacific Northwest. For three days, a heat dome hovered over Multnomah County, Oregon, broiling Portland’s drizzly landscape of evergreens and ferns with temperatures of 116 degrees Fahrenheit—much hotter than any resident, even the centenarians, had ever experienced there. Along the rocky ocean shore, sea stars and clams cooked alive in the tide pools. The heat buckled highway lanes and melted the cables that route power to the light-rail network. Not everyone had air-conditioning, and as the days wore on, getting an ambulance or a hospital bed became more and more difficult. Sixty-nine locals died.

At the county’s administrative headquarters, in East Portland, the board of commissioners got to thinking that someone should pay for all of this damage and loss of human life. Until recently, it would have been absurd to hold a person or company financially responsible for a sweltering spell of summer weather. Heat waves were like storms or fires: acts of God or products of chance, depending on your metaphysics. But climate scientists’ understanding of Earth’s atmosphere is getting only more sophisticated, and a growing number of them have come to believe that specific local weather events are traceable to human actions. They say that, with state-of-the-art computer models, they can now link a stifling heat dome like the one that settled over Multnomah County all the way back to the derricks, refineries, and glass office towers of specific oil companies thousands of miles away.

The scientists who launched this field knew that it could have legal ramifications, and hoped that it would. Now their work—still new, and not without its challenges—is being put to the test. During the past five years, dozens of American cities, counties, and states have used it in claims filed against ExxonMobil, Chevron, and scores of other companies, demanding not only payment for disasters that have already happened, but also mitigation funds for those that are bound to strike in the future. At a Multnomah County board meeting in June 2023, the commissioners voted to join them. Only a few hours later, the county filed a claim for more than $51.5 billion against 17 defendants: 14 fossil-fuel producers and sellers, two industry trade associations, and one management consultancy, McKinsey & Co. No such case has yet made it to trial in the United States, but legal experts told me that the Multnomah suit is among the most promising of its kind.

These lawsuits represent a new front for the climate-change movement, and they’ve arrived at a time when the movement is desperate for one. President Trump has withdrawn the United States from the Paris Agreement, dashing all hopes of a near-term global effort to limit emissions. His administration has not only refused to regulate atmospheric carbon, but actively encouraged Americans to vent more of it into the sky. Trump’s appointees have reversed fuel-efficiency standards, slashed investments into solar and wind, and, just last week, proposed the removal of all greenhouse-gas emissions limits on the nation’s power plants.

Meanwhile, summers keep getting hotter. A new environmental nihilism is setting in among Americans: Although a record number of us now see global warming as a “serious threat,” more than 60 percent tell pollsters that humanity won’t do what’s necessary to prevent the planet from careening toward a once-in-an-eon catastrophe. Against this backdrop of despair, the cases brought by Multnomah County and other plaintiffs like it could seem like a godsend for those who are desperate for some kind—any kind—of action. They sidestep Congress and the president and take the fight directly to the prime polluters. With jury awards that might conceivably run into the tens of billions of dollars each, a successful run of these lawsuits could function like a carbon tax on steroids. They could force energy companies to make more urgent investments in renewable energy, if only to limit their own liability. The more extreme climate-change scenarios would become less likely.

But that’s only if the science holds up.

When Noah Diffenbaugh was first starting graduate school in Earth sciences, in the year 2000, everybody already knew that human beings were heating up the planet. Climate scientists had discovered a smoking gun of sorts in the vertical layers of Earth’s atmosphere: The warming was distributed through them in exactly the way that you’d expect if our carbon emissions were the culprit. But no one dreamed of finding human fingerprints on a single weather event, such as a heat wave. “The posture was just: That’s not something we do,” Diffenbaugh told me.

A few years later, in 2003, that posture would change. In an editorial for Nature called “Liability for Climate Change,” Myles Allen, an atmospheric scientist at Oxford, argued that the attribution of extreme events to global warming was not only scientifically possible, but an urgent priority for the field. He urged his colleagues to pursue the idea precisely because it could be useful to future litigants. Epidemiologists had already shown the way. They’d long ago stopped asking whether smoking caused a person’s cancer; instead, they asked by what fraction it raised a person’s risk of getting the disease. Allen argued that climate scientists should do something similar by asking how smokestacks and other sources of atmospheric carbon had raised the planet’s risk of experiencing a flood or heat wave.

As if on cue, within months of Allen’s editorial, a singularly lethal week of extreme heat struck Europe. In France and Italy, tens of thousands of people died. In the aftermath, Allen joined up with Peter Stott, a climate scientist with the United Kingdom’s national weather service, and a young Oxford researcher named Dáithí Stone to analyze the heat wave’s potential link to climate change. The team looked at European temperature records for every June, July, and August going back to 1851, and found that 2003 had indeed been an outlier. For a large swath of the continent, the mean summer temperature had been 4 degrees Fahrenheit higher than normal. The team used a computer model to calculate the probability of that kind of spike in two different worlds: one with human carbon emissions and one without. By simulating summers in both worlds, the researchers determined that in ours—the one with emissions—the risk of Europe’s brutally hot summer was at least twice as high.

[Read: Climate change is moving mountains]

With that paper, a new science was born. “It’s now a recognized subfield with dozens of research groups and hundreds of papers,” said Diffenbaugh, who is currently a professor and senior fellow in Stanford’s Doerr School of Sustainability. In July, the National Academies of Sciences, Engineering, and Medicine published a report on the state of climate-attribution research, after more than 20 years of scientific work. Efforts to link extreme weather events to climate change had markedly improved, it said, thanks to more sophisticated models. Some extreme weather events are still difficult to attribute: Wildfires and hurricanes are tricky because they involve lots of different factors beyond just heat. But the report made clear that for heat waves, such as the one that hit Multnomah County, climate-attribution science is now on solid ground.

Indeed, if all the county had to do was prove that its brutal summer weather in 2021 was attributable to human-caused climate change, that would be a cinch. But the county’s lawsuit wasn’t filed against the whole of humankind; it names specific energy companies. To succeed in court, the plaintiffs will need to prove not one but two connections: first, between the heat wave’s death and destruction and industrial activity writ large, and second, between that industrial activity and individual polluters.

Proving the second link requires input from a different line of research—a tedious accounting project that unfolded over the very same period in which climate-attribution science went mainstream. In 2003, the same year that Allen published his Nature editorial, a geographer named Richard Heede was asked to prepare an unusual report. An environmental lawyer, Peter Roderick, wanted Heede to quantify every ton of CO₂ that Standard Oil had put into the atmosphere, beginning in 1882 and continuing through the company’s merger-enabled transformation into ExxonMobil. By then, humanity’s total carbon contribution to the atmosphere had already been estimated at about 400 billion tons. Heede’s work was meant to show what percentage of that mind-bending total could be traced back to the operations of a single large company.

Heede was able to retrieve most of the data that he needed from Exxon’s own records. After the Securities Exchange Act of 1934 was passed, publicly traded firms began to report their annual production of oil, gas, and coal to shareholders. To get the remaining pre-1934 records, Heede tracked down early histories of Standard Oil in dusty library collections in different parts of the country. Some of the early years remain a mystery. “I don’t think even the company itself has that full record,” he told me. But fossil-fuel production was relatively minuscule during those decades anyway, so omitting that had little effect on the final total.

[Read: What climate change will do to America by mid-century]

After adding up the numbers for Exxon, Heede told Roderick that the company had likely contributed about 5 percent of all the carbon that humans have ever pumped into Earth’s atmosphere. Roderick had commissioned Heede’s study because he’d had a hunch that it would be useful in a lawsuit against the company, but that end result seemed too small to get traction in court. So Heede set out to run the same analysis on other companies too.

In 2013, he submitted his results—a database completed with the help of volunteers and colleagues around the world who visited libraries on his behalf—to a scientific journal called Climatic Change. His manuscript reported that a full 63 percent of the world’s industrial CO₂ and methane emissions could be traced to just 90 entities. When Multnomah County filed suit for heat-wave damages in 2023, its attorneys cited this research; almost all of the named defendants appear on Heede’s list.

Climate scientists had achieved a remarkable bit of planetary detective work. They’d found the human hand in global warming, traced its effects all the way to specific weather events, and then apportioned blame for those events among the Exxons and Chevrons of the world. But not all of those links are straightforward—and in a courtroom, reasonable jurors may find them to be abstract or otherwise unconvincing.

For one thing, whatever advances that climate-attribution scientists have made, their work will never be definitive. They can’t run randomized controlled trials on extreme weather events: No one has access to a second Earth, identical to ours in every respect except for the extra carbon in its atmosphere. So instead, they give us probabilistic analyses that compare the weather that we see today with what we would have expected to see without human emissions, based on records from the time before we pumped so much carbon into the sky. The problem is, scientists have at most a century’s worth of good weather data, and that’s only for North America and Europe. That may not be enough to capture the whole range of natural variation, and even if it were, the data would still describe only one of many possible ways that Earth’s atmosphere might have behaved in an emissions-free universe.

To get around these limitations, scientists try to simulate a host of possible scenarios. They use climate models to generate thousands of ways the weather could have unfolded over time, either with or without human-generated carbon emissions in the mix. Then they count how often an event like the one they’re trying to attribute—a historic heat wave in the Pacific Northwest, for example—appears in each pile of simulated worlds, and analyze the difference.

But a lot depends on how the researchers define what they’re looking for in those piles. To complete their studies, scientists must set parameters for the heat wave that they’re trying to attribute—an exact peak temperature, a number of days, a geographical area—and then search for these across existing data, or in their pile of simulations. They take steps to control the risk of bias in the ensuing analysis, but by defining the event in a certain way, researchers could, in theory, make it look more unusual—and thus more attributable to human emissions.

Even setting that aside, an attribution analysis can say only that climate change makes a certain kind of heat wave more likely to occur. It can’t say whether climate change actually caused each and every such event. Roger Pielke Jr., a climate-policy expert at the American Enterprise Institute, told me that to slide from one to the other is like saying that a high batting average caused a home run.

Ted Shepherd, a climate scientist at the University of Reading, in England, who has been critical of this mode of attribution, has been pushing the field toward what he calls “storyline” attribution. In this method, scientists re-create a specific weather event by doing their best to put the details of its atmospheric mechanics into a model. Then they watch how it plays out, with and without human greenhouse gases in the mix. These storyline attributions avoid some of the issues that plague the other type, but they have their own simplifications and potential biases.

In the Multnomah County case and others like it, judges will have to evaluate whether attribution science, however it’s done, is admissible in court. Robin Craig, an environmental-law professor at the University of Kansas, told me that she’s not sure whether every one of them will reach the same conclusion. Even some of the researchers who are working in this field are wary of its use inside a courtroom. Park Williams, a hydroclimatologist at UCLA who has done important work on attributing wildfires and droughts to climate change, told me that studies of this kind are impossible to verify. In a sense, the research methods themselves will be on trial in these cases.

Climate science is already politically contested. For most of its history, researchers have been able to argue, accurately, that when they describe the need to reduce carbon emissions, they’re just following what the data indicate—and offering neutral counsel as to how a planetary disaster can be avoided. But from the very start, extreme-weather attribution has been pitched as a tool for activist lawsuits: Allen’s editorial from 2003 began by asking, “Will it ever be possible to sue anyone for damaging the climate?” This intermingling of research work and legal practice has been a driving force for the whole endeavor.

One of the original members of the committee that reviewed the state of climate-attribution science for the National Academies was the lead scientist at the Union of Concerned Scientists’ Science Hub for Climate Litigation. (After Pielke pointed this out in a Substack post, she was removed from the review panel.) Oil-funded advocacy groups have besieged the remaining committee members with public-records requests, presumably to ferret out any other conflicts of interest or embarrassing exchanges. Organizations connected to major oil companies have similarly criticized Heede’s database of individual polluters as having been specifically created to enable litigation. They also point out that even going by Heede’s own data, the carbon emissions of individual American companies are dwarfed by those from nationalized oil producers, including those in the former Soviet Union and in China.

When I asked Ryan Meyers, general counsel for the American Petroleum Institute, the largest U.S. trade association for the oil industry, about the attribution-science litigation, he responded with an emailed statement calling it an “ongoing, coordinated campaign to wage meritless, politicized lawsuits against a foundational American industry.” He added: “Climate policymaking belongs in Congress, not in a patchwork of courts.”

What if it works? What if U.S. courts accept the science of climate attribution, with all of its tortured counterfactuals? What if some jury somewhere is inclined to see the oil companies as villains and wants them brought low? A single successful claim could inflict astronomical liability; in a warming world, almost every town and county in America might eventually be eligible to file one, and each would have the prospect of producing an enormous verdict.

As satisfying as that outcome would be for the beleaguered climate movement, it may not represent justice in full. By suing companies such as ExxonMobil and Chevron, Multnomah County and the other plaintiffs are blaming climate change and extreme weather on the companies at the very top of the energy supply chain. This makes a certain kind of sense; the companies that extract and sell fossil fuels have certainly enriched themselves while Earth has warmed, and they can afford to pay a lot of damages out of their extraordinary profits. The plaintiffs will argue—as climate activists have for more than a decade now—that these companies understood the damage they were doing and deceived the public about it. This was the strategy that worked against the tobacco companies. When those companies were sued, the discovery process shook loose internal documents showing that they had long known that nicotine was addictive and cigarettes caused cancer, and had concealed this knowledge.

But if energy giants did understand global warming’s risks and hide them from the public—and some evidence that this is true has already been made public—would that let the rest of us off the hook? For decades now, we’ve all had good reason to believe that carbon emissions are heating Earth, and very few of us have renounced the use of fossil fuels, no matter how much we profess to care about climate change. We fill up our gas tanks, heat our homes in winter, and board cross-country flights to see family. The construction of our built environment would not have been possible without fossil fuels, nor would the functioning of our agricultural system. The lesson of attribution science might be that climate change is attributable to all of us.

Oil companies don’t want to pay for their share of it, but neither do most people. Degrowth arguments that call for humans to consume less or embrace a future with fewer people are woefully unpopular, and so are policies that would require Americans to pay more to burn carbon or compensate developing countries for the carbon that we have already burned.

Even supposing that energy companies are not just responsible for climate change, but uniquely so, that doesn’t necessarily mean that they should be held liable for all the harms that flow from it. “Disasters are never just weather,” Greg Lusk, a philosopher at Durham University, in England, who has written extensively about the epistemology of climate models, told me. They’re also a function of human choices about infrastructure and emergency response. During the ultra-lethal European heat wave of 2003 that killed nearly 15,000 people in France alone (and went on to be the test case for the first climate-attribution paper), many health workers happened to be away on holiday, which reduced France’s capacity for dealing with the crisis. Similarly, if more of Multnomah County’s residents had owned air conditioners in 2021, its heat dome would have had a much lower human cost. Our global conversation about who precisely ought to be held accountable for events like these, and to what extent, is likely only just beginning.

Either way, in the here and now, the energy companies are fighting these cases as if their survival depends on it. “There is a view that if you can get one to trial, you can make the fossil-fuel companies look bad and sway a jury,” Jonathan Adler, a law professor at William and Mary, told me. But whether Multnomah County’s lawsuit, or any other like it, ever gets to trial is an open question. Next month, the U.S. Supreme Court will hear arguments as to whether federal law preempts a claim that Boulder, Colorado, filed against several oil giants in state court for the costs of adapting to wildfires and drought. The plaintiffs in Portland will be watching closely. “A broad ruling from the court in the Boulder case could make a lot of the others go away,” Adler said.

Oil-and-gas companies may be able to weather a few losses. They could end up shelling out billions in damages, and still make enough profit to allow fossil-fuel production to survive and thrive. Whatever happens, this wave of climate lawsuits likely won’t be the last of its kind. The legal context for these cases will evolve. Company records will keep piling up, and courts may come to recognize new harms. Scientists will get better at measuring all the ways that Earth is changing. The weather will very likely worsen: People will experience more lethal heat domes hovering over cities that were once known for their mildness; more floods that wash away homes; more wildfires that orange the sky and fill it with lung-damaging smoke. More people are going to want someone to pay.

City Rats Are Going on Birth Control

2026-09-18T11:14:15-04:00

A city can have one of two levels of rat. There are cities where rats are glimpsed at night, furtively scurrying into a dark corner or among the train tracks. Then there are cities where the rats are so numerous and so brazen that they work out private disputes by brawling midday on a train platform, or they lounge on your front porch, eating fallen fruit. When rats reach that second level, city officials need some answer, any answer, to their rat problem. Washington, D.C., has reached that level, and Mayor Muriel Bowser’s answer involves rat birth control.

Contraception for rodents has become more popular over the past few years because the most commonly used rodenticides kill painfully, as the poisoned rodent internally bleeds to death over a period of days. Not a lot of people love rats or mice, but people don’t love causing that much suffering, either. Many are fond, too, of cats, owls, and hawks, which can be poisoned when they eat poisoned rats. New York City’s Flaco, the beloved Eurasian eagle-owl sprung from a cage in the Central Park Zoo, was found to have high levels of this kind of poison in his system when he died. And to top it off, some rodents are evolving resistance to poisons.

D.C. hopes that rats will lap up a liquid contraceptive called ContraPest from plastic bait stations. It is not the only city to try the approach: Chicago is having luck with rat birth control in Wicker Park; Boston had mixed results in Jamaica Plain. Most of the available research on ContraPest has been done by the staff of the company that makes it, SenesTech, sometimes in collaboration with external researchers. One study at a poultry farm found that a year of continuous baiting with ContraPest led to a 94 percent reduction in rat activity as measured by camera traps. The company says that if the rats continuously eat the bait, they stop having pups, a finding confirmed in a laboratory setting.

But that’s a big if. A prior attempt to use ContraPest in D.C. was inconclusive, according to Andre Pitman, the program manager for the city’s Rodent and Vector Control Division. In rat hot spots such as Adams Morgan, where evening revelers have been known to throw greasy pizza plates on the ground, rats simply had too much other food around for the birth-control liquid to hold much appeal. “We couldn’t get the rats to drink it,” Pitman told me.

So this time, Pitman said, D.C. is making what, to his knowledge, is the most aggressive push against rats in a very long time. First, the city is going to try to kill rats by heaving poison directly down their burrows, which will then be filled in. City workers will return multiple times, looking for burrows that the rats have reopened, adding more poison, then filling them again. The next week, the same procedure will be followed with a different poison, which clings to rat fur and is ingested as they attempt to lick themselves clean. Finally, starting in week three, the city will put out ContraPest for 15 days. All the while, city workers will clean alleys and encourage residents to remove ground cover, containerize trash, and generally boost hygiene levels so that rats are more likely to take the bait. Post-bar snackers are requested to keep things tidy too.

It is understandable that the nation’s capital, facing a high number of resident complaints, has chosen to throw everything it’s got at the rats. But the downside to the guns-blazing approach, from a scientific standpoint, is that if it succeeds, no one will have any idea why. If burrow numbers are down by the new year, no one can say that birth control works—only that the combination of two poisons, ContraPest, hygiene, and other measures works. And Pitman acknowledges as much. “It is not a fertility pilot,” he said. “It is a rodent blitz.”

Fertility control of wildlife, which is used in elephants, wild horses, kangaroos, and pigeons, among other species, does work sometimes, and advocates for this method worry that premature deployment of oral contraceptives for rodents risks making the entire field look bad. Ultimately, no real independent studies show that SenesTech’s products are safe or effective, Lawrence Green, the interim managing director for the Botstiber Institute for Wildlife Fertility Control, told me. (SenesTech CEO Michael Edell responded, “We have been doing this a long time and have case studies and data to back up our claims.”) Assuming ContraPest can work in cities, rats would likely have to take it for longer than 15 days to see lasting results. As people who take birth control generally know, if you miss a dose, it’s less effective, and you have to keep taking it for as long as you want to not have babies. ContraPest is really meant to be used indefinitely: SenesTech even offers a “Subscribe & save” option.

And that familiar e-commerce option points to a larger truth that everyone I spoke with agreed on: There is no end to this war. The quest to control rats is not about winning. It is about nobly devoting oneself to an unwinnable fight. Because rats spread disease, we must struggle against them, clear-eyed in the knowledge that, unless we are dealing with a small island, eradication is not possible.

The reason that rats will never be conquered, at least not in our lifetime, either with poison or with birth control, is easily explained by two facts of rat biology: They have many babies, and they love to explore. You can nuke rats with poison or slowly usher them into a population crash with contraceptives, but all you will have done is create unoccupied territories for young, exploratory rats to come occupy. D.C.’s current blitz is focused on Adams Morgan, Chinatown, and Eastern Market, but rats live in other places in Washington. They will be happy to move in once those neighborhoods have space.

Niamh Quinn, a human-wildlife-interactions adviser at the University of California Division of Agriculture and Natural Resources, puts it this way: “We live in an infinite sea of rats.” Just beyond the boundaries of any rat-control project lie more rats, waiting to move in. Pitman, a longtime rat expert since before he joined the city, knows this too. “Rodents were here long before humans,” he said. All we can hope is to control them.

Why a Sea-Lion Epidemic Is a Bad Sign for Humans

2026-09-18T11:06:00-04:00

You have likely never heard of synanthropes, though you surely know them. And they know you. Synanthropes, whose name is derived from the Greek for “together with man,” are not our pets or our herds of domesticated animals but something like our oldest frenemies: beings from which we have long sought and failed to decouple ourselves, that now reside within our communities, and that have become an ineluctable part of our lives. They include insects like bedbugs, which have evolved within urban environments to become uncommonly dependent on humans for their survival, as well as house sparrows, seagulls, blackbirds, and pigeons, all of which have adapted themselves to take advantage of cities while evading their dangers. In Singapore, synanthropic river otters have begun rampaging through local markets and aggressively pursuing tourists for food. Black bears are bathing in backyard swimming pools in California.

All of these creatures straddle the human world and the wild. And that can make them uniquely perilous to human health.

In Chile, which has a 4,000-mile-long coastline, the most conspicuous synanthrope is the South American sea lion. About 130,000 of these pinnipeds, which can weigh 660 pounds and resemble giant waterlogged chocolate Labrador retrievers, live in raucous colonies on the water’s edge. On account of their ubiquity, pretty much everyone in Chile has had some kind of run-in with the barking beasts.

“As a Chilean, you can easily see sea lions,” Victor Neira, a veterinary scientist and virologist, told me from his lab at the University of Chile in Santiago. Sea lions have long been habituated to human handouts in Chile, and many of the urban colonies they’ve established are located near fish-processing plants, where the animals gorge on fish guts, bycatch, and other detritus dumped back into the ocean. “They’re not aggressive by nature,” Neira said. “They see people all the time, so they generally just move out of the way for you.” He thought for a second and seemed to change his mind. “That being said, if you mingle with them, they will try to bite you. So usually you have to kind of run or you have to scream.”

Neira has had more than his fair share of sea-lion encounters, because he’s spent the past decade tracking novel viruses in wild and domesticated animal populations across South America to sound the alarm about potential threats to humans. The more blood and fecal samples he and his team collect from seabirds and marine mammals, the more likely they are to detect novel pathogens or viral variants that have just appeared in the region. This sort of epidemiological research is undertaken in countries across the world because scientists now recognize that synanthropes are among the most important vectors transmitting novel pathogens to human populations. The hope is that a network of scientific early-warning systems, like the one Neira and his team have designed, can give humanity a head start in identifying new pathogens and stamping them out before they explode into the next pandemic.

In December 2022, Neira discovered a disturbing anomaly in blood samples he had been routinely collecting from migratory birds in Chile. H5N1—an influenza virus that causes acute respiratory symptoms and death in birds—appeared to be circulating among pelicans, cormorants, and terns. All three species regularly cross paths with other birds and sea creatures, and the viruses they harbor, along their migration routes. Neira was terrified. In the preceding 20-plus years, H5N1 had killed tens of millions of wild birds around the world, and hundreds of millions of poultry had been culled to contain its spread. He girded himself for a devastating avian outbreak. When the viral wave arrived, though, the casualties went much further.

Neira told me that in early January 2023, clusters of dead seabirds began washing up on shore. Later that month, Neira’s team witnessed the first cases of H5N1 among Chile’s sea-lion colonies. After years during which the monthly number of sea lions stranded on land (a measure that marine biologists use to assess disease severity and death counts) never surpassed 100, January 2023 saw a sudden spike to about 250. One month later, that number had doubled, and in June, more than 4,500 stranded sea lions were observed by scientific authorities. Overall, somewhere between 10 and 25 percent of the sea-lion population in the country died in the six months after the first H5N1 case was detected among Chilean seabirds, though Neira estimates that the proportion could be double that. It was a sudden and deeply frightening mass-mortality event, caused by a virus that had gained the ability to infect and kill mammalian hosts as well as avian ones. The corpses were piling up on shore, attracting flies and other scavengers—and risking further spread of the virus. Eventually, Neira said, the Chilean government built halls to house the remains, a sign of the scale of the crisis and its lasting impacts on Chile’s ecosystem.

More epidemics were caused by novel viruses in the past two decades than in all of the 20th century, suggesting that human vulnerability to opportunistic pathogens has shifted dramatically. The H5N1-fueled mass-mortality event among Chilean sea lions is an unsettling illustration of how synanthropes are ratcheting up this risk by bringing novel viruses carried by wild and migratory animals close to humans. H5N1 is still spreading through animals worldwide, including in the United States. More than 1,000 American dairy herds have tested positive for H5N1 over the past two years; cases among poultry and cattle alike may tick up this fall during the seasonal migration of wild birds.

So far, that hasn’t translated into a major human outbreak. Since 2024, only 71 human cases of H5N1 have been reported in the U.S. Two people have died. No evidence indicates that the virus currently infecting cattle can spread between people. It’s a lucky break, given how ubiquitous H5N1 is among hundreds of wild bird species that regularly come in contact with poultry and cattle herds, which are perfectly positioned to infect human populations with opportunistic pathogens. That luck might not hold forever, though. A recent report from the World Health Organization suggests that H5N1 viruses are continuing to mix and match genetic material, and a 2024 study from Scripps Research Institute in San Diego found that the H5N1 variant circulating in U.S. cattle is only a single amino-acid mutation away from being able to efficiently infect humans.

As animal populations continue to come into closer proximity with humans, the threat posed by synanthropic viruses will only grow. Viral pathogens can now be efficiently borne from the most remote places on Earth into the heart of human metropolises. It just takes an animal to bring it from there to here.

This article has been adapted from Dan Werb’s new book, Our Wild Familiars: How Animals Are Adapting to Cities and Reshaping the Natural World.

The AI Pandemic Isn’t On Its Way

2026-09-11T13:12:58-04:00

Last month, a team of scientists published a high-profile paper describing how they’d taught an AI model to help produce brand-new, deadly viruses. The AI—a “genome-language model” trained on DNA sequences—had designed multiple new genomes that scientists then synthesized into functional viruses. And some of those viruses were capable of killing bacteria in a dish.

News of AI-made viruses went over poorly. The loudest reaction to the study (which had been released as a preprint about a year ago) was fear. Now that AI can design viruses that go after bacteria, perhaps someone could use the technology to manufacture pathogens that harm people—sparking a major outbreak, even a devastating pandemic. This potential threat evoked some of the same concerns that politicians, researchers, and tech leaders (including Bill Gates) have raised. And these fears are only gaining traction: On Thursday, Anthropic released a report that sparked a round of doom-laden headlines, detailing several instances in which the company banned accounts that were using its LLM, Claude, for what could have been reputable research efforts—or efforts to develop biological weapons.

AI, which has repeatedly proved itself capable of exceeding user expectations and bypassing guardrails, is a formidable force; many fears about its potential biological threat stem from its relentless pace of development and the possibility that it could make independent breakthroughs, exceeding researchers’ own understanding, especially if AIs eventually achieve recursive self-improvement—that is, models designing newer and smarter models. Kevin Esvelt, a biologist at MIT and a co-founder of Secure Bio, a biosecurity nonprofit, told me that although he thinks the probability of an AI-driven pandemic remains quite low, the consequences of such a catastrophe would be so large that “it’s very important that we drive that probability down as low as we can.”

But many other biologists, especially infectious-disease experts and virologists, told me that the threat of a pandemic need not loom particularly large on the list of AI-made threats. To the scientists most familiar with the minutiae of growing, modifying, and testing viruses in a lab, and the havoc they can wreak in the real world, the viruses that have already arisen through evolution, or will soon enough, still present much more immediate perils. In other words, humans should be investing aggressively in pandemic preparation, but not because of AI.

To be clear, viruses made with the help of AI could pose some threat. The experts I spoke with agreed that, in the wrong hands, AI could be used to speed infectious destruction, to a degree. To create the bacterium-killing virus, scientists led by a group at Stanford University and the Arc Institute fed the AI model, called Evo, only viral genomes that attack bacteria. Those genomes are much simpler and smaller than ones coding for viruses that go after humans. In theory, though, with a similar technique, researchers could train AI on the genomes of viruses that infect humans, with the aim of manufacturing novel pathogens. Or perhaps—as some people worry is likelier, easier, and more dangerous—users could ask AI to adjust specific traits of viruses known to cause disease, until those pathogens can skirt available vaccines.

Scientists have long been able to manipulate viral genetics without AI. This kind of “gain of function” research has stirred its own political and scientific controversy—especially in discussions about the origins of COVID-19—and the Trump administration has restricted federal funding for it. But current AI models have the capacity to “put a supreme battery in the back of the things that people were already doing,” Brandon Ogbunu, an evolutionary biologist and infectious-diseases modeler at Yale University, told me. Tom Inglesby, the director of the Johns Hopkins Center for Health Security, worries about bad actors using AI to surmount some of the limits of their own technical expertise. Ideally, models powerful enough to suggest new viral genomes should be subject to strict, formal review, Inglesby said; perhaps they could even be limited to training on only certain types of benign data. (For instance, the team that built Evo trained it only on viruses that kill bacteria, and withheld any data about viruses that can harm more complex life forms, including humans.) Esvelt advocates for restricting access to certain models entirely.

Other researchers, however, including the virologists I spoke with, pointed out that another set of safeguards is already in place: how little we humans—the creators and users of AI—understand about pathogenic viruses. For AI to map out more effective pathogens, people must be able to feed it data on what makes viruses highly transmissible, dangerous, and able to circumvent immune safeguards.

But these concepts are complex and murky enough that the chasm AI would need to cross is wide; researchers lack much of the basic knowledge they’d need to train an AI properly on such topics. “Right now, the data is junk going in,” Seema Lakdawala, a virologist at Emory University, told me. Generations of virologists have not come up with firm answers to key questions about what makes certain viruses better at causing diseases than others, or why some people are more susceptible to particular infections than others. Even expert virologists may struggle to ask AI the best prompts to guide it in useful directions.

In her own research on viral transmission, Lakdawala hasn’t found LLMs such as Claude and ChatGPT particularly helpful or correct, she told me, even when she’s asked them questions far simpler than “How do I make a pandemic virus?” (for example, basic questions about the flu-virus genome). Brian Hie, the lead researcher on the team that designed Evo, told me he’s also unconvinced “that AI would even be the method of choice for someone trying to make a dangerous pathogen.” It does not know enough about pathogens, because we don’t know enough about pathogens.

Esvelt, at MIT, argues that AI might not need to understand viruses perfectly, or even that well, to stumble across a dangerous iteration of one. AIs might also improve rapidly enough to make more sense of the existing data than human researchers ever have. But no matter how good AIs get, they so far have no way to build and test new pathogens, Gigi Gronvall, a health-security expert at the Johns Hopkins Bloomberg School of Public Health, told me. Even a very motivated human—whether they wanted to spark a pandemic for a government or terrorist organization, or for their own ends—would still have to pick through AI-generated possibilities, successfully synthesize them, assess them for enhanced (or at least sufficient) transmissibility and virulence, produce a sufficient quantity to do serious damage, and then release that payload of new virus into the world. All of that requires substantial skill in the lab, as well as expensive equipment and lab animals. It’s not impossible, but it certainly isn’t easy, either.

The practice of building new viruses in a laboratory is fickle: “Even if I give two experienced scientists the same protocol, they’re going to do it differently, and the results could come out differently,” Linsey Marr, a viral-transmission researcher at Virginia Tech, told me. Tinkering with viruses the old-fashioned way has also taught researchers that many hypotheses about what should make a virus more contagious or dangerous don’t pan out. Researchers cannot simply tune up a virus’s transmissibility or virulence like a dial; such changes can come with serious costs, including rendering the entire pathogen nonfunctional—incapable of infecting or hurting a person at all.

Evolution is already a master at tinkering with life forms and building upon their most successful—or dangerous—traits. Viruses are constantly trying out new versions of themselves and aren’t limited by human expectations about what “dangerous” agents of disease look like. Evolving pathogens also need no assistance from malevolent actors to start epidemics; in nature, where species constantly mingle, new outbreaks begin all the time. And nature has an elegant way of dealing with mistakes: Unviable iterations of life simply never come to exist at all. As good as AI is getting, “I can’t look anyone in the eye and say that process is any more harmful than what nature is already operating,” Ogbunu, the Yale evolutionary biologist, said.

The virologists I spoke with did agree that AI models should be monitored and regulated, but were wary of limitations that would hinder AI’s ability to speed the development of new vaccines, shore up viral surveillance, boost epidemiological modeling, or otherwise benefit public-health and infectious-disease research. (Even the types of bacterium-killing viruses that caused such an uproar could be deployed against antibiotic-resistant infections.)

Worries about killer viruses are already interfering with such lines of research. Among the examples of potential threats in Anthropic’s report were gain-of-function work on chikungunya virus and research about how flu viruses adapt to mammals—projects that could have been malicious or, in theory, could improve therapeutics or outbreak detection, respectively. (The company said in the report that it couldn’t confirm whether the queries represented the pursuit of knowledge for the public good or attempts at bioterrorism, and had erred on the side of caution.) Lakdawala and Marr told me that, recently, when they have asked Claude or ChatGPT about, say, genetic quirks in the flu-virus genome or using UV light to inactivate viruses, the AI flagged or halted their conversations. “It happens to me almost daily,” Marr said. (An OpenAI spokesperson wrote over email that “today’s models are specifically trained to withhold biological information that could enable harm” and that researchers “pursuing legitimate biological and life sciences work” can apply to its Trusted Access programs, which require proof of identification. Anthropic did not respond to a request for comment.)

Lakdawala and Marr are now trying to develop their own framework to help AI models distinguish legitimate scientific pursuits from nefarious ones. Both say that it’s difficult: What would the criteria be? Who would be the final arbiters of those criteria and of ambiguous cases? And none of those guardrails would address the most extreme concerns about viruses created by AI—that a self-improving intelligence might eventually be able to deliberately design, manufacture, and release a pathogen. For now, though, AI is still a tool. We’re still in charge of the questions we ask it, the data we feed it, and the fate of any virus that it might suggest we create.

Zoos Are Doubling Down on Very Unnatural Habitats

2026-09-08T08:00:00-04:00

Photographs and video by Cody O’Loughlin

Hudson, a 10-year-old arctic fox at Massachusetts’s Stone Zoo, was not doing all that much when his caretakers came by his enclosure one sunny morning last month. But the moment he detected their approach, he scurried over. His eyes focused on the pink, kibble-filled puzzle box they were holding—a programmable, lab-made device that’s part of an experiment in letting captive animals play around with technology.

In the past year, in partnership with a team led by Rébecca Kleinberger, an engineer and a computer scientist at Northeastern University, Stone Zoo has introduced Hudson to his puzzle feeder, two black bears to a motion-sensitive bubble machine, a family of colobus monkeys to an interactive cabinet of mirrors, and more. The zoo is not alone in its tech curiosity. At the Maryland Zoo, caretakers have debuted a hunting simulation for polar bears, in which a robot arm extends and retracts a romaine leaf smeared with peanut butter—mimicking a prey animal poking its head out of an ice hole. At Zoo Atlanta, keepers have partnered with engineers at the Georgia Institute of Technology to develop an interactive audio wall where elephants can insert their trunks into motion-sensitive holes to trigger a variety of low-frequency sounds. At some facilities, penguins have been given access to iPad games designed for pet cats, and orangutans have been taught to play on an Xbox.

Cody O’Loughlin for The Atlantic

A gibbon inspects a motion-sensitive bubble machine.

Cody O’Loughlin for The Atlantic

Rébecca Kleinberger, the lead researcher on a team at Northeastern University studying how technology can enhance the lives of captive animals

Hudson the arctic fox makes a case for the benefits that captive animals can derive from tech. His quarters at the zoo, where he lives alone, offer few of the thrills he’d experience in the wild—chasing after live rodents and birds, wooing mates, fleeing wolves and bears. In the past, when I’d come to his enclosure as a visitor to the zoo, he had always been snoozing; simpler toys rarely captured his attention for more than a few minutes at a time, his keepers told me. This time, he trotted. He sniffed. He wriggled his tail. He reared up on his hind legs and used his long-clawed front paws to spin the pink wheel of his puzzle counterclockwise, then clockwise, until it yielded crunchy bits of kibble.

When Kleinberger’s team first approached Pete Costello, the assistant curator at Stone Zoo, and his colleagues about the prospect of introducing the zoo’s residents to technology, his mind swam with images of creatures clutching iPads, doomscrolling their days away, he told me. But as Hudson spun his box—which can be toggled up and down through four levels of difficulty and programmed to release food at set intervals—I glanced over at Costello and the other keepers. They were beaming like proud parents and complimenting the fox’s deft moves: Ooh, good spin … Yup, he’s got it right away!

Regardless of how well wild animals are cared for in captivity, their lives by definition shrink. In enclosures, removed from nature’s shifting circumstances and perils, and, often, the company of their own kind, they “lead very impoverished lives,” Clara Mancini, a pioneering researcher in the field of animal-computer interaction at the Open University, in the United Kingdom, told me. Technology may at first seem to make captive animals’ lives more artificial. But it’s starting to look like a good way to coax them into behaving more wildly.

The field of technological enrichment for animals might have taken off decades earlier, Joey Golden, the curator of animal-behavior programs at the Maryland Zoo, told me. In the 1970s and ’80s, for instance, the scientist Hal Markowitz tinkered with devices that taught Diana monkeys to dispense food from a vending machine, or flung “flying meatballs” for servals. The animals were, by all appearances, interested in Markowitz’s devices. But “what he was building,” Golden said, “was incredibly artificial” to others in the research community who were focused on better replicating wild habitats.

Animal-behavior researchers have since come around to the idea that the aesthetics of an animal’s surroundings are less important than how well those surroundings serve them. Creatures need the unpredictability of the natural world; in the monotony of captivity, where humans are in charge, a device that can shift with an animal’s needs can offer that creature “a little more control,” Kleinberger told me. A wobbly chew toy stuffed with a mini meatball might entertain a fox like Hudson for a few minutes. But it can’t operate on a shifting schedule or become more difficult as the user learns its quirks—unlike a teched-up puzzle feeder.

Cody O’Loughlin for The Atlantic

A black bear at Stone Zoo, in Massachusetts

Kleinberger got her start by creating voice-technology tools for musicians and people who stutter. But several years ago, keepers at the San Diego Zoo asked if she would develop a device to help a beats-loving hyacinth macaw named Sampson control his own playlist. Kleinberger was pleased to see not only that the device not only gave Sampson access to the tunes he wanted to hear, but that he’d bob his head along with the music to attract the types of zoo visitors whom he seemed to particularly enjoy. “He really liked children or people wearing red T-shirts,” Kleinberger said. Sampson, in effect, had made the technology his own.

Determining what animals may get out of different technologies requires some guesses and leaps. At Stone Zoo, a pair of black bears, Smoky and Bubba, are certainly interested in the device that Kleinberger’s team built for them—a motion-sensitive machine that blows scented bubbles. The day I visited, both bears lumbered over to the lemon-scented, then bacon-scented bubbles streaming into their enclosure; they took big, huffing whiffs; flicked their ears; and occasionally licked their chops. The bubbles were stimulating the bears’ keen noses, engaging a sense they might not otherwise use very often. But, of course, the bears couldn’t tell us how much they actually liked the experience. When I asked Mancini how researchers know the full effects that technology has on other creatures, her answer was simple: “We don’t.”

Cody O’Loughlin for The Atlantic

Liam Kelly, an animal-well-being adviser at Stone Zoo, holds a puzzle box for Hudson the arctic fox.

Cody O’Loughlin for The Atlantic

A motion-sensitive bubble machine positioned outside the black-bear enclosure at Stone Zoo

Animal behaviorists generally don’t like to use words such as enjoyment and boredom to describe their study subjects. “Those are very human concepts,” Kleinberger said. But if the goal is to improve captive animals’ lives, sentiments at least akin to enjoyment and boredom are what enrichment researchers must try to measure. Scientists are left to use the animals’ actions as a code, Malu Celli, Zoo New England’s vice president of animal care, told me. Researchers look for certain behaviors that they might feel are lacking in captivity: more swimming by a penguin, more walking by a bear, more foraging by a fox. They’ll also pay attention to how a creature’s senses are engaged—whether an animal’s ears lift or its eyes focus on a machine—while watching for subtle signs of distress. A positive result can be as simple as “Do they come back to it?” Kleinberger said. Or, presented with the same system, do they run away?

Engagement doesn’t necessarily mean fulfillment. An animal that’s spending long periods of time with a screen might just be drawn in by the lights. And in the same way that chasing a laser might infuriate a cat who can never catch its blinking red “prey,” touch screens that ask animals to chase virtual objects might never satisfy predatory instincts. Tech might also suck up so much of an animal’s attention that it doesn’t do much of anything else, or engross a group of creatures so much that they begin to compete for access, Marcela Benítez, an animal-behavior researcher at Emory University, told me. At some zoos, in fact, caretakers have entirely banned smartphones around their great-ape enclosures out of concern that the animals might get too much screen time.

Researchers are still trying to understand the extent to which different species can recognize two-dimensional images or videos of themselves or others of their own kind. Certain brainy creatures can be good candidates for screen-based interactions: Kleinberger and her colleagues have had success with a video-chatting system for pet parrots, highly social birds that commonly end up isolated in a human home. Many of the parrots they studied would chatter or sing loudly to each other through the screen. One parrot in the study, Kleinberger said, started flying more often after a series of calls with another bird that frequently flitted about. Some of the parrots, though, just lay down next to the device or fell asleep.

Several of the scientists I spoke with for this story also happened to be parents of very young children; more than once, they compared the types of enrichment they were tinkering with to the games and toys they’d shared with their toddlers, for better or for worse. Some wondered about the potential for addiction, or even about skewing certain creatures’ perception of reality by overloading them with artifice. But projecting too much of humans’ own complicated relationships with technology onto animals can unnecessarily limit research. Just as animals might derive the same benefits that people do, they also might not experience the same harms. With Hudson, Kleinberger at first worried that the fox might become unhealthily obsessed with his puzzle box. But the presence of the device, she said, instead seemed to boost his enthusiasm for activity in general. It’s hard to say for sure, but maybe the technology reminded him that there was still fun to be had inside his artificial world.

Cody O’Loughlin for The Atlantic

A black bear investigates the bubble machine as it spews bacon-scented bubbles.

The Dark Side of Mindfulness

2026-09-04T09:00:00-04:00

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

In 2004, Willoughby Britton, then a doctoral candidate in clinical psychology at the University of Arizona, began recruiting participants for a study that, she felt sure, would show meditation improves sleep.

Britton, a meditation teacher, was so devoted to the practice that she saw herself as a “double agent,” someone who uses science to promote meditation. Although she had grown up in a nonpracticing Episcopalian home, she’d become deeply invested in Buddhist teachings. She read meditation books in her spare time and at one point considered becoming a Buddhist nun. She was confident that meditation improved most things, including sleep; it had already improved much about her own life.

For her study, Britton split 26 participants into two groups: one that would go through an eight-week-long meditation class, and one that would serve as a control group. She brought them all into a lab to sleep overnight while sensors measured their brain waves, muscle activity, and eye movements.

But when she reviewed the results, she was dismayed to find they were the opposite of what she had predicted. The participants in the meditation group slept more fitfully, with more frequent awakenings, than those who didn’t meditate. They spent more time in “Stage 1,” the lightest sleep phase, from which people are easily roused. “Graph after graph after graph, no matter how you looked at it—number of awakenings, number of micro arousals, number of Stage 1 minutes,” Britton told me, the amount of meditation that the subjects practiced was directly correlated with worse sleep quality.

Britton wasn’t sure what had happened. Had she done something wrong? Designed the study poorly? Was the result a fluke, a random finding that didn’t mean much? Or perhaps meditators just need less sleep. At the time, promoting meditation was more important to her than scientific ethics, and she worried that publishing the results would mean “harming the dharma”—criticizing the teachings of the Buddha. So she didn’t.

On a meditation retreat not long after this, she was chatting about her study with a meditation teacher when the teacher said, “I don’t know why all you psychologists try to make meditation into a relaxation technique. Everyone knows if you meditate enough, you stop sleeping.”

On some level, Britton must have known this. At that point, she had spent years both meditating and studying the brain. Because most kinds of meditation aim to enhance concentration, the practice can act like a stimulant. And just as chemical stimulants such as caffeine and Adderall can interfere with sleep, so, too, can excessive meditation.

What else, she wondered, did meditation teachers know that they weren’t telling people? Gradually, starting with that sleep study, the tiers of thought propping up Britton’s worldview—that Buddhism is a cure-all, and that meditation was always beneficial, and could never cause harm—would collapse. Over time, she would find that meditation, a multibillion-dollar industry that is promoted as a treatment for anxiety, chronic pain, alcohol abuse, addiction, and many other ailments, can sometimes, for some people, cause severe and lasting side effects: depression, anxiety, even acute psychosis. Losing the belief system that had been foundational to her life would lead her to produce a trailblazing body of research about the adverse effects of meditation, and then to found a clinical practice geared toward helping the people—a lot of them, as it turns out—who believe that meditation has harmed them.

Britton has straight blond hair, blue eyes, and a muted affect. When you say something to her, she pauses, as if considering whether or not to smile. A colleague once suggested to her that she might be autistic. “And I’m like, ‘Okay,’” she told me one day recently.

She grew up on a farm in Massachusetts with parents who believed that a person should find a problem that isn’t being solved and embrace it as their niche. Her father’s niche was fermenting waste from local businesses and selling it as compost.

Britton was, by her own description, a weird kid who spent a lot of time by herself. She was always asking strange questions like “What is looking out of my eyes?”

One day during the summer break after her junior year at Colgate University, where she was majoring in neuroscience, Britton went over to the house of her best friend, Sarah, a champion athlete at Dartmouth, and discovered her lying lifeless on the bed, a hunting rifle on the floor and a splatter of blood on the wall. Overwhelmed by the pressure of competitive sports, Sarah had taken her own life.

After Sarah’s death, Britton developed symptoms of PTSD. She grew afraid of crossing the street and of driving. She didn’t like opening doors, because she feared she might find a dead body behind them. At one point, while killing rats at her bench in the National Institutes of Health lab where she worked between college and graduate school, she watched the blood drain from their eyes and wondered aloud to a colleague, “Where does it go?” “In the freezer,” the colleague responded. But Britton wasn’t asking about the rat’s body, but its life force. Where did that go?

She was reluctant to take medication, so her father sent her the book A Path With Heart, by the famous meditation teacher Jack Kornfield. She appreciated Kornfield’s explanation for why humans are so prone to rumination—“over and over we feel the familiar tug of thoughts and reactions that take us away from the present moment,” he writes—and that he offered a way out via meditation. She began meditating for half an hour to an hour each day.

In graduate school, Britton initially focused her research on the neuroscience of near-death experiences, but for her dissertation she pivoted to studying meditation, which had been so steadying for her personally.

Meditation was, even then, having a moment. In 2005, the filmmaker David Lynch established a foundation that aimed to promote Transcendental Meditation in colleges and schools. Soon, tech CEOs were flocking to mindfulness summits to learn about “wisdom in the workplace” or “conscious leadership,” and a mindfulness class for Google employees regularly had a long waiting list. High-powered celebs including Oprah and Arianna Huffington credited meditation for centering them amid their empire building. LeBron James partnered with Calm, a multibillion-dollar meditation-and-sleep app. Today, according to one study, about half of all Americans have tried meditation, and about a fifth meditate regularly.

Meditation’s popularity is well-founded. Although the practice has long been supported by the more woo-woo corners of the wellness industry, as well as by Eastern philosophy, an accumulating body of scientific research has suggested that it has all kinds of measurable benefits. Major studies show that meditation can alleviate anxiety and depression—at least about as well as the antidepressant Lexapro can, but without the cost and with fewer side effects. Even short periods of meditation have shown notable effects on the brain: Several studies have found that eight weeks of meditation training can reduce reactivity in the amygdala, the emotion center of the brain, and another showed that meditation could help activate brain regions involved in attention, memory, and perspective-taking in people with social-anxiety disorder. Other reputable studies have linked meditation to positive mood; improved immune function; reduced chronic inflammation; less back pain; lower blood pressure; enhanced self-awareness; better attention spans; greater compassion; and even reduced feelings of loneliness in older adults.

Beyond the scientific evidence, Britton believes that meditation became popular because it’s a health hack that people can do independently, without relying on expensive medical care. The practice suggests that “underneath all of your thinking is some beautiful, basic state” just waiting to be discovered, she said. It probably helps, she added, that meditation is “this ancient thing from Asia and therefore seems more legit or real or something.”

In graduate school, Britton became a serious meditator in the Western insight tradition, which includes sitting and focusing on an “anchor,” such as one’s breath, sometimes for hours at a time. Britton describes it as an austere practice intended to “train the mind.” She molded her life around ancient Buddhist principles. When colleagues remarked on her plain appearance, she said, “Why would I adorn myself?”

Soon, Britton immersed herself in Tibetan studies. Instead of reading novels at night, she’d read dharma books. Her belief at this time, she said, was that “getting enlightened is the most helpful thing you could do for the world.” She considered dropping out of academia to move to a monastery in Asia, which would have meant renouncing her former life, taking a vow of celibacy, and living under a new name.

But she stayed in the United States and continued her academic work, while also training to teach mindfulness-based stress reduction (MBSR), a secular meditation program developed in the ’70s by a molecular biologist named Jon Kabat-Zinn, and a related course called mindfulness-based cognitive therapy. These classes promise to help everyday people with their stress, low moods, and chronic pain; alongside daily meditation practice, they typically include some non-Buddhist teachings from psychology and literature.

Britton moved to Rhode Island to begin her clinical residency at Brown University’s Warren Alpert Medical School. She didn’t know anyone, and, working 12-hour days, she felt intimidated by the cutthroat Ivy League academic life. To fill the time she had to herself, she started meditating for three hours a day.

One night, after meditating on her balcony for an hour, she went back inside and noticed that her house felt like someone else’s home. A picture of her with two friends looked like a picture of three people she happened to know.

Britton decided to take a walk, and as she strolled the streets of Providence, she felt like a bird sitting on her own shoulder, watching herself. She walked into a creperie where she noticed two people arguing. When they both turned and glared at her, she thought, Oh my God, I forgot that I am a person.

What Britton was experiencing is called dissociation, a state that can occasionally be evolutionarily useful—it can help people push through a task when faced with overwhelming terrors. But according to Britton and other researchers, it can happen to extreme meditators too. Britton told me that extensive meditation can prompt the prefrontal cortex to essentially shut off the brain’s emotional centers, causing feelings to seem blunted. And without the typical emotions that we project onto objects, familiar things—our hands, our homes—can start to look foreign. Heleen Slagter, a cognitive neuroscientist at the Vrije Universiteit Amsterdam who studies meditation, says that part of the way the brain gives rise to our sense of self is by predicting how we will interact with the world in the future. If you spend enough time zeroing in on the present moment, as one aims to do during meditation, Slagter says, your sense of the future can start to recede—and with it your sense of self. This can produce dissociation.

Britton began dissociating a few times a week. She had trouble driving because she couldn’t sense where her body began and ended. She’d watch a sappy commercial and suddenly notice her face was wet, not realizing she had shed tears. She felt porous, like other people’s states of mind were her own, a fact that became especially disturbing when she was treating suicidal patients. The dissociation continued for six or seven years, sometimes happening during meditation, but other times at random.

When she told meditation teachers about the dissociation, she said they told her she had entered a new stage of enlightenment—that she was “riding the ox backward.” She had “seen the truth of no-self,” they said, and this was a good thing. They told her to keep meditating, which made the dissociation worse. Somehow, Britton got through her postdoc, but she wasn’t entirely functional. She didn’t date anyone for a decade.

Lauren Lancaster for The Atlantic

Willoughby Britton at home in Vermont. Now retired from Brown University, she runs an organization that counsels people who have had distressing experiences during or after meditation, such as insomnia, anxiety, depression—even psychosis.

Even before she started dissociating, another layer of Britton’s worldview began to crumble after she moved to Rhode Island and was working at an inpatient psychiatric hospital as part of her residency. One day, someone came in from a meditation-retreat center in a state of florid psychosis. Three months later, another psychotic meditator came in, from the same retreat center. That’s weird, Britton thought. Do they know each other? They did not.

It’s well known that psychedelics such as LSD can spark psychotic breaks, but meditation can produce altered states of consciousness almost as powerful as those induced by psychedelics. The brain likes and expects a certain amount of input; go long enough without doing anything and the neurons might start to fire spontaneously, producing, say, flashes of light or made-up memories. Inuit people in the Arctic Circle have been known to experience a type of hysteria known as piblokto, which might be brought on by the sensory deprivation of long winters. No one seems to know why, exactly, psychedelics and meditation can both occasionally prompt psychosis, but it might have to do with the fact that both can have the effect of “rewiring your brain,” which can be a positive thing but can also mean your brain just doesn’t work the way it used to. It might become more easily influenced, for example, or impart grave significance on unimportant events.

At this point in her research, Britton’s studies were still mostly focused on the benefits of meditation, such as on how it can be helpful for depressed people and substance-abusing teens. But now, when she spoke about her studies at conferences, she’d sometimes briefly mention some of the potential adverse effects of meditation. Afterward, people would corner her in the halls or in the bathroom and tell her they’d had negative experiences with meditation—everything from increased depression to psychotic breaks.

She started to wonder whether the unintended effects of meditation might extend beyond a few anecdotes and her own experience with dissociation. Centuries ago, meditation was practiced mostly by monks and other spiritual seekers in East Asia and India; it wasn’t something ordinary people did, Pierce Salguero, a historian of Buddhism and medicine at Penn State Abington and the editor of the book Meditation Sickness, told me. The point of meditation was never merely to relax or just “clear your mind” but to wholly rearrange one’s relationship to one’s thoughts. This might make sense for a Buddhist monk pursuing the great unbinding, but not for a random Brooklynite who just wants to be more productive at work. The “no-self” teachings of Buddhism can be at odds with the goals of Westerners, who are generally aiming to be their best selves, not to disconnect from themselves.

Britton began to perceive that the research on mindfulness contained a giant lacuna: its potential risks. Meditation researchers, she felt, tended not to find adverse effects among their study participants because they didn’t ask their participants about them. What’s more, many meditation studies are funded by pro-meditation organizations, and some study participants might feel pressured to say that their mental health improved after meditating. There’s no such thing as a placebo for meditation, which complicates studying its effects. When Britton looked for past studies on the risks of meditation, she found mostly isolated case reports and small studies—one of which found that two of 27 long-term meditators had had “profound adverse effects,” including confusion, depression, and severe shaking.

Britton still believed in many of the potential benefits of meditation, but she began to rethink her role as a booster for it. “It just seemed like the world does not need another meditation researcher who’s going to promote meditation,” Britton told me. “We’re like Starbucks—there’s one on every corner.” Assessing the extent and severity of possible negative effects of meditation, she decided, would be her version of a problem to solve—her niche. In 2007, she began interviewing meditators for what would become the first comprehensive study of meditation’s adverse effects.

In 2010, Britton met a religious-studies scholar named Jared Lindahl at a conference. At that point, Britton had been meditating so much that she saw stars even with her eyes open—usually either one blue star or one white star—a phenomenon that Buddhists call nimitta. Lindahl had written his dissertation on visual light experiences of this sort. Meeting him, Britton said, “was having somebody understand a part of my experience that no one else had understood.” They wrote a paper about meditation-induced light experiences together, began meditating together, and, eventually, fell in love. The year she met Lindahl, she also finally published the sleep study she’d conducted in Arizona—the one showing that too much meditation can keep people awake.

Lindahl joined her on the meditation-challenges project, for which they conducted interviews about various types of adverse effects experienced and witnessed by 60 meditators and 32 meditation teachers—a smallish sample that is nevertheless rather large for qualitative studies such as this one. They finally published their first study, called “The Varieties of Contemplative Experience,” in 2017, nearly 10 years after they started it.

The most common adverse effects, reported by 82 percent of the meditators (which, importantly, included only people who had experienced an adverse effect of any kind), were “fear, anxiety, panic, or paranoia.” Second-most common were “changes in motivation or goal,” experienced by 78 percent. (Britton later talked with one meditator who sat on her couch for two years, unable to do much of anything, though most of the motivational effects reported were less dramatic.) Smaller numbers reported experiencing delusions, hallucinations, physical pain, rage, and suicidality. The effects lasted for a median duration of one to three years, and the majority of the sample indicated “moderate to severe” impairment in at least one area.

Although the study suggests that what’s most likely to cause problems is intensive meditation—sitting in silence in a retreat hall for hours or days on end—28 percent of the meditators had issues arise from their everyday non-retreat practice, and 25 percent of those in the study were meditating less than an hour each day.

Being an experienced or even expert meditator didn’t necessarily protect people. “I was like, ‘You’ve written a bunch of dharma books. How come you never mentioned this?’” she told me she asked one of the teachers. “And he’s like, ‘Well, it’s not good advertising.’” Notably, much of Britton’s sample of people harmed by meditation had graduate degrees; Britton believes that this is because high achievers tend to try to “ace” meditation, by, say, ignoring any internal sense that something might be wrong. (This struck me as similar to the endurance it takes to earn a Ph.D.: Power through years of pain and tedium for the promise of a big payoff at the end.)

As a study, “The Varieties of Contemplative Experience” had some inherent limitations. Because it recruited people who had had negative experiences, the study could conclude nothing about their prevalence. The study also couldn’t definitively prove causality—for instance, it’s possible that the meditators’ symptoms happened after meditation, but for unrelated reasons. Some Buddhists would also say that experiencing negative emotions or other bad effects is supposed to be part of the practice. But by and large, the study was well received and widely cited in the field. It acknowledged its own limitations and, if anything, respect has grown for it over time.

In part that’s because subsequent studies, some of which Britton was involved with and some not—and some of which were conducted in part by the most well-known academic boosters of meditation—have provided sturdy evidence that some percentage of meditators do experience negative effects from it, ranging from mild to severe. The frequency of bad effects found varies, but most studies, and most experts I talked with, estimate that about 10 percent of all meditators experience them, with the prevalence being considerably higher in certain situations. Richard Sears, a clinical psychologist and Zen master who’s been practicing meditation for some 40 years, has said that he thinks Britton’s research is important. “Something is happening” to these meditators who have bad experiences, “and we should be aware of it. This is opening up a conversation.”

Britton and Lindahl went on to publish many other studies on meditation’s dark aspects, including delusion-like ideation and feeling bolts of electricity traveling through the body. One study subject grew paranoid at a meditation retreat, left, and on the way home believed that street signs were delivering secret messages just to her; another became convinced that she needed to save the world from a “black Tantric plot” contrived by retreat-center staffers, threw water in her meditation teacher’s face and slapped the center’s director; a third began experiencing “messianic visions” about the “need to sacrifice myself in some way for the benefit of saving mankind” and was convinced that meditation teachers were communicating with him in secret code; a fourth reported that meditation awakened an “energy” inside him with “a will of its own” that once almost compelled him to jump from a balcony because it “wanted to lift up and fly.” None of these people had previously experienced any psychosis or been treated for mental illness.

Different types of meditation seem to have different adverse effects on the people who experience them. In one of Britton’s studies, a “noting” practice—labeling thoughts or body sensations—was more likely to trigger anxiety, whereas focusing on an anchor such as one’s breath was more likely to prompt a depressive episode. Overall, Britton has found that about one in 10 people who have meditated, even once, had an adverse effect that lasted more than a month—a rate that, were meditation a drug, would qualify as “very common” under the standard classification used for adverse drug reactions. More frequent meditation tends to increase negative effects, Britton said.

The body of evidence that Britton and Lindahl have amassed showing that meditation can sometimes cause harm is now bolstered by others’ research. A Clemson University study found that mindfulness can paradoxically increase distress. (Cultivating a hopeful outlook was more effective than mindfulness at shoring up workers’ resilience and well-being.) A 2020 study found that 3 to 7 percent of schoolteachers and university students in the U.K. who took a mindfulness class reported “unpleasant experiences”—including anxiety, distressing memories, obsessive thoughts, and physical pain; another study found that a quarter of meditators experienced “unwanted effects” such as panic attacks, nausea, headaches, dizziness, insomnia, and dissociation, though most of these effects were transitory. A German research group studying more than 1,300 regular meditators around the world found that 13 percent experienced adverse effects, including a small number who experienced “very severe effects with lasting consequences” or “extremely severe effects” that were life-threatening or required hospitalization. Yet another large survey found that a quarter of meditators, and a higher percentage of nonreligious meditators, had a “particularly unpleasant meditation-related experience.”

To some extent, this is to be expected. It would be odd if a single treatment, billed as the solution for dozens of mental and physical problems, alleviated all of these ailments without any side effects. In medicine, “if it’s not doing any harm, then it’s probably also not helping,” Nicholas Van Dam, a psychologist at the University of Melbourne who has co-authored papers with Britton, told me. Along with a team of researchers in Australia, Van Dam found that about 9 percent of meditators in their survey reported a functional impairment arising from a meditation-related problem, including sleep changes and a sense of disconnection from others. Van Dam’s team suggested that people should be warned that meditation is contraindicated for certain groups, just as certain medications should not be given to some people. Miguel Farias, an experimental psychologist at Coventry and Oxford Universities who specializes in the science of meditation, told me that the early literature on cognitive behavioral therapy, from the ’70s and ’80s, recommended against meditating during a depressive episode. “Physical activity is a wonderful antidote to depression,” Farias said. But “meditation makes you physically be more passive.” In a systematic review, Farias and his co-authors found that adverse events among all meditators were “not uncommon,” appearing in 65 percent of the studies reviewed; the most prevalent symptoms were anxiety and depression.

Some of the side effects these researchers are finding may be attributable to the fact that, in the years since Eastern traditions of meditation were introduced to the West in the 19th century, this once primarily religious ritual has expanded far beyond its original intent; its more sweeping remit may have consequences that the practice’s original Buddhist context might have prevented. But as Salguero and others pointed out, early Buddhist texts cautioned that “meditation sickness” plagued even the devout contemplatives of earlier eras: In one eighth-century Chinese writing, the Buddha warns of 50 “demonic states” in which meditators could become trapped, including hallucinations, sorrow, anxiety, suicidal thoughts, and possession by ghosts.

While working on my most recent book, about personality change, I wanted desperately to change one thing about myself in particular: to bring down my stratospheric level of neuroticism, the personality trait associated with anxiety and depression. When I Googled how to reduce depression and anxiety, the first drug-free thing that came up was meditation.

So I signed up for an online mindfulness-based-stress-reduction class—the kind of meditation that Britton also once taught, and which now has hundreds of classes and instructors to choose from around the country.

I found some of the class’s Buddhist teachings, though watered down for newbies, very helpful, and by the end of the eight-week class, a test revealed that my neuroticism score had, indeed, ticked down. This surprised me because I’m bad at meditating, so I would spend entire 45-minute sessions reviewing my to-do list, my worries, and other earthly matters.

What annoyed me the most, though, was the way the meditation teachers talked, which tended toward New Age–isms in passive voice. “There’s a subtle striving in what you shared,” one of the teachers responded when I asked a question about how to make “mindful walking” more restorative. “Maybe you want for it to be relaxing and not challenging. But that’s not the reason we practice. We practice for the sake of practicing and to meet with whatever is arising.”

This is what Britton calls a “thought-terminating cliché”: a profound-seeming phrase deployed by people in power to quash debate—or, as Britton put it, a way for teachers to politely convey “Shut the fuck up and do what I say.”

But many Buddhists and meditation teachers would agree with my instructor: Feeling, and observing, discomfort is part of the point. In keeping with this, a good number of the meditators whom Britton has interviewed told her that when they had brought up their symptoms with meditation teachers, the teachers dismissed them, saying things like “It gets worse before it gets better” and “What you resist persists.”Their teachers would suggest that all negative experiences are inherently growth opportunities, or that perhaps the person was meditating incorrectly. The mentality seemed to be “Try our product and if you don’t like it, then you did it wrong,” Britton said.

This attitude tracks with what I heard from a few experienced meditators who suffered acute distress while meditating. One man told me that after a long retreat, when he began experiencing spasms that would seize up half his face, his meditation teacher just told him to relax more. Two other experienced meditators I talked with, Terran Palmer-Angell and Carmen Maria Alvarez, also told me that some fellow meditators at a long retreat they attended, organized by the Center for Contemplative Research, had experienced psychotic breaks or developed tics, including one man who would rotate his head around compulsively. The teacher and one of the center’s co-founders, B. Alan Wallace, they said, told them that it was just “an involuntary movement” and nothing to be concerned about.

Palmer-Angell and Alvarez emphasized that their experiences reflected this specific retreat, and not meditation in general. Eva Natanya, the center’s other co-founder, told me that “no one has been diagnosed with a psychotic break by a psychiatrist” at the retreat since its founding, in 2020. However, she said that some “individuals develop physical tics” because of a “burst of energy” while meditating. She said these tics are sometimes referred to as “involuntary movements,” and that initially, meditators might be encouraged to “let it go and go back to meditating because that energy will work itself out by itself.” She added that she herself worked with the person with the compulsive head rotations by “asking questions and listening.” She said the center has multiple forms of support available, including meetings with herself and Wallace, psychological counseling, appointments with doctors, and physiotherapy.

Britton remembers peddling a version of this logic as an MBSR teacher herself. “I was trying to tell people ‘Nothing is good or bad, but thinking makes it so,’” channeling Hamlet and Stoic philosophy. For a while, this idea seemed radical, and constructively life-changing: The secret to a good life is to just be okay with whatever! But then her own meditation practice started to make her think “like, Okay, but wait; I don’t feel like I exist now.”That didn’t seem like a good thing, no matter how you thought about it.

Lauren Lancaster for The Atlantic

Having moved away from Buddhism, Britton now finds her own kind of tranquility in the natural world, taking care of her plants and chopping wood.

After Britton and Lindahl published the “Varieties” study, the phone in their lab started ringing. People experiencing psychosis, dissociation, tics, and other problems after meditating were asking for help. The meditators felt confused and ashamed, like they had damaged themselves. Traditional psychotherapists, they said, didn’t understand what they were going through—and neither did anyone else. How can you hurt yourself meditating?

The surge in outreach prompted Britton to open what she called Cheetah House—a play on citta, the Sanskrit word for “consciousness”—and develop it into a full-fledged nonprofit aimed at helping people who have had difficult meditation experiences. Based at first out of her house in Providence, it now serves 200 people a month, entirely over Zoom.

The Cheetah House program consists mostly of talks with experts, sessions with a “care team” member, and support groups with other meditators who have had distressing experiences. The program is priced on a sliding scale; most people pay $50 a session with a care-team member, and $100 or more for a session with Britton herself.

Many of Cheetah House’s clients have (relatively) manageable problems such as anxiety and insomnia, but Britton told me that, once or twice a month, she gets a call from someone whose family member has had a psychotic break triggered by meditation. Far more rarely, people have killed themselves after intensive meditation retreats. “When people say that I’m fearmongering,” Britton told me, “my response is, ‘We haven’t even gotten into the suicides yet.’” She recently decided to start a support group for the parents of meditators who experienced severe psychiatric symptoms after 13 families, four of whose children had already died, expressed interest.

Between half and three-quarters of Cheetah House’s clientele began having symptoms after a meditation retreat, Britton told me. People often go on retreats to try to deepen their practice or to “relax”—but these are generally not relaxing affairs. Many retreats involve waking up at dawn, eating very little, and sitting perfectly still while focusing on one’s breath for hours on end. Several days of this, Britton said, can cause neurons to become more sensitive, such that minor stimuli, like the ticking of a clock or the sound of another person’s breath, can start to feel overwhelming—as can emotions such as rage and grief. Kieran Fox, a psychiatrist at UC San Francisco and an experienced meditator, told me that retreats are “some of the hardest things I’ve ever done in my entire life. And that’s including med school.”

Among the Cheetah House clients I interviewed, retreats were a common trigger for adverse effects. One meditator, Dan Millstein, told me that after a 10-day retreat, he spent years feeling emotionally shut down and disconnected, like he was “living behind a glass wall or a veil.” Another man, Billy, who asked that I not use his last name, had been meditating regularly for a decade when a seven-day silent retreat ratcheted up his physical and mental anxiety: Shivering relentlessly, he started feeling better when he stopped meditating.

About 15 years ago, a woman who asked to remain anonymous was on her second 10-day silent retreat when she began feeling a twitch in her face that developed into an urge to roll her head around. She felt like electricity was shooting through her body. She would shake uncontrollably and at times writhe on the floor. Eventually, she was hospitalized for psychosis. She was later diagnosed with complex PTSD, and although she’s learned how to manage those symptoms, she told me that she still experiences involuntary muscle movements today. Another woman, Rebecca Fischer, told me that, more than 12 years after a retreat she attended, she still has chronic pain in her legs and back.

A good number of Cheetah House’s clients do suffer ill effects outside of retreats. In the days after meditating at home with an app, a client named Susan Small told me she was driving when “my whole sense of myself disappeared,” and she went on to experience “waves of terror” for months. A man named Eoin told me that after eight days of doing 30-minute audio meditations on his own every morning, he started feeling like a “nervous wreck,” on the edge of psychosis, and like he didn’t really know who he was, a phenomenon called depersonalization.

These people and other clients said they appreciated hearing from fellow meditators with similar experiences, and that Cheetah House helped them feel less like a “misfit toy,” as Millstein put it. They’d had a hard time finding support elsewhere, in part because one of the most common recommendations for emotional distress is to try meditating. Many of her clients, Britton said, seem relieved when she tells them it’s okay if they want to stop.

One recent New Year’s Eve, Britton hosted what she jokingly called a “parallel play” party—a kind of get-together for introverts where everyone works on a craft so they don’t have to talk much. She made margaritas, and friends and colleagues brought over magazines to cut up for collages. A few hours in, someone held up an article from the magazine Psychotherapy Networker and said, “Hey, is this about you?”

The story was, indeed, about Cheetah House, and Britton waited until everyone left to read it. One quote in particular stopped her cold:

“There should be a trigger warning for the Cheetah House website,” says renowned Buddhist teacher and clinical psychologist Jack Kornfield. “Too much of it is misleading information and over-the-top fearmongering.”

This was the Jack Kornfield, the mindfulness guru who wrote A Path With Heart, the book that had launched Britton’s journey into Buddhism. He was like a father figure and a mentor to her. Jack Kornfield thought she was fearmongering. Now all she could think was “Thanks, Dad.” She spent a week in bed after that. (Kornfield declined to comment.)

As she continued with her research, Britton gradually started to feel iced out of the Buddhist circles that had once comprised her entire personal and professional networks. She said a Buddhist organization sent her a threatening letter after she interviewed members of their community. One time, after she gave a talk about adverse effects at a conference, Britton said a prominent meditation researcher cornered her outside and yelled, “I don’t believe you!” (The researcher denies saying this.)

Invitations to conferences started to dry up—and when she did receive them, she said, it would turn out to be an invite to debate whether meditation-related challenges even existed—“just an insult to everyone that I’ve interviewed,” she said.

“If you want to build a bridge, you can’t just be like, A bridge is gonna be so cool. You have to think about where could it possibly fail,” Britton told me. “And it’s so unpopular to be that person.” Her former postdoc, Roman Palitsky, told me that Britton suffered for pointing out the problems with the bridge. “Nobody was doing meditation difficulties before her, not seriously anyway,” he said. “And I think large parts of the meditation world kind of turned on her.”

Some of this criticism can be chalked up to differing interpretations of the science, and of how best to treat stubborn afflictions such as anxiety and depression when all of the remedies have potential downsides. “Is it better for an anxious person to be meditating or on Lexapro?” is not an easily answerable question. Fox, the UCSF psychiatrist, told me that some of the neuroscientific effects that Britton attributes to meditation, such as increased activity in the insula, can also be interpreted as a sign of better “autonomic awareness”—basically, being more attuned to the inner workings of your body—rather than a sign of trauma or panic.

But some of the criticism seems to emanate from a reflexive defensiveness on the part of meditation’s promoters, who—perhaps understandably, given the abundant benefits that have shown to be associated with the practice—don’t want to tarnish the reputation of a good thing.

I’ve noticed that people are apt to defend meditation with a zeal I haven’t witnessed since evangelical youth group in my teens. You don’t enjoy meditation? Good, you’re not supposed to enjoy it. You say it doesn’t work for you? Well, you’re just not doing it right. The answer to every problem with meditation seems to be more meditation.

This touchiness can extend to meditation researchers, most of whom are meditators themselves; academic meetings Britton attended would sometimes kick off with a meditation. Nicholas Van Dam, the University of Melbourne psychologist, told me that meditation communities tend to carry a sense that “if you say anything bad, you’re negating the good.” Farias mentioned that after he included a chapter about meditation’s risks in his book The Buddha Pill, some of his friends “were very, very annoyed,” and a couple of them told him that “we could not be publicly seen as associates.”

Some pro-meditation researchers are more evenhanded—recognizing the value of Britton’s findings even if they quibble with aspects of them. Eric Loucks, the director of the Mindfulness Center at Brown, pointed out that a person can have a challenging experience during meditation and still find their overall meditation practice to be beneficial. He also noted that several studies have found mindfulness-based interventions to have fewer side effects than pharmacological treatments, including Lexapro. Still, he called Britton “innovative” and “courageous,” and said her work influenced not only his own views and scholarship—including updates to the standard MBSR curriculum, which the Mindfulness Center now maintains—but also mindfulness research as a whole. “Just looking for the good stuff,” he said, is not his goal. “We really want to know the truth.”

But several meditation teachers I interviewed did seem protective of the practice. Carol Greco, a psychologist at the University of Pittsburgh and an MBSR teacher, once co-authored a paper that aimed to “dispel the notion” that anxiety, pain, and psychosis “are adverse events due to mindfulness meditation.” She seemed chagrined that I was working on this story, writing in an email, “I would rather talk with you about how meditation can be helpful.” When Britton’s studies began coming out, she said, it “brought a heightened awareness of the potential of trauma in a person’s background and how we can create a safe environment.” But her “heart hurt” because “it felt like it was making mindfulness and meditation seem like it’s really dangerous and really bad.” She told me that in her two decades as a meditation teacher she’s never seen a serious adverse event like psychosis or dissociation.

A few of the scholars I interviewed who defended meditation acknowledged some of Britton’s findings, at least implicitly. Richard Davidson, a University of Wisconsin at Madison professor of psychiatry and psychology who is a well-known mindfulness researcher, told me that, as far as relaxation techniques go, meditation is fairly benign. He believes that for most people there’s a greater risk of harm from not meditating than there is from meditating, because the physical and emotional benefits of the practice are so robust, and he’s co-authored studies supporting that claim. But, he added, people with serious psychiatric disorders should meditate only under the guidance of a mental-health professional who also happens to be a meditator—advice that would probably come as news to the many thousands of people who meditate by using apps at home.

The Buddhist scholar and teacher Gil Fronsdal told me that although he doesn’t think meditation causes adverse effects, retreats may exacerbate preexisting psychological problems. (Britton said adverse effects can crop up even for people without prior mental-health conditions.) Fronsdal recalled one meditator who gradually lost touch with reality after a monthlong retreat: She stalked Fronsdal’s wife, became homeless, and spent hours in a medical-law library researching how it was that Fronsdal had managed to exchange his wife’s spine for hers. Having schizophrenic people on meditation retreats has been “one of my most challenging situations as a teacher,” he told me. He said that people with psychosis shouldn’t go on meditation retreats, and neither should people who are “hyper-anxious”—a quality that described me when I sat for a daylong MBSR retreat for my book.

The substantial and sometimes hostile pushback against her work has made Britton hypervigilant. Maybe excessively so: To my eye, she sometimes detects criticism where there isn’t any. When I asked her to describe what it was like working at Brown’s largely pro-meditation research group while publishing study after study finding that meditation can be risky, she wouldn’t say much more than that the situation could be “tense.” (Loucks, the Mindfulness Center’s director, denied that there was tension.) “When it comes to somebody choosing their religion or your friendship,” she told me at one point, “they will choose their religion every time.”

Starting in the early 2020s, Britton and Lindahl struggled to find funding to continue studying the downsides of meditation; they said an organization once rejected their proposal for a grant on “transformative experiences” for being too critical of religion. She says that when they submitted papers to academic journals, the peer-review comments for the papers focusing on the risks of meditation were “six times longer” than for those focusing on the benefits.

Equanimity is a cornerstone of Buddhist thought; Britton herself had repeated the words to rooms full of meditation students many times: “Nothing is good or bad, but thinking makes it so.” But she was starting to think that, actually, some things are good and other things are bad. Sure, you can just accept everything, but sometimes you shouldn’t. She had made a big contribution to the field; it seemed to me like she was feeling done with Brown, with academia, with living in Rhode Island, with all of it.

When Britton first met Lindahl, he persuaded her to start taking one day off a week, something she had not done in two decades. Then it was two days. Then she stopped teaching in the summers. Soon, they were scouting for a plot of land in Vermont they could move to. They now live in an off-the-grid cabin in the woods, where a botanical drawing greets visitors on a wall by the entryway. What is the human soul, the caption reads, but a claw pointing back to the forest?

Britton is consulting on a research project with her former postdoc, Palitsky, who is now at Emory University and running a much larger, longitudinal study of meditation-related challenges. This summer, both she and Lindahl officially left Brown. They’re enjoying exploring “other sections of the bookstore,” as she put it. Running Cheetah House takes up the bulk of her time now.

Britton stopped meditating in 2017, around the time she published the “Varieties” study, and she “began to have a less romantic and more nuanced view of Buddhism.” Now she finds her own kind of tranquility in the natural world, taking care of her plants and chopping wood.

It’s important to emphasize that neither Britton’s research nor her work with Cheetah House clients has turned her against meditation per se; it’s just attuned her to the importance of doing it safely. Meditating for longer than about 30 minutes a day tends to be associated with more negative side effects, she said, and certain types of meditation can be a poor fit for certain people. (For instance, some anxious people shouldn’t do a “body scan” type of meditation, because they are already too focused on their body.) People considering extended meditation retreats should try meditating for a few hours at home first, to be sure they can tolerate long stretches. And those who experience challenges with their meditation practice should consider changing the kind of meditation they’re doing, modifying their practice, or just not meditating. “Treat meditation like any other relationship,” she said. “If you feel like shit with that person, maybe it’s time to break up.”

Spending so much time with Britton made me want to revisit my own adventures in Buddhism and what I had found valuable about it. Mostly, I found old notes about how much I hated meditating: “This is the part of the book where I’m not having a lot of fun”; “I would literally rather give birth than do nothing for 45 minutes”; “I’m just an anxious immigrant who fears you can mindful yourself onto a park bench.”

But I also came across some teachings that I did appreciate, and which I’m going to oversimplify here because I’m not a Buddhist. Among them: The Buddha taught that people should pursue a “middle way” between decadent indulgence and rigid austerity—to not live mindlessly, but to not be so ascetic as to immiserate yourself. “Avoiding these two extremes,” one sutta reads, “the Realized One understood the middle way of practice.” Maybe he was onto something.

When you buy a book using a link on this page, we receive a commission. Thank you for supporting The Atlantic.

The Missing Ingredient of Modern Life

2026-09-04T08:00:00-04:00

Sign up for Ordinary Extraordinary, Ian Bogost’s newsletter on making everyday life vivid.

You have, by now, probably come across the notion of maxxing—looks-maxxing, sleep-maxxing, protein-maxxing, potassium-maxxing, sun-maxxing, testosterone-maxxing. Etymologically speaking, maxxing seems to be a descendant of somewhat recent slang—for those of us who remember the late 20th century, “maxin’ and relaxin’ ” and “to the max” both come to mind—but in practice, maxxing is a reductionist worldview cloaked in expansionist language.

The term is often invoked to describe a narrowly focused goal: If you’re maxxing, you’re likely trying to make yourself smaller, sharper, more chiseled. Maxxing requires sacrifice. It requires careful measurement. And in this way, maxxing is a consequence of our broader optimization-obsessed culture, a fixation fueled by every product, device, app, and service that trains people to collect and analyze data generated by their own behavior.

This culture of optimization has been building for some time now, since at least the 19th century, when industrialism made efficiency a core ideal of working life. Later, when offices overtook factories as the emblem of work, efficiency became more abstract. Productivity began to be measured not only in profits maximized, but also in meetings shortened, messages answered, and bottlenecks removed. Over time, workers started managing themselves preemptively in accordance with the mandate of efficiency, optimizing their own calendars and lunch breaks. This optimization mindset leaked out of business and into everything else. People stopped asking what activities were for and started asking how to do them better, faster, or more completely.

[Ian Bogost: People are thinking about looksmaxxing all wrong]

Efficiency now infuses every aspect of contemporary life, if you let it. Instacart and DoorDash can replace grocery shopping and dining out. People optimize their vacations, maximizing credit-card points and airport-lounge strategies. Kindle streaks and Goodreads goals have turned books into items on a checklist. Wearables have turned exercise into VO₂ or HRV maximization. The more nutrition-obsessed among us have transformed eating into macronutrient acquisition. Every pleasure can become instrumental—with every instrument focused on the outcomes it might produce.

Optimization is terrific if you want to get somewhere else in the future. But it’s terrible for being where you actually are right now. When goals overtake experience, you lose the freedom to encounter your life as it happens. And when you lose that ability, you are no longer living in full. We used to optimize work so we could do everything else. Now we optimize life as though living were another job. Too many of us seem to be maxxing our way out of the experiences and perspectives that give us maximum meaning.

I’ve come to believe that wonder, the most immense of our sensations, is the antidote. Wonder can rise up anywhere. It might arrive after you hike to Hopi Point, nearly a mile above the Grand Canyon’s river floor. Or as you take in the northern lights from Tromsø, Norway. Or, on a different scale, wonder might be expected while you hold your newborn child for the first time, recognizing human life in its primal form and knowing that your own has now changed entirely. For children themselves, wonder remains abundant—in the bigness of a Tyrannosaurus rex in a museum, or the flicker of fireflies in the summer dusk. But for many adults, wonder is too often compartmentalized: a feeling we associate with extraordinary and rare experiences, banished to a realm outside the mundanities of daily life.

[The Wonder Reader: A weekly guide to Atlantic stories that will fascinate and delight]

For me, several aspects of modern life have made wonder seem out of reach lately: the rise of “frictionless” automated services for chores; the decline of serendipity in a world where every trip is Google Mapped; the ease with which everything can be done remotely, averting the happenstances and textures of regular routines; the ability to opt out of ambient noise (and impromptu conversations) by popping in AirPods; and, of course, our tendency to fill the gaps in our days by staring at our phones. But most of all, wonder feels scarce because contemporary life has depleted the value of the moment. Time has become instrumentalized, and as a result we have taught ourselves to treat the present primarily as a waiting room in which to plan for—and, yes, optimize—the future.

Efficiency culture has muscled out the present so completely that it feels difficult to occupy a moment at all, let alone fully. Even the seasons, which help us mark the present amid the endless passage of time, seem to be anxious to overcome one another—pumpkin-spice creamers and “Pumking” imperial pumpkin ales now appear in supermarkets in July. Wonder interrupts that impulse to rush to the future before it arrives. It makes the world stop serving our self-centered plans or intentions, and instead offer something worth encountering for its own sake.

For years, I have struggled with the problem myself. The urge to optimize caused me to flee to the future constantly. I realized that I felt best able to overcome efficiency culture when I treated supposedly unimportant sensations, objects, and events as vital ones instead: sliding my feet into the plush softness of slippers in the morning, feeling the warmth of a mug of coffee in my hands, hearing the gravel crunch under my feet as I walked. None of these acts compromised purpose or plans. All they required of me was to construe what I was already doing as valuable.

[Atlantic Reads: The Small Stuff with Ian Bogost]

Eventually I came to recognize that my long-standing obsession with ordinary things—including many that I have written about in The Atlantic, among them toasters, gas-station signs, backpacks, slushies, and gearshifts—offered a reason to take the present seriously. These objects show that beauty and even genius are often expressed through design choices and simple engineering. Tools we might think of as “useful” are in fact wondrous. The A BIT MORE button on modern toasters is a perfect example: so useful that you might overlook it as obvious, yet absolutely transformative in the making of toast. For some readers, this may generate an eye roll. Cynicism is pervasive, and not everyone has time for or interest in the scenic route. But I believe there are more of us wonder-seekers than the cynics and maxxers among us might like to admit. And besides: The route I am suggesting does not require an out-of-the-way trip. It simply asks you to accept gifts that the universe is already handing you.

Wonder may not improve your résumé, make you hotter, or maximize your income. It will simply remind you that you are a body alive in the world, right now, and that is a better place to begin than most self-help advice would have you believe.

Once you allow wonder to arrive, it proves to be astonishingly promiscuous. Over the past year, I’ve made a point of asking people about what inspires daily wonder. I expected themes to emerge, but I also assumed that those themes would be predictable: sunsets and forests; notable structures or works of art.

Many people did describe the familiar icons of awe—the moment when an airplane breaks through the clouds, and all you can see is the great expanse of sky ahead; the experience of strangers gathering silently along a shoreline to watch the sun vanish below the horizon. But just as many have told me that they’ve found wonder in the mundane—watching a procession of ants carrying things they value, for instance. I’ve corresponded with people who described their affection for an old refrigerator; a sewing machine whose hums and clicks became a friend’s familiar voice; pots and pans that carried decades of family meals; and even a toilet whose owner began to admire it after a repair revealed its elegant method of operation.

Whenever you cease treating the world as scenery or equipment, you’ll find that you encounter the world as something inexhaustibly, constantly present. Oceans and ants, sewing machines and sunsets, strangers and screen doors, microwaves and mountains, rocks and racquets, lamps and laughter all prove equally worthy. Our expectations for wonder sometimes wane, but it is with us all along, waiting to be rediscovered.

This article appears in the October 2026 print edition with the headline “The Missing Ingredient of Modern Life.”

The Netherlands Will Never Be Rid of Louisiana Crawfish

2026-09-03T09:51:56-04:00

Confronted with a seasoned, boiled, arguably delicious crayfish, some of the people of the Netherlands are not sure exactly what to do. “Do you suck the head out?” I heard from the table next to me earlier this summer, at Two Chefs, a restaurant in Amsterdam that hosts a monthly crayfish boil.

Yes is the answer: The best way to eat a crayfish is to shell the tail, then slurp the flavored juices out of the head, and pick out the remaining meat with your hands.

This strategy might not be an immediate transatlantic hit. At the boil, one 29-year-old from Maastricht likened the meal to shrimp but with more work. Two twins told me that the crayfish would need to look a little more chic on the plate to appeal to the Dutch public. But, given the ever-growing population of crayfish in the Netherlands, coming around to eating more of them could benefit the country.

Because if Dutch consumers aren’t sure how they feel about crayfish, crayfish inarguably love the Netherlands. They have invaded the country’s waterways so thoroughly that they have become a real problem, and dealing with them will mean persuading more creatures—including Dutch people—to eat more crayfish.

The crayfish invasion started decades ago, but in recent years, the red swamp crayfish (also known as the Louisiana crawfish) has become a particular pest, Pim Lemmers, an ecologist at Radboud University who has dedicated his career to researching crayfish, told me. In the 1980s, it was brought to the Netherlands as food, and people naively released the animals into the environment. The waterways in the Netherlands are very interconnected and became a “crayfish highway” of sorts, Ivo Roessink, an aquatic ecotoxicologist at Wageningen University & Research, told me. But crayfish can walk and breathe on land, too, and have been spotted in parks, along bike lanes, and at Schiphol airport, where the birds use the runway to crush them.

And they can dig—burrows up to three feet long that can collapse the edges of fields into the water and that some people fear could damage the country’s extensive system of dams, dikes, and canals. Scientists emphasize that the risk to dams and dikes is low; from their point of view, the bigger risk is to other creatures that live in the water. Crayfish are omnivorous: “They totally destroy underwater vegetation,” Roessink said. They also eat detritus, amphibian and insect larvae, snails and other invertebrates, small fish, and one another—a sight I witnessed for myself.

But crayfish cannibalism will not end the invasion; in fact, although the Netherlands’ central government is working on a program for crayfish management, getting rid of the species will likely be impossible. “I know of one example where they effectively eradicated crayfish, and that was in the U.S.,” Roessink said. “There was an infected pond, and they turned it into a parking lot.” (Resorting to this solution in the Netherlands would require paving over the entire country.) Research in the Netherlands and elsewhere has shown that even extensive trapping won’t necessarily eliminate crayfish. The most anyone can hope for is fewer of them.

Some water boards of the Netherlands are trying to persuade all sorts of creatures to eat more crayfish, including herons, eels, and European catfish, which are themselves something of an invasive threat. And, in theory, one of the crayfish’s most effective predators could be Dutch people.

Stefan van Ekris and Maarten van der Sanden, who are both managing directors at CrawFish Farm Holland, are trying their best to make it happen. This July, I joined them as they put on waders and immersed themselves in irrigation ditches to set and check their crayfish nets in the Groene Hart, or “Green Heart,” of the Netherlands, the thinly populated pastures and peatlands among some of the country’s largest cities. They get paid per kilogram caught—the expedition I joined yielded about 150 kilograms, or 330 pounds, of live crayfish—but this is really a passion project for them. CrawFish Farm Holland is approved to sell live and prepared catch directly to restaurants and other wholesalers, and hopes to sell its crayfish in supermarkets in the next two years. “We’re not cultivating them,” Van der Sanden told me. “We’re fighting them and eating them.”

In America, Van der Sanden pointed out, Louisiana was able to quickly ramp up crayfish production and demand in the 1980s. But the scientists I spoke with were skeptical of this strategy. “You will never eat yourself out of this crisis,” Roessink said. And if people like crayfish too much, Lemmers said, they might start cultivating crayfish in their backyard or trapping them without oversight, which could promote further spread. Even if more Dutch people come around to the pleasures of a messy boil, the crayfish still could come out ahead.

The image originally accompanying this article did not show a red swamp crayfish. That image has been replaced.

The Growing Lawlessness in California’s Forests

2026-09-01T07:00:00-04:00

This article was featured in the One Story to Read Today newsletter. Sign up for it here.

On the day of the kidnapping, the campground at Gumboot Lake happened to be nearly empty, so Joseph Charles Henrichsen was one of the only people to see the two Forest Service biologists pull up in their pickup truck. Federal employees didn’t visit all that much—campers who’d been coming to the small alpine lake for years told me they’d never seen a ranger on patrol in these Northern California woods. But some of the regulars had noticed the “frog people,” as one called them, quietly collecting their specimens the previous summer. These particular biologists, both women, were there to survey the Cascades frog, whose well-being can help signal the health of the overall ecosystem.

People at Gumboot Lake tended to give one another space, but that July morning, Henrichsen approached the biologists as they worked. He was holding, authorities would later allege, an AR-15-style assault rifle, and told the biologists to turn around and put their hands in the air. He zip-tied their wrists behind their backs and ordered them to walk to his camping trailer, parked at a site strewn with crates and refuse, the clothesline heavy with towels. His 23-year-old son, Phoenix, made room for the biologists inside.

What Henrichsen’s intention was that day—what he thought he might achieve by abducting two government scientists, whether he had a clear motivation at all beyond an impulse of self-righteousness and a loaded gun—remains uncertain. Before long, he escorted one of the biologists back to their Forest Service truck. He drove several miles with her down a winding road along the Sacramento River. Glimpses of Mount Shasta, the volcano looming to the east, reared up through the trees. Along the way, he made calls, according to the FBI. He spoke with a Forest Service ranger and left a message with another person who has not been identified. “I’ve taken two Fed biologists hostage from the Forest Service,” he said, according to authorities. “What I’d like to have happen, I’d like to have you send me a fucking Fed investigator, with a goddamn badge, preferably from the FBI. And we can talk things through a little bit.”

Over the next few hours, Gumboot Lake and the surrounding Shasta-Trinity National Forest would swarm with law enforcement in helicopters and Bearcat armored vehicles. Local SWAT teams and snipers, then more than 50 FBI officers, poured into the national forest. A Boeing 757 flew across the country, from Virginia, with an elite FBI hostage-rescue team.

All of these officers were confronting a man who had a history of erratic behavior: Henrichsen had been released from a Washington State psychiatric hospital a few years before, after allegedly tormenting a neighbor by firing a gun, throwing fireworks on his porch, and placing zip-tie handcuffs on his property. But Henrichsen had also managed to stay at Gumboot Lake for months—well past the official seven-day limit—apparently without triggering any official response from the Forest Service. Many of the campsite’s regulars had crossed paths with the Henrichsens over the summer, and although Charles held himself apart, he struck them as a familiar type, not so out of place here among the free spirits and drifters living in their cars. That he’d turned out to be unhinged or anti-government enough to kidnap a couple of federal frog biologists didn’t seem to shock them. In fact, with the way things are going on the woollier fringes of America’s public lands, maybe something like this was bound to happen.

Siskiyou County, the area that includes Gumboot Lake, has a reputation for fierce independence and the hostility toward government meddling that often comes with it. The county leans conservative, though Mount Shasta is a magnet for a crunchier type: crystal enthusiasts, New Age spiritualists, pilgrims on the road to Burning Man. Hikers stream down the Pacific Crest Trail, which runs along a ridge above the lake. On Sundays, some of the campers at Gumboot Lake like to gather for a drum circle, and when the full moon arrives, people convene to celebrate.

The Gumboot Lake campground is free and has no assigned spots, no running water, limited cell signal, and a lone pit toilet. It’s one of dozens of alpine lakes in the Trinity Divide accessible by car or off-road vehicle. Others are more popular. The Forest Service described the campground in one write-up as a “typical meadow type lake, not very beautiful.” But it’s relatively easy to get to, just a 40-minute drive up from Interstate 5.

I made that drive last month, not long after the Forest Service had reopened Gumboot Lake, less than a week after the kidnapping. I saw old vans and school buses parked along the Forest Service road leading to the campsite; people had strung up hammocks between the trees. At the lake, several campers had already set up their tents and parked their mobile homes. One camper, a 24-year-old backpacker from Australia, told me the campsite was a place to “drop in with the land, drop in with each other, cook meals with each other, and just be super present.” As we talked, a large shirtless man, whom she called Zephyr, appeared and handed her a slice of papaya.

The Forest Service is supposed to be a steward of its public lands, preserving them while still allowing people to enjoy the woods and use their resources. Dispersed camping—outside of established campgrounds—is allowed in most national forests. And although people here are supposed to move on after seven nights, locals say that plenty of people live out of their cars for much longer. Informal settlements are widespread on the national-forest lands of Oregon and Northern California. People tend to cluster along the river corridors, most of them peaceable campers. “But the last 10 percent of them are pretty sad,” says Gabriel Howe, the executive director of the Siskiyou Mountain Club, a group that maintains trails and campgrounds in the region. “People living out there, people manufacturing drugs and all sorts of nasty stuff.”

Policing a remote forest is unlike any other job in law enforcement. The Shasta-Trinity spans more than 3,000 square miles, and the Forest Service is perennially under-resourced. Carmen Kinch, a Forest Service law officer who patrolled the area around Gumboot Lake in the late 1990s and 2000s, told me that back then, seven law-enforcement officers worked across the entire forest. Today, there are four. Not one, she said, is currently based in the Forest Service management unit that includes Gumboot Lake.

The drop in personnel is not just a problem in this one forest. Since last year, the Forest Service’s staffing across the country has fallen by about 20 percent as a result of firings and departures encouraged by the Trump administration, part of its push to reduce the federal workforce. In the Shasta-Trinity forest, excluding firefighting roles, the permanent staff—employees who maintain campsites and conduct wildlife research—has been reduced by about a quarter, one former U.S. official, speaking on condition of anonymity because of the sensitivity of the topic, told me. (Bob Doucette, a Forest Service spokesperson, told me that staffing levels are “similar” to what they’ve been in the past few years and said that “all employees continue to meet public needs.”) Howe has worked in the region for 15 years, and he sees the Forest Service as a slowly dying patient. “Around COVID, it went into palliative care, and now it’s on hospice,” he told me.

A sense of lawlessness seems to be growing on the public lands in the region. People on Howe’s staff have returned to find their cars stripped at trailheads—no wheels, no batteries, gas siphoned. They’ve encountered mountain roads blockaded by hostile civilians. Michael Kauffmann, an ecologist and the executive director of the Bigfoot Trail Alliance, told me that he’s recently instituted more training for his researchers and trail crews on safety measures and defusing confrontations. More than anything, Kauffmann has noticed a greater disregard for the rules—more graffiti, a lot more garbage, and less appreciation for these lands as places to care for rather than to joyride over with guns and four-wheelers. One person who regularly hikes in the area around Gumboot Lake told me that he’d felt safer during “close encounters with bears and lions” than with some of the armed and unfriendly people around.

The Henrichsens had apparently lived at Gumboot Lake since at least early May, when Todd Heie, another camper who visits the lake regularly, first encountered them. In early June, Joseph Charles Henrichsen, who goes by his middle name, asked a neighbor at the campground, Kevin Airoldi, for a lift into town. On the 45-minute drive, Charles said that he and his son were “just traveling around, trying to do it on the cheap.”

Charles mentioned that they’d come down from the Pacific Northwest but not what he’d left behind. He’d once been a successful mortgage-loan broker in Orange County, California. But he had run into business trouble, and his relationship with Phoenix’s mother had fallen apart. By 2022, he was divorced and living near Bellingham, Washington, up by the Canadian border. That year, he spent six months in a county jail, charged with a hate crime for allegedly harassing his neighbor. Police said that he “appeared to be spiraling in his mental health,” according to court documents. Henrichsen, who kept multiple guns at his home, would rant about world events, including the war in Ukraine; an evaluation found that he likely had schizophrenia or another psychotic disorder. Although a judge ordered him to be admitted to a hospital for psychiatric treatment, bureaucratic delays kept him out for months. Even then, he was admitted only briefly, after the charges had been dropped.

At Gumboot Lake, both Charles and Phoenix Henrichsen tended to keep to themselves in their Feather Ultra Lite trailer, parked amid the lodgepole pines. The other campers would watch them pacing around the lake, lap after lap, as the father appeared to loudly lecture, or preach, to the son, who kept his head down and rarely said a word. Many found Charles intense and standoffish. He would burn red sage during the day, and at night he would play a type of chanting Indian music that Heie, a former director of the county’s department of corrections, sometimes listened to as he drifted in his kayak, looking up at the sky. At one point, Aaron Sullivan, a software engineer who was working remotely out of a Winnebago along the lake, stopped by the Henrichsens’ campsite. He’d been impressed by Charles’s strong, lean frame and “big, beautiful beard.” Sullivan invited them to one of the full-moon gatherings. We won’t be doing that, Charles told him.

The Thursday morning of the kidnapping, on July 16, Charles Henrichsen must have been watching the biologists before he approached them—it’s unclear how long. After he had captured the women and herded one of them into the pickup truck, he kept her tightly restrained. He told the unidentified recipient of the message he left while driving down the mountain that “her hands are behind her fucking back, and I’ve got live rounds ready to fucking [inaudible] anyone who fucks with me,” according to the criminal complaint.

When Siskiyou County Sheriff Jeremiah LaRue was alerted to the situation that morning, he thought the man could be swatting, he told me—attempting to trick his office into scrambling to respond. His county is larger than Connecticut, and more than half of it is public land. Sheriff’s deputies soon arrived at Gumboot Lake to find the campsite gate closed and a gray Subaru blocking the road. (The Subaru, police would later learn, was stolen, possibly by the Henrichsens, and likely parked there to block others from entering during the kidnapping.) The sheriff’s deputies flew a drone overhead and quickly spotted the Forest Service truck that Henrichsen had commandeered, LaRue said. A man stepped out of the trailer, carrying a rifle, and later the deputies saw him with someone whose hands were bound. “He’s now pointing a rifle at the drone,” a deputy said over the radio. LaRue knew he had to get more people up there.

Police set up in a dirt roundabout on the north side of the lake, within shouting distance of Henrichsen’s trailer, where he had holed back up with Phoenix and both biologists; they blocked access to the campground, and law enforcement streamed in from California and elsewhere, their cars crowding the narrow Forest Service road. Hank Magnuski, a retired tech executive who owned a nearby cabin, came upon the pileup of armored vehicles and heavily armed agents in the early afternoon, at the end of a long drive from his home in Palo Alto. He was startled by the military-style occupation of this out-of-the-way wilderness, but he gave two SWAT-team members a ride up the road. “What’s the occasion?” he asked one of them.

At about 3:30 p.m, more than four hours after his initial calls, Henrichsen’s voice came over law-enforcement radio. He had taken one of the biologists back to their truck, where she helped him access the radio to contact the Forest Service’s dispatcher and the FBI. He seemed calm; he repeated his request to speak with FBI officials, “so I can release these hostages. Over.”

The reception was choppy enough that he started negotiating with the FBI agent Russell Quiniola about how they might speak in person. “Are you comfortable putting the rifle down and coming to a point where you feel safe?” Quiniola asked.

“Listen, you come and talk to me,” Henrichsen responded. “I would suggest coming unarmed, because if you have one, that might trigger something in me, you never know. Or you may have a rifle, and you see me, that might trigger something in you. But I suggest we get the show on the road. These ladies would like to go home. Over.”

The fact that Henrichsen had initiated contact with authorities by requesting to talk with the FBI was, to some degree, encouraging. It suggested that his intentions went beyond simply harming the biologists, Gary Noesner, a former chief of the FBI’s crisis-negotiation unit, told me. In such situations, every hour that negotiators can keep the kidnapper talking is one without violence. Unless they sense that the risk is increasing, Noesner said, “we’ll talk ’til the cows come home.”

Quiniola and other FBI agents spoke with Henrichsen over the next nine hours. By 1 a.m., the standoff had gone on so long with such little progress that LaRue was surprised when someone in his vehicle said, “One of the females is coming out,” and the first biologist opened the door of the trailer and walked toward the police. Her colleague followed six minutes later. Another hour and a half would pass, during which FBI agents exploded flash-bangs around the trailer—a commotion heard by hikers on the Pacific Crest Trail above the lake—before Phoenix and then Charles finally turned themselves over and the government vehicles lining the road to Gumboot Lake started heading back down the mountain.

By the time I arrived at the lake, almost all signs that this place had been the site of such force were gone. The Henrichsens’ trailer had been towed away; their debris had been bagged and left in the outhouse, along with a note that read, “Trash from the kidnappers campsite.” One camper had held on to three wooden billy clubs found amid the mess. I set up my tent at a campsite next to the Henrichsens’, where I could see the lake through the pines. Everything was serene. The day was cloudless, and the sun sparkled on the water.

Charles and Phoenix Henrichsen eventually pleaded not guilty in federal court in Sacramento to charges of kidnapping federal employees and other felony counts that carry the prospect of life in prison. Charles’s lead defense attorney, Mia Crager, told me that her client had intended no harm to the biologists; according to police, the two women, who have not been named publicly, did not appear to suffer major injuries during their 15 hours as hostages. “What happened on July 16 was preventable and resulted from system failures on a number of levels,” Crager said in a statement. She did not elaborate on the type of failures she was referring to, and didn’t respond to questions about any specific allegations against Henrichsen. She said that Charles “hopes that some good will come of the case, including that his son Phoenix will get the mental health care he sorely needs.”

Phoenix’s lawyer, Kresta Daly, told me that Phoenix had been an unwilling participant in the standoff: He had no gun, didn’t know what was going to happen, and, once the crisis had started, he hoped that it would resolve peacefully. His main thought, Daly said, was “How do I stop anyone from dying today, me included?” His father, Daly told me, had hurt him before; he’d been in the hospital earlier this summer after Charles beat him up. Charles Henrichsen, through his lawyers, did not respond to questions about that accusation, but agreed that his son “was not at all at fault for what happened that day, nor was he an active participant.”

Henrichsen’s specific demands during the hostage crisis have not yet been fully explained. He did not leave a manifesto. An FBI spokesperson declined to describe those hours of discussions with Henrichsen, citing the ongoing investigation. Sheriff LaRue, who had been inside a Bearcat with the FBI negotiators during part of the process, said he couldn’t hear Henrichsen “discussing anything as to why he was doing this, what his motives were.” At least in that moment, he seemed to want the attention of the federal government. He got it.

The day of the biologists’ release, the chief of the Forest Service, Tom Schultz, wrote an email to agency staff expressing his gratitude to law enforcement and his support for the two women after their ordeal. A safety manager with the agency followed up the next week, with links to security tips and videos with titles such as “Avoiding Trouble” and “Building Trust in Small Towns.” Despite such internal gestures, some former Forest Service leaders heard little public support for the staff from an administration that has shed civil servants without much remorse. “There’s not an ounce of empathy for federal employees,” Vicki Christiansen, a former chief of the Forest Service during the first Trump administration and the Biden administration, told me. “Not an ounce.”

As I talked with people at the lake, the hostage episode began to seem inseparable from the deepening neglect of our country’s public lands. Whatever the Henrichsens’ personal motives may have been, the erosion of day-to-day management has made ruptures in the fragile relations among strangers in the woods all the more likely. “They shouldn’t be letting someone stay here for months on end,” Heie, the former corrections officer, told me. In the early evening, I watched him paddle his kayak to the middle of the lake. Another man stood in the shallows, casting a fishing line. Through the trees, I heard a couple laughing. I kept looking around at the peaceful scene, unable to shake the unsettling feeling that something was off. As darkness settled, I sat in my camping chair, listening for the sounds of the frogs.

This El Niño Will Be Unlike Any in the Past Eight Decades

2026-08-30T07:00:00-04:00

In the middle of the Pacific Ocean, along the equator, scientists have drawn a box. It’s a big box, two-thirds the size of China, yet hardly anyone lives inside it. This rectangular swath of open ocean is monitored by satellites and a network of moored buoys that measure minute changes in the temperature of the water, with an eye toward anomalies. Every few years or so, driven by changing winds and currents, the water temperature in this part of the ocean starts to rise; even a half degree Celsius above normal over months warrants a designation that an El Niño phase has begun. That’s enough warming to unleash floods and fires, drought and famine, around the world. The box is called Niño 3.4. Right now, it’s hot in the box.

“Like a baby pool,” Kristina Tietjen, who spent most of August scuba diving in the middle of this mass of abnormally warm water, told me. Tietjen has been coming to the world’s largest atoll, Kiritimati, for more than a decade to study its coral reefs. She usually dives in a medium-weight wetsuit; this trip, her team needed only rash guards. It is still early in this El Niño cycle—which is likely to be the strongest ever recorded—but the waters around Kiritimati have already been 3 to 4 degrees Celsius above normal.

Tietjen’s time on Kiritimati (pronounced like Christmas) has allowed her to watch, in brutal detail, what these repeated warm-water cycles, playing out in a steadily warming ocean, mean for the great tropical coral reefs here, which were once near pristine. During her first El Niño, in 2015–16, 90 percent of Kiritimati’s coral reef died. Years of slow, partial recovery followed. This month, Tietjen and her colleagues, part of the Canadian scientist Julia Baum’s lab at the University of Victoria, discovered that nearly half of the island’s remaining corals had already bleached—a prelude, if warming persists, to death. When Tietjen comes back, she may well be swimming over a graveyard.

Most of the world experiences El Niño’s ripple effects as the heat building up in the equatorial Pacific radiates into the atmosphere and around the globe. This means great gouts of rain in supercharged storms in some places; in other areas, crop-killing dry spells that can extend for months. Along the Pacific Coast of South America, torrential rains and landslides have forced the temporary closure of a main route to Machu Picchu, in Peru; deadly floods have swept through arid regions of northern Chile. Countries such as India and Ethiopia are enduring some of their driest summers on record. El Niños tend to mean more rainfall across the southern United States and drier stretches across a band in the North. Strong El Niños can mean record-breaking heat, and this one is shaping up to be unlike any cataloged in the past eight decades. A 2-degree-Celsius increase in the Niño 3.4 box typically earns the title of “Super El Niño”; we’re already past that threshold, and experts project that the warming could peak at roughly twice as high. “We are now pretty certain this will be an exceptional event,” Malte Stuecker, the director of the International Pacific Research Center at the University of Hawaii at Manoa, told me.

Part of the reason Stuecker expects this warmth to keep building is that the buoys along the equator can detect water temperatures well below the surface. Down more than 100 meters, an enormous mass of warm water has built up and will rise to the surface in the months ahead. The amount of heat brewing in the ocean has produced one of those graphs in which the current year departs drastically from past trend lines, wriggling up erratically into unfamiliar territory.

Although scientists continue to debate the extent to which climate change makes El Niño cycles more frequent or intense, they agree that these cycles are occurring in an ocean that is warmer than it was. Last Saturday, the daily global average sea-surface temperature reached 21.1 degrees Celsius (70 degrees Fahrenheit), its highest level on record, according to Europe’s Copernicus Climate Change Service. Even if El Niño cycles are not changing at all, as they spiral around that ascending baseline, the temperature peaks they hit keep getting higher, causing repeated blows to marine life and disrupting ecosystems in ways that may be irrevocable.

Carlos Zavalaga, the director of the seabird lab at the Universidad Cientifica del Sur, in Lima, began to comprehend the magnitude of this year’s warming in mid-June, when he visited Macabi and Guanape Norte, two islands off the northern coast of Peru. He’d been hearing from colleagues about dead birds washing up by the hundreds on the coasts of Peru and Ecuador. He’s seen the news stories about flocks of hungry Peruvian pelicans swarming the fish markets in port cities looking for food. On the two islands, which just last year were thronged with pelicans, Inca terns, and Humboldt penguins, Zavalaga found empty nests, some with eggs and starving chicks. “They were abandoned islands,” he told me.

These seabirds feed on schools of anchovies swept up in the Humboldt current, the cold, nutrient-rich upwelling waters that create bountiful fisheries off South America’s Pacific Coast. During El Niño years, that current gets displaced by warmer waters pushing east across the ocean, and these great schools of feeder fish can vanish from the coasts. The decline has been severe enough that Peru this month shut down an anchovy fishery that is a key economic engine for the country, with much of the year’s quota unfilled. Long before satellite monitoring, Peruvian fishermen noticed these cycles of warmer waters and dwindling fish stocks; they named the phenomenon, which peaked around Christmas, after the Christ child. References to El Niño in Peru’s scientific literature date back to the 19th century.

In his three decades studying seabirds, Zavalaga has witnessed die-offs before. But it is the rapid-fire nature of these blows and the severity of the disruptions that make him so concerned. Just three years ago—the same year avian flu killed more than 500,000 seabirds—another El Niño slammed into the coast of Peru. The birds that had made a tenuous recovery since then now face the prospect of another die-off. “It’s like a boxing match: You have a first knockout, you try to recover quickly, and then you have a second knockout,” he said. “I cannot imagine what will be the scenario by the end of this year.” He doubted that the birds could go extinct, but some species might move farther south in search of colder waters. The coastline in Peru and the culture that depends on the fish and the birds could be permanently altered.

When the warming started around Kiritimati, one of the few populated islands in the Niño 3.4 monitoring zone, Bill Newton, an island resident, was fascinated by how evident the changes were. He told me that the normally dry island turned sopping wet, and regular floods started washing out the roads. The wet brought swarms of mosquitoes (prior El Niños have led to outbreaks of dengue and chikungunya) but also clouds of dragonflies, which he hadn’t seen before. Higher-than-normal tides poured into seafront homes. Tuna, which can do well in warmer water, attracted dozens of fishing boats off the coast. Prevailing winds changed direction.

Corals have a particularly hard time when water warms; they can’t move elsewhere, and they are sensitive to heat stress. As little as four weeks in water just a degree warmer than normal can cause bleaching, the loss of the symbiont algae that supply corals’ food and brilliant colors, according to Baum, whose lab Tietjen works at. Baum, an ecologist, has studied Kiritimati’s reef since 2009. After the El Niño a decade ago killed off nearly the entire reef, Baum began to consider the possibility that she would witness the obliteration of an ecosystem in her lifetime. In a 2018 paper in Science, she projected that successive bleaching events could occur frequently enough that coral reefs, which take roughly 10 to 30 years to recover from mass mortality, would not have time to regrow. In the years since, she had been hopeful that nature could fight back as she watched surviving corals take up symbionts that were more heat resistant. This year’s warming and the pace of bleaching already under way have dispelled that hope.

What happens next is in some ways predictable. El Niño is not a haphazard force of destruction; it’s a warming phase in a specific area in the Pacific Ocean that scientists have determined shifts the probabilities for certain weather outcomes in various parts of the world. Those probabilities are reliable enough that, for example, the U.S. Navy already plans to send a hospital ship to the coast of Peru in February to be in position to help after expected floods. But as Baum and other scientists who study and love the delicate ecosystems in the equatorial Pacific are finding, just because El Niño is a familiar cycle doesn’t mean things will be the same when this one is over.

A New Way of Seeing Alien Worlds

2026-08-29T08:00:00-04:00

Tomorrow morning, if all goes to plan, NASA’s Nancy Grace Roman Space Telescope, its latest and greatest orbital observatory, will lift off from its launchpad and add a little rocket-booster light to the Florida sunshine. A lot is riding on this mission. If the Roman is successful, it could give humanity a new way to see the planets around other stars and a fresh understanding of the largest-scale objects that we know to exist. But if it fails, NASA’s science program, which is already fending off ruinous budget cuts, will need to seriously rethink its future. The Artemis mission to the moon was a big win for human spaceflight, and now the science side of the house badly needs one.

NASA’s engineers have spent more than a decade readying the Roman for this moment. Before the telescope arrived in Florida, it spent more than a year in NASA’s largest clean room at the Goddard Space Flight Center in Greenbelt, Maryland. To get near it, you had to be mummified alive. When I visited in December, NASA staffers led me to a sterile, fluorescent-lit locker room. They tossed me a white jumpsuit and two sets of booties with tight-fitting straps. They told me to put on a beard net and hair net, and triple-wrapped the gap between my jumpsuit sleeves and gloves with blue tape.

The astronomers didn’t want my skin exposed in the clean room. Dry flakes of it could drift off and settle onto the telescope’s ultrasmooth primary mirror. “You’re going to be the dirtiest thing in here,” one of them told me, as we walked through a little chamber where more than 30 nozzles blasted us with air. I jotted down this remark on special mint-green paper that sheds fewer particulates than a regular notebook, and then stepped into a cavernous eight-story white space soundtracked by the hum of six school-bus-size fans. At its center, the Roman stood more than 40 feet high, and it was surrounded by more than a dozen people, all in identical jumpsuits. Half were standing in clusters, talking, while the other half were up on motorized lifts, fussing with the telescope, applying duct tape and making final adjustments. More than a thousand people have worked on this project, and those who are lucky enough to get close to the hardware seem to derive a physical pleasure from the experience.

[Read: Inside Donald Trump’s attack on NASA’s science missions]

The Roman has a pretty shady backstory for a scientific instrument. Its primary mirror was part of a surprise gift bestowed on NASA by the National Reconnaissance Office. The NRO is one of America’s most secretive government agencies, and it apparently had a Hubble-class spare just lying around. A cash-strapped agency can’t afford to be too choosy about provenance: NASA took this mirror that was supposed to point down at the ground, for surveillance, and will instead point it out at the stars.

Coronagraph delivery and unboxing (Chris Gunn / NASA)

The NRO’s spare mirror has also solved a bit of internal politics for NASA’s science program. The Roman is built to monitor the universe at the very largest scales, to map the cosmic superstructure of galaxy clusters that stretches out from us in every direction, to see how it has expanded over time. These observations might lead scientists to a more refined sense for gravity, and may even help them discover new physics. But the idea for a mission like this was originally conceived of around the early 2000s, shortly after the discovery of dark energy, and a survey of the large-scale universe seemed like the obvious next thing to do. When NASA was finally ready to build the telescope, in the early 2010s, the action in astronomy had shifted closer to home. Scientists were in the midst of discovering thousands of new planets around nearby stars, and they wanted to find more. Although the Roman’s original mission plan did allow for some planet-hunting, the new mirror would allow them to do it in a whole new way.

[Read: The Nobel Prize winner who thinks we have the universe all wrong]

Finding a little world with its own atmosphere in the glare of a nearby star is tricky work. Even the very largest planets will be a billion times fainter than the suns they orbit. When I spoke with Dominic Benford, Roman’s program scientist at NASA Headquarters, he likened the challenge to spotting a firefly buzzing around a powerful searchlight from thousands of miles away. To make this possible, astronomers have lodged an exquisite new kind of coronagraph into the Roman’s metallic guts. Coronagraphs originally helped astronomers see the plasma flares emanating from our sun, by blocking out the far more brilliant light from the solar disk. In a sense, the device creates a brief artificial eclipse inside the telescope. The Roman’s coronagraph is designed to do something similar for more distant stars so that astronomers can see the planets that surround them.

When the telescope is trained on a star in our Milky Way, it will catch a lot of light, and then run it through the coronagraph’s system of barriers and mirrors. Two of the mirrors that it uses may be even more beautiful in their design than the ultrasmooth, eight-foot primary. Each is smaller than the average makeup compact, and has more than 1,600 actuators that can individually deform tiny portions of the glass in tiny increments, less than the diameter of a helium atom. Astronomers will employ hundreds of these warping movements, spread across the mirrors, to deflect a specific star’s glare, creating an image of the star within a ring of nearly pure darkness. Deformable mirrors like these have previously been used on the ground to help telescopes see through the warble and blur of Earth’s atmosphere. But no one has ever sent one of this complexity to space.

The Roman team believes that up there, where the viewing conditions are more pristine, the Roman will be able to deliver the first direct images of distant planets in visible light. Until now, most of the extrasolar planets that scientists have observed have been detected indirectly. In some cases, a planet’s existence is inferred from the gravitational force that it exerts on its host star, causing the star to wobble. Or astronomers might see a host star dim, repeatedly and predictably, as an orbiting planet passes in front of it. We have been able to get direct snapshots of only a few young planets that were still roiling hot and glowing brightly, and mainly in infrared. The Roman telescope’s coronagraph should be able to see worlds that aren’t glowing at all. Like Earth, they will be illuminated solely by the soft visible light of their host stars.

[Read: When a telescope is a national-security risk]

I asked Vanessa Bailey, the coronagraph’s instrument technologist, if there were any particular stars that she was excited to inspect. Because the coronagraph isn’t part of the telescope’s main mission, just three of the first 18 months will be set aside for its observations. Bailey said that the astronomers will use that time to look at previously detected planets that are only tens of light-years away. She mentioned the Epsilon Eridani system, where a Jupiter-size world is known to orbit an orange dwarf star. With the coronagraph, they’ll stare at it for many hours, letting photons from the planet slowly pile up. “It’s just going to be a little fuzzy smudge, just barely there,” Bailey said. “But we might be able to tell if it’s cloudy or clear, and we might be able to say whether it’s more like Jupiter or Neptune.”

All of this could go wrong. Bailey told me that the coronagraph has never been tested outside the lab, and in space, it may not work. If for some reason the mirrors are misaligned, or the instrument malfunctions, any planets would be lost in the astral glare. NASA’s long-running quest to find another living world in the universe would be set back, perhaps by a decade or more. But if the coronagraph does fulfill its mission, it will serve as a proof of concept for NASA’s next really big telescope, the Habitable Worlds Observatory, which the agency has been developing for more than a decade.

Roman’s coronagraph will be able to find only large planets—Saturn-size or ideally bigger—that circle their stars in chilly, distant orbits. NASA scientists want to place a much more precise one on the HWO. That mission will fixate on more than 100 nearby stars that share the sun’s rough age and brightness. Its coronagraph will cast a much tighter ring of near darkness around them, to see if any planets show up in the closer, warmer orbits that might be more hospitable to life—and it will even look for signs of life in the planets’ atmospheres. But the HWO won’t lift off until the late 2030s at the earliest. For now, the Roman has to prove out. It has to survive the violence of launch and fly a million miles from Earth so that its mirrors can catch the light of a nearby star and then surround it in darkness so deep that a faint alien world appears.

America’s Rural Lands Are Burning

2026-08-18T09:22:16-04:00

The summer of 2026 has been one of extraordinary fires. In Europe, they have burned hundreds of thousands of acres in France and Spain and destroyed homes across England; they killed 13 people in Spain last month, and an elderly couple in Greece this weekend. In Canada, fires have consumed an expanse of forestland larger than Belgium—which is also battling its biggest recorded wildfire. In the United States, about 7.2 million acres have burned, the largest year-to-date area in the past decade and nearly double what it was at this time last year, according to data from the National Interagency Coordination Center. Allison Rayburn, a program coordinator for the Oregon Department of Forestry, told me that in the eastern part of the state, “you can barely see the sun” through the smoke: “It’s just devastation, and miles and miles and miles of blackened landscapes.”

In Europe, many of the fires have threatened populated areas, but in the U.S., much of the area burning in the West has been rural. Near the Oregon town of Jordan Valley, the Big Grass Fire has been burning since July 23 and is among the largest fires in the state’s modern history; toward the end of last month, it hopped a canyon and crossed the border to Idaho, where the smoke from this and other fires made the air quality the worst in the country. The complex of fires in Spokane, Washington, flattened entire neighborhoods; three other wildfires 10 times its size are burning elsewhere in the state.

If 2025 was a year of urban fire that destroyed large swaths of Los Angeles, 2026 is a year of rural fire, Daniel Swain, a climate scientist at the University of California Agriculture and Natural Resources’ research unit, told me. “This is an extraordinary fire season in the American West,” Swain said. “It just hasn’t been in California”—which usually accounts for much of the nation’s fire activity and its media attention.

Many of the biggest fires this year started because of back-to-back lightning strikes and have grown because of extreme heat, drought, and dry fuel left behind after an unusually warm winter. They’ve destroyed grazing land and crops, killed livestock, and incinerated fruit that was just shy of harvest. Many of the ranchers and government employees I spoke with compared 2026 to Oregon’s devastating fire season two years ago, when 1.9 million acres burned across the state; already this year, residents have lost hundreds of thousands more acres than that. When I spoke with Dave Upthegrove, Washington’s commissioner of public lands, he was driving from the fires in Spokane to one near Walla Walla that was threatening the town’s water supply; the week before, he’d visited the burn area of the 150,000-acre Sinlahekin Fire. Firefighters there had used a road as a fire line, he said, and on one side of it, all was flat and black. On the other side, crops were wilting from the fire’s heat.

Because so many of the fires this year are burning on rugged terrain without many roads, the elite firefighters who rappel from helicopters and hike through miles of forest to reach a fire line have been in high demand, Upthegrove said. But these hotshot crews cannot hit every fire, and in many cases, ranchers are the first to get there. Silas Skinner, a sixth-generation cattle rancher, told me that when the Big Grass Fire broke out, three government trucks were there; the rest were rigs from the Jordan Valley Rangeland Fire Protection Association. RFPAs—about 40 of which are spread across Oregon, Idaho, and Nevada—are independent, volunteer firefighting nonprofits composed of local ranchers and farmers who are trained by state and federal agencies to help fill the firefighting gap in rural areas. Their equipment comes largely from military surplus—literally, in some cases, bulldozers that didn’t get sent to Vietnam, “and some that did,” Rayburn, who is one of the RFPAs’ main contacts in the Oregon government, said. In the weeks since lightning struck and started the fire, Skinner and about 100 other volunteers have been using military trucks, outfitted with water tanks and sprayers, to help government agencies fight the Big Grass Fire. It is currently burning an expanse of sagebrush and juniper trees nearly the size of Luxembourg.

Federal agencies have also been sending in thousands of employees to fight the fires. But with so many fires burning simultaneously, government resources have been stretched thin and diverted to those in more populated places such as Spokane. The volunteer firefighters, then, have taken on more of a share of the work than they have in the past, they told me. (Both the Forest Service and Wildland Fire Service said that it’s standard for RFPAs to take on more work during a fire season this active.) When I spoke with Morgan Kromm, a rancher and the secretary of the RFPA in Crane, Oregon, and Tom Doman, the association’s chairman, they’d just come off a 10-day stretch of back-to-back firefighting. They told me they’d been working in 36-hour shifts, and when I asked where they’d catch naps during those stretches—in their trucks?—Doman clarified: “We were not sleeping.” The Crane RFPA has already handled 18 fires this year—normally the total they’d have seen by about October—and “we still have 90 more days of fire season,” Doman said.

Fire seasons vary by region, but in general, they have been getting longer and less predictable. Both RFPAs and state officials told me that this has made it difficult to get resources when and where they’re needed. Typically, Oregon enters into mutual-aid agreements with states in the Southwest and along the East Coast that have different fire seasons and can help one another out. But the whole country has been so dry that “they’re busy with their own fires,” Michael Curran, the chief of fire protection for the Oregon Department of Forestry, told me. RFPAs provide mutual aid to one another too, while members also try to manage their own ranches and farms alongside firefighting. “Yesterday, we were moving cows, and the fire was burning right above us,” Skinner said. “So everybody kind of had an eyeball on it.”

Some of his grazing land has already burned. Other ranchers have found that their cattle died in the fire. In Jordan Valley, community members have been donating hay to those whose land burned, so their cattle will have something to eat. Rayburn told me that since that terrible 2024 fire season, she has watched landowners “for two years rebuilding fences and reestablishing water holes and reestablishing their cattle herds.” Now the fires are burning those same places—land where families have lived for generations, and land upon which their livelihoods still depend.

Nobody Knows What Ultra-Processed Food Is

2026-08-12T13:27:27-04:00

You might know a tomato when you see one, but for the federal government, identifying one is harder. Is a tomato a fruit or a vegetable? In 1893, this question went all the way to the Supreme Court, which decided that, for tax purposes, at least, a tomato was a vegetable. What about milk? For a bureaucrat setting standards, a straightforward explanation such as “liquid from a cow’s udders” won’t quite cut it; according to the FDA, milk is “the lacteal secretion, practically free from colostrum, obtained by the complete milking of one or more healthy cows.”

Even for the government, however, tomatoes and milk are relatively simple to define compared with ultra-processed foods. Although many researchers agree that these creations are negatively affecting people’s health, they cannot agree on what foods do and don’t belong in that category. That’s not keeping Robert F. Kennedy Jr. from trying: After months of promising the federal government’s first formal definition of ultra-processed foods, the secretary of Health and Human Services announced on Monday that the department is finalizing one.

He stopped short, though, of sharing the exact text in either the public announcement or at a “food policy celebration” held Monday at HHS’s Washington headquarters; the definition is still being reviewed by the White House. This effort may have real upsides. Having a regulatory definition of ultra-processed foods (or UPF, as they’re often referred to) could at the very least advance inquiries into the harms of these products. But the effort also raises the question: Can the government adequately define a problem before researchers know its true cause?

So far, scientific efforts to define UPF have revealed only how fuzzy the category is. The most commonly used definition, which was developed by Brazilian researchers, takes the most expansive approach: It lumps all foods with industrial ingredients, such as sweeteners and emulsifiers, into the ultra-processed category. A classification system created by academics at the University of North Carolina at Chapel Hill, meanwhile, defines “highly processed foods” as products “processed to the extent that they are no longer recognizable as their original plant/animal source.” Yet another well-respected system, this one designed by a French analytics company, attempts to incorporate additional factors, such as a food’s salt, sugar, and fat content, as well as the “function, number, and potential health risk of additives.”

These rubrics are distinct enough from one another that flavored yogurt is an ultra-processed food under the Brazilian system and a moderately processed food under the UNC system. Under the French system, the classification depends entirely on the yogurt itself (even though a paper from the group found that 95 percent of flavored yogurts in French grocery stores were indeed ultra-processed).

Researchers disagree on how to define these foods, because the science around how and why they may be harmful is ambiguous. Epidemiological studies have found that UPF-heavy diets have consistently been associated with a range of maladies including cardiovascular disease and type 2 diabetes. But although researchers suspect that certain additives are causing dietary issues, the number of industrial ingredients or additives in Americans’ food is so large that many more may also be causing trouble. Nor do researchers know what processing steps cause everyday food ingredients to go from healthful to harmful. That means Kennedy and his department are essentially trying to define a food group that no one really understands.

Kennedy is under no illusions about how little we know about the American food supply. He has frequently criticized a decades-old loophole that allows food companies to put new additives in their food without even telling the FDA. On Monday, he announced a new proposal to fix this issue; if finalized, the rule would require companies to let the FDA know when they’re introducing a new additive. But even that won’t fully solve America’s additive problem.

Researchers at the Pew Health Group estimated in 2011 that America’s food supply already features more than 10,000 additives. The new rule would require manufacturers that quietly introduced such compounds to tell the FDA what those additives are but not to submit any scientific proof of the additives’ safety. And although the FDA can do its own review of such ingredients, that process is often glacially slow. (The agency still hasn’t completed its rereview of several chemicals nearly a year after announcing that it was taking a second look at their safety.) Plus, certain additives may be safe when consumed on their own but harmful when consumed alongside other additives. A study published last year found, for example, that consuming a cocktail of emulsifiers, a preservative, and a dye was associated with higher rates of type 2 diabetes but that several other industrial concoctions were not.

If anything, researchers know less about how processing changes the ways humans digest certain foods. As the term ultra-processed implies, researchers believe that the ways food companies process individual ingredients has had a profound impact on our health. But because that concept is so new, very little research has been done on it. The Brazilian researchers, for example, have cited pre-frying as a mark of ultra-processing, even though it’s hard to imagine that precooking chicken nuggets could meaningfully affect how they’re digested. And although research does suggest that extrusion, which is used to make products such as cereal by pushing ingredients through a heated tube with rotating screws, makes it easier for the body to break down starches, the evidence of any ill effect on humans is limited.

All of this makes Kennedy’s push to define UPF difficult. Any earnest effort to “Make America Healthy Again” cannot ignore the fact that a majority of the calories Americans consume come from ultra-processed foods, and that something in those foods seems to be making people ill. Creating a definition is an obvious first step toward addressing the issue, and the previous definitions for ultra-processed foods have undoubtedly advanced researchers’ understanding of those foods’ impact on our health. A 2019 study using the Brazilian definition found, for example, that people ate faster and gained more weight when fed an ultra-processed diet compared with an unprocessed one.

But for researchers, the definition of UPF serves more as a hypothesis—an idea that they can test and adjust about how to think about this category of food. A government definition of UPF has a different purpose—it can prompt research, but it can also be used to make rules about how foods are tested, advertised, and sold. An overbroad definition could cost food companies millions in unnecessary costs and stoke unsubstantiated fear in consumers; a narrower definition, based on a hunch, could cause the same problems.

Government definitions are not written in stone: In 2024, the FDA overhauled an outdated definition of healthy foods, which was issued amid the low-fat craze of the 1990s and excluded foods such as nuts and fatty fish. But the process of revising the definition took eight years. As Kennedy said Monday, “If you can’t define a problem, you can’t solve it.” But if you define a problem in the wrong way, you probably can’t solve it either.