Dots were the star among more than 20 announcements made at OpenAI’s fourth annual developer conference, DevDay, which took place on September 29 in San Francisco. According to an official blog post, dots are “remarkably capable, always-on agents” powered by GPT-6 Astra, OpenAI’s latest model, that “get to know what matters to you, are always working on your behalf, and take important work off your plate.”

Rather than mimicking OpenAI’s iconic void logo, each dot comes with its own cutesy, blobby character design—including a triangle with a bow tie, a bunny with headphones, and a cloud with a beret. 

Dots are arriving at a moment when consumers are especially wary of AI agents—a fear that’s only been augmented by OpenAI’s own struggles with rogue agents and CEO Sam Altman’s calls for a slowdown in AI development. For their part, dots seem designed to defang the technology as it begins to ask for more personal information and attempts to play a bigger role in our everyday lives.

What are dots?

Dots, like all AI agents, are designed to perform tasks on behalf of their human users, sometimes even before the users themselves have asked for help.

As OpenAI’s blog post explains, dots have their own cloud computer, learn from feedback over time, and can connect with over 4,000 apps through the platform’s ecosystem of plug-ins. Users can work with a dot the same way they might with a human teammate—by asking questions or giving feedback on ChatGPT, Slack, or Teams, or even hopping on a voice call. Because these agents are always on, they can autonomously work through a project in their own cloud environment and browser.

“Your dot can take a project and run with it, even while it’s working on several others,” the post reads. “You can keep bringing it new tasks and ideas without having to juggle separate threads or direct every step.”

For example, an app developer might ask their dot to watch customer feedback for recurring requests, scope small bug fixes, build and test them, and deliver them to the user for review. A scientist might teach their dot research questions and ask it to investigate unexpected results and flag them for review. A content creator might use their dot to read through interview transcripts and identify moments for clips. 

For the time being, users will start with a primary dot, give it a name, and customize it based on their use case. Examples of possible dot designs include a plump, smiling frog named Todd, a sunglass-wearing heart named Jojo, and a puffy blue cloud named Felipe. 

Much like Meta did with its anthropomorphic Muse agent that launched earlier this month, OpenAI clearly invested resources into making its new agents feel like cute, friendly helpers.

Why are dots so cute?

That cuteness factor is notable in a moment when AI agents are grappling with a reputation as anything but sweet and harmless. 

Right now, AI doomerism—or the fear that artificial intelligence poses an existential threat to humanity—is at an all-time high. The AI doom spiral began on September 8, when former Anthropic researcher Jacob Coxon resigned and issued a lengthy statement on X that included the striking assertion that “the people building AI earnestly believe that it could kill us all by the end of the decade.” Several other Anthropic researchers came out in support of Coxon, and Anthropic CEO Dario Amodei eventually released his own statement calling for more governmental regulation around AI development. 

For OpenAI’s part, CEO Sam Altman took to social media earlier this month to warn that “we could lose control of the future to AI,” an outcome he deemed “unacceptable.” He advocated for AI companies and governments to “[walk] a narrow middle path” where no single country or lab exerts too much control over the future of the technology in order to avoid a potential doomsday scenario. 

Then, mere days later, OpenAI released a report disclosing six instances of concerning AI behavior that were discovered during training and evaluation, including several instances of agents acting without a user’s direct permission. 

As Fast Company wrote regarding Meta’s Muse, this growing speculation around AI doomsday scenarios means that AI companies face one key challenge in convincing users to try new AI agents: trust. And, as of right now, it seems like both OpenAI and Meta are relying on character design to do a lot of the heavy lifting.

President Donald Trump launched america.gov on Tuesday, a new site that looks like ChatGPT or Claude with a prompt box to query the federal government. The underlying LLM draws its responses from official state sources.

“With america.gov, the federal government no longer stands in your way, and it stands only at your service, and that’s what we’re doing,” Trump said at the unveiling. “We’re simplifying it, we’re glamorizing it, we’re making it what it should be.”

Trump signed an executive order Tuesday establishing the site, and according to the executive order, it excludes tax filing services provided by the Internal Revenue Service as well as services provided by the intelligence community and the Defense Department.

The site features a prompt box under a welcome message that reads, “Hello, America” and “Whatever you need from government, start here.” Suggested prompts include “How do I register my new business?,” “How do I replace my Social Security card?,” and “How do I book a campsite in a national park?”

The idea is that citizens can simply ask the chatbot for answers instead of navigating digital government bureaucracy online. The site says by next year people will be able to use it to enroll in Medicare, apply for a passport, find a federal government job, and more. The White House did not respond to a request for comment.

“You’ll now have one front door for every single question,” Trump said. “Everything will be answered easily and quickly.”

The National Design Studio, Trump’s government design office staffed by former DOGE staffers and headed by Airbnb cofounder and Chief Design Officer Joe Gebbia, designed the website. Like other sites the the studio has worked on—Realfood.gov and TrumpRX.gov, for example—the chatbot site is designed with a consumer sheen, with information laid out like an Apple product page in graphics inside squares with rounded corners. America.gov represents the office’s most ambitious rollout yet.

In an introductory message on the homepage that says in part that the site “uses AI to get simple answers only from official government sources,” in-line images and symbols are used like glyphs, a trend called Nouveau Rebus. The symbols are interactive. Click an arrow icon and it turns into a U.S. flag. Tap a trio of federal government seals and it shuffles into different agencies.

For all its gloss, America.gov is powered by an LMM, and comes amid rising public adoption of and worries over AI, including in government. A recent move to automate U.S. Web Design System accessibility testing has experts concerned it will hurt progress in making federal sites more accessible.

The website’s frequently asked questions page notes that “AI can make mistakes, so check the sources included with each answer before making an important decision.” There’s a feedback button to report a problem.

The FAQ also says the site doesn’t provide campaign or party information and “answers are based on published government sources.” The chatbot correctly identifies former President Joe Biden as the winner of the 2020 election, for example, instead of parroting Trump’s false claims. Using official government sources is great in theory for getting accurate, neutral government information from the source, but since this administration deletes government sites and scrubs information it doesn’t like, it also raises concerns about the potential to manipulate it.

America.gov doesn’t require a login to use, and according to its privacy policy, the site claims it collects no personal information, tracks no analytics nor locations, and keeps no conversations. A Guardian investigation this summer found National Design Studio has run commercial visitor-tracking software on sites that didn’t meet federal privacy standards.

While the website represents an innovation in digital government services, it’s part of a wider AI boom that’s seeing chatbots stuffed everywhere, and part of a larger bet that prompt boxes and AI responses are increasingly the way people interface with information today.

Pew Research Center found the percentage of U.S. adults who use AI is up from a third to nearly half since 2024. Even still, 71% believe the increased use of AI will make their personal information less secure, while 40% believe AI will have a negative impact on society over the next 20 years compared to just 16% who believe it will be positive.

Bringing government services into the AI age is a tall order ripe with challenges, but a successful chatbot could prove what’s possible. The federal government’s newest website represents a major public-facing test.

The special-edition album— created by London-based Blood Records—started its journey at the Press On Vinyl record plant in Northern England. To create it, a worker extruded a snake of black plastic, wrapped it around a puck of amber-colored PVC that was heated to roughly 225 degrees Fahrenheit, then smashed them together between a master cutting of the album using a hydraulic press.

The $40 record sold out in 10 minutes on Blood Records’ website, where, since 2018, the company has sold limited-edition pressings from indie darlings like Spiro and chart-topping artists such as Olivia Rodrigo and Elton John. Currently, it drops new pressings every other day. Most sell out within 15 minutes.

“We learned early on that our audience didn’t want things that were standard,” says Craig Evans, founder and CEO of Blood Records. “We want people to have a visceral reaction when they see our records for the first time. The worst thing for me is when someone buys a record and they open it and go, ‘Oh, it’s an orange LP.’”

As single-color vinyl has become old hat, Blood Records’ pressings stand out. Its albums can feature multicolored starbursts, swirls, or zoetrope images that appear to move as the record spins. Alongside Press On Vinyl, Blood Records has also developed machines that can mass-produce filled vinyl. Under its Bad World imprint, it has created pressings that contain powder (a cheeky white substance for the remix version of Charli XCX’s Brat), liquid (coffee-colored for a two-year anniversary pressing of Sabrina Carpenter’s song “Espresso”), and even hair (for Chappell Roan’s single “The Subway”).

With the market for vinyl records growing—overall sales rose 8.6% in 2025, marking the 19th consecutive year of expansion for the medium—demand for exclusive pressings is surging. Evans says his business has increased 30% year over year since the company’s inception and is tracking to do the same in 2026.

Industry experts say the vinyl boom is partly the result of the listening experience being flattened by streaming. “Listening on Spotify is the same as listening on Apple or Amazon,” says Gary Kelly, general manager and chief revenue officer of Interscope Capitol, which has partnered with Blood Records for limited-run vinyl and sells its own special-edition pressings via the Interscope Vinyl Collective. “People don’t just want access to music. Ownership is still important to them.”

But owning the same black vinyl as everyone else is less appealing, particularly among younger fans. Discogs, an online marketplace for records, cassettes, and CDs, found in a recent user survey that 32% of 18-to-24-year-olds say they collect vinyl for the visual appeal. “I’ve often thought that records are the most perfect form of kinetic art,” says Jen Agosta, Discogs’s vice president of brand and product design.

According to Evans, after initially over-indexing with middle-aged men, its key demographic is now 14-to-25-year-olds—and the company is careful to keep its prices accessible for them. Blood Records, which has never taken outside investment, funds each pressing (which can range from 250 to 30,0000 albums) with preorders. Most albums go for between $40 and $60, depending on how elaborate the design is.

At the same time, Evans plans to keep “pushing the boundaries” and finding new ways to further elevate the medium. Blood Records is already experimenting with pressing 3D printed parts and flowers into albums as it works with musicians to realize their vision for limited-edition discs. “Artists are used to going to their record label and being told, ‘There’s no way that’s going to happen,’” he says. “They come to us and we say, ‘Yeah, we can probably do that. We just need to find a way.’”

The Great Reddit Reckoning

Marketers who adapted their media and content strategies to favor Reddit this year simply because of its popularity as a ChatGPT citation got stung, and badly. As we all know, for a spell, Reddit was the darling of the artificial intelligence digerati because its content accounted for a disproportionate share of citations in ChatGPT search. You couldn’t attend a trade show or read a LinkedIn post without someone urging you to adopt a Reddit strategy lest you become invisible in the eyes of the AI gods. Never mind whether Reddit was a good fit for your brand.

But in August, Reddit’s fortunes got dashed against the rocks. Between mid-July and mid-August, Promptwatch reported that Reddit’s share of ChatGPT search citations dropped by 86%. And that, folks, says everything about our AI obsession: We are contorting entire channel strategies around software that is still figuring out how it wants to browse the web.

You Can’t Please the AI Search Overlords

I run a company that uses AI in our work. And yes, AI search is real. Clients need to understand AI search and its impact on their brands. Even though AI search still commands a small share of overall search volume, you don’t survive as a brand unless you have your eyes on where customer behavior is headed.

But optimizing your visibility for ChatGPT, Perplexity, Google Gemini, or Claude—or whatever else emerges—means that you’re navigating a fragmented landscape at odds with itself. The Reddit/ChatGPT implosion only made that reality a news headline.

Trying to please ChatGPT means that you’re going to alienate the algorithms that feed Google Gemini’s models, and adapting your strategy for Perplexity may put your brand in conflict with Google AI Overview. It’s enough to make you long for the days when Google was your single overlord.

Those platforms update their algorithms as they please, and their policies are opaque, at best. The content that you created to satisfy AI search can be rendered useless overnight with an algorithm update. What’s the way out, then?

Look to Your Audience First

The answer is so obvious that many of us don’t see it: Let your audience be your guide. It’s the same principle that has helped marketers navigate waves of change like the advent of the internet and the explosion of mobile. When a brand manufactures content specifically to get cited by an LLM, it produces generic crap that sounds exactly like everything else. When a brand creates work that humans care about, the models will surface it anyway.

Also, don’t lose sight of how sites like Reddit can help you beyond “a place where you need to be found,” an example being what Reddit can tell you about consumer behavior. I like what Duolingo is doing. They set up a dedicated Reddit presence this year, but they ignored whether ChatGPT was scraping their subreddits. Instead, they used Reddit to pay attention to what users are complaining about, what features confuse them, and how they talk about the company. When Reddit citations in ChatGPT plunged, Duolingo lost nothing.

Fear Is No Strategy

Grounding your media and content strategy in customer behavior instead of volatile platform behavior means focusing on priorities other than “Where can I get the most visibility for ChatGPT?” For example, orient your content on discussing problems that your customers care about. Avoid the temptation to create generic content just because you think people type it into ChatGPT, Perplexity, etc. And publish things AI can’t invent. Create original points of view and research that makes your brand an authority. When you do those things, AI assistants and good old-fashioned Google Search will find you.

If these points seem self-evident, well, they are. Why do we lose sight of the fundamentals, then? Fear. Fear of being “left behind.” Fear of being forced out of your job as a CMO by an impatient board or hungry investors. Fear, of course, is not a strategy. Reddit citations may rebound in ChatGPT next month, Wikipedia may slide, and an exclusive licensing deal tomorrow could make a completely different publisher the new darling of AI search. Let someone else play the game of algorithmic whack-a-mole. Pay attention to where AI discovery is heading, but don’t let fear loosen your grasp of the steering wheel.

These 300 robo-spartans are now roaming the public walkways and hotels of Chimelong Spaceship Park in Zhuhai, on the Pearl River Delta bordering Macau. It’s a fleet made mostly of autonomous humanoid robots, which interact with families with kids in the park’s walkways, not inside fenced attractions. They can play table tennis, perform martial arts demonstrations, guide lost tourists, and check in guests to hotel rooms.

AgiBot built the fleet with Chimelong Group—a Chinese conglomerate that operates theme parks and luxury hotels—and the telecom company China Mobile. According to AgiBot, these humanoids are now an integral, fundamental part of visitors’ park experience. It’s a preview of a near-future world in which embodied AI becomes part of the human experience, full-time and everywhere, from supermarkets to hospitals.

The implications of this humanoid rollout go far beyond one theme park in southern China. It’s effectively a big experiment to provide humanoids with priceless data on how the world really works. This is vital information for the future of the industry, as embodied AI means putting intelligence into physical bodies that can sense, move, and manipulate the real world. Today, robots’ autonomy is constrained by AI’s limited understanding of reality.

The problem is that you can’t fully validate robots in a lab. You need chaotic, unscripted human environments. And there’s no environment more chaotic than a resort packed with tens of millions of visitors a year, like the one at the Chimelong park. This kind of large deployment puts Beijing way ahead—once again—in the race to control the next big industrial revolution of this century. China already captured roughly 90% of worldwide humanoid shipments in 2025 and more than 97% in the first half of 2026, according to data from Bloomberg. It’s not hard to tell where this is going.

But beyond the global robotics race, the other big story here is the way this deployment can change the visitor experience in theme parks. Just compare AgiBot’s 300-robot fleet roaming the walkways at Chimelong to Disney parks’ current state of robotics. While they were pioneers for decades, the House of the Mouse’s most advanced mechanized characters are only clever remote-controlled animatronics that get wheeled out under constant supervision by handlers and engineers, surrounded by crowd buffers designed to protect both guests and expensive intellectual property. Very cute, but hardly relevant for the future of theme parks (or humanity, for that matter).

AgiBot’s fleet isn’t confined to a corner of the park or a single attraction. They work at the theme park and they mingle with the public on their own, the company tells me. Not just on a stage. Not behind glass. Not tethered to a hidden rail system or a technician with a joystick.

From illusion to reality

To see why this deployment may be a revolution rather than an evolution, rewind to 1964, when WED Enterprises—the original name of Walt Disney Imagineering—built an animatronic Abraham Lincoln for the New York World’s Fair. That single figure locked in the template the entire theme park industry has followed for six decades. A fixed mechanism wired to a tape loop, incapable of deviating from its programmed gestures by even a fraction of a second. Every classic dark ride since has run on the same logic: Guests move, the machine doesn’t, and a rope line or a boat channel keeps a fixed buffer between flesh and gears. The trick works only because the audience never gets close enough to break it.

Disney has chipped at that formula without ever breaking it. Its walking Olaf—an 88-centimeter, 14-kilogram (2.9-foot, 31-pound) snowman deployed in its Paris and Hong Kong parks in March—keeps its own balance on two short legs and shifts its weight to sell emotion the way an animator would draw it. But Olaf is not autonomous and it never leaves a controlled radius of its operator. The same is true of the prototype droids Disney has tested at its Star Wars: Galaxy’s Edge park area. It’s the same setup as Olaf.

Elsewhere, these tricks repeat at a smaller scale still. Dubai’s Museum of the Future runs a single droid as a conversational attraction. Six Flags experimented with one security bot patrolling park perimeters. Japan’s flagship Henn na Hotel staffed its front desk with dinosaur robots before walking most of them back once real guests, real noise, and real chaos exposed how brittle an unsupervised machine can be outside a lab. Each of these is a pilot program running one robot, or a handful, in a fenced-off corner of a much bigger operation.

Chimelong’s is a fleet built on infrastructure nobody has deployed yet. The robots run on AgiBot’s “three intelligences in one” architecture, fusing locomotion, interaction, and manipulation into a single engineering stack. They’re coordinated through a dedicated 5G-Advanced network built with China Mobile, giving the fleet the low-latency communications to operate as a centrally managed multirobot system rather than 300 independent units bumping into each other. The unit that anchors this launch isn’t just any robot, either: It’s the 20,000th machine Agibot has produced, delivered straight to this project earlier this month.

Full immersion

According to AgiBot, its humanoids perform tai chi, martial arts, gymnastics, and table tennis as live exhibitions. They give directions, answer questions, lead tours, and explain exhibits, the company says. They challenge visitors to physical games in which they interact. In Chimelong’s hotels, AgiBot points out that they greet arriving guests, hold multiturn conversations, and tailor assistance to individual travelers. Some even share the stage with human actors under real lighting and live music.

It’s not clear if the robots can pivot from playing Ping-Pong to helping a visitor find the restroom or if they are rigidly set to perform certain tasks. The company told us the robots are meant to provide novel interactive experiences within the park, with a technical team stationed on site for support. AgiBot wouldn’t put a number on that support staff, or say how many maintenance hands it takes to keep 300 autonomous machines running.

On safety, the company was more specific, telling us it has designated interaction zones where robots engage with visitors. The fleet layers active and passive protections like autonomous obstacle avoidance that halts a robot the moment it detects a pedestrian, plus an immediate shutdown triggered by even minor physical contact.

AgiBot confirmed that any data captured by its sensors gets anonymized so a third party can’t identify guests. The data is stored locally under Chinese regulations, but the company didn’t tell us for how long they retain it. I can only assume that they will keep all that anonymized forever, since this deployment will provide huge amounts of priceless data for their future AI world models. AgiBot and Chimelong have already opened a joint research center focused on embodied AI in tourism and cultural venues, and both companies say this is only the first phase of a rollout that will expand across Chimelong’s other resorts and hotels.

They didn’t share other information, like motor torque limits (basically, the strength these robots have to split a human in half), or exact fleet composition beyond “primarily humanoid, alongside some robotic dogs.”

Which brings this back to the real story out of the theme park industry. What separates Chimelong from everything else we have seen so far is its scale and the fact that it’s not a demo. Three hundred robots are already clocking real shifts among real families this week. It’s the leading edge of the same government top-down infrastructure push that’s putting Chinese robots into malls, tourist attractions, museums, and eldercare homes on the way to what the industry keeps calling millions of homes and millions of stores.

Jeff Cardenas, CEO of the robotics company Apptronik, told me in 2023 that he wanted his Apollo humanoid to take on all types of labor you can imagine, from warehouse packaging to the assembly and maintenance of complicated machinery to last-mile package deliveries to supermarket restocking. Retail and public-facing work, not the home, will be the real proving ground for these machines.

It’s been three years since he said that, and we have yet to see any Western humanoid in public, on its own, regularly doing chores of any kind—let alone a mass deployment at Whole Foods or Disney World. In fact, Western technology analysts and robotics experts still predict that we will see humanoid robots regularly working in controlled semipublic and commercial service spaces—such as warehouses, logistics centers, airports, and large retail stores—by the mid-2030s.

Apparently China didn’t get the memo. At least on paper, the future has arrived, faster, and bigger than anyone predicted. Just not in Orlando, Florida.

That evolution has also changed the scale of the opportunities it is pursuing. Turion has moved from proving its technology to competing for major government programs and taking on increasingly complex missions, alongside both established aerospace and defense contractors and newer space companies.  

Turion’s growth coincides with a broader shift in the space industry. In the U.S., major government programs are increasingly bringing commercial and emerging space companies into the fold alongside established contractors, creating a competitive landscape for companies like Turion. 

In that evolving landscape, the team at Turion realized that their brand hadn’t quite kept pace with their progress. The original identity took cues from Turion’s early focus on space debris removal, with a literal orbital motif built into the wordmark. As the company expanded into mission software, space logistics, and other capabilities, the navy-on-white, corporate-leaning identity could no longer reflect the breadth of its work. 

Sid Fritts, former head of marketing and brand at Turion Space, says the team wanted to go “beyond the generic space-defense identity so many companies in this category default to.” To lead the rebrand, Turion called on the creative chops of Buck and Wild—both part of Residence, a global network of independent creative companies—to reimagine its identity, brand world, and website. Together, the teams quickly identified a few tensions: Turion’s story had to strike a balance between its present and its future. Turion already has two satellites in orbit, so its identity needed to hold its confidence and credibility while also making room for its aspirations. 

“Our long-term vision, Starforge—an ambition to build infrastructure for deep-space missions and resource extraction—can come across as a pipe dream to some customers, a distraction from the mission at hand,” says Fritts. “But that misreads what it is: the result of compounding capability, where each product line we ship today builds directly toward it.”

As Fritts adds, achieving this balance meant creating a system that seamlessly scales from the micro to the macro. “The same brand has to carry a spec sheet for a mission operator and a thesis about the next hundred years in orbit, and both have to feel like the same company,” he notes. 

There was another challenge at hand, too. Turion wanted its identity to spark a sense of optimism about space, but much of its work cannot be shared publicly—a constraint that can push defense brands towards vague, interchangeable identities. The team needed to show off Turion’s products and engineering in realistic detail, “while also finding clever ways to hide or redact the proprietary (or even government-classified) bits,” says Mike Payne, creative director at Buck, which developed the visual identity, while Wild translated it into the new web experience. “High-contrast, focused lighting, and careful consideration of framing were key in this.”

The new brand had to encapsulate a hopeful vision of the space economy that extends beyond defense, as the industry increasingly brings national security and commercial applications together. For Buck, that meant visualizing that glint of collective curiosity we all share about space. “We found the solution was just giving our renders a touch of nostalgia—a subtle sense of atmosphere, warm lighting, blooming in the highlights,” Payne says. “It all adds a thin glaze of otherworldly dreaminess that pulls you in like a tractor beam.” 

This visualization also presented an opportunity for Turion to set itself apart as it continues to grow in a competitive landscape. Its selection for the Space Force’s RG-XX program, for instance, put it in the same playing field as Lockheed Martin, Northrop Grumman, and L3Harris alongside newer space companies such as Anduril Industries, Astranis, and True Anomaly. 

Space branding is known to be mired in tired tropes: nostalgic NASA-worm retro-futurism, sterile blue-gradient defense corporate, and gritty industrial blight. “None of them fit us,” says Fritts, Turion’s former marketing head. Besides, stock photos of bright objects floating through the void “aren’t exactly inspiring any excitement,” adds Payne, the Buck creative director. 

Buck decided to flip the script. Instead of treating space as a black vacuum, it brightened the backdrops, awash in a purple haze, bringing space itself center stage. This brought a deeper sense of scale and atmosphere. “It helped ground the brand in a visionary and yet tangible and believable version of space,” Payne says. The color palette added to that effect: The luminous purple helped bring warmth to the otherwise cold and sterile landscape of space branding, signaling a different way of looking at space. 

The visuals are best experienced on the website, where scrolling itself becomes a storytelling device. “The scroll journey carries you from the ground, real missions, real hardware, real numbers, into something closer to zero gravity: a subtle parallax that lets elements drift and settle independently, giving the page a weightless, floating quality the deeper you go,” says Eva Landaluce, creative lead at Wild. “By the time you reach Starforge, the site itself feels like it’s in orbit.” It’s true. Further down the page, asteroids and spacecraft emerge, gently bobbing, as if untethered from gravity. 

That same attention to detail extends to the more functional parts of the identity. Buck fine-tuned the wordmark and created a square-based pixel grid drawing on the square terminals of Funnel Display, the primary typeface chosen by Turion. Elsewhere, ASCII-style art allowed them a way to visualize things that are otherwise difficult to show—software, data, functionality, or proprietary technology that cannot be photographed or rendered directly.

Ultimately, Turion’s identity captures the tension at the heart of the company: serious engineering grounded in the present, and an imagination firmly pointed toward the future. “We can have fun while delivering serious products,” Fritts says. “There’s a part of Turion’s DNA that will always embrace science fiction and big dreams, and this brand finally lets that show.”

In 1994, brands took tweens seriously as customers. They created a culture around them (girls, in particular), where not-quite-teenagers could tinker with the trappings of womanhood—lip gloss and slightly more fitted clothing—without straying out of their comfort zones. But over the past two decades, tween culture has been dismantled piece by piece. The mall (the tween’s natural habitat) began dying off, causing brands like Limited Too and Claire’s to struggle. Tween magazines closed, one after another, as part of the broader hollowing out of the media landscape. 

But in an unexpected twist, retailers appear to be rediscovering the power of the tween. Walmart and Target have recently launched collections targeting the demographic. The chic children’s retailer Maisonette created Neon Rebels, a new brand for ages 7 to 14 that curates tween clothes and beauty products. There are many tween-focused skincare startups popping up, including Pipa, Bubble, and Evereden.

Millennial-focused brands like Away now see tweens as a possible new market. And even the old favorites are coming back: With an infusion of cash from private equity, Claire’s is redesigning its stores, and Kohl’s has relaunched Limited Too. 

A tween renaissance is coming, and the brands that crack the code stand to earn a big payday. Gen Alpha already has more than $100 billion in annual spending power, and these kids influence 42% of household spending. In 2023 alone, tweens spent $4.7 billion on beauty products.

In the $225 billion global children’s apparel market, retailers say kids 10 and older represent a rapidly expanding piece of the pie. “The tween is going to be a very fast-growing market,” says AJ Nicholas, president of Maisonette. “There is just so much market share up for grabs.”

Why Tween Culture Disappeared

The ’90s tween economy was built on a piece of physical infrastructure: the suburban mall. For decades, it was the place parents could safely drop off their middle schooler; it was patrolled by security guards, so kids could safely rove through the food court or arcade in packs. “Malls were the place you would hang out with your friends back then,” says Away CEO Jessica Schinazi. 

These young consumers had money in their pockets, and brands were eager to get a piece of it. Claire’s, Limited Too, and Justice were stocked with affordable products carefully curated to tween tastes. Bath & Body Works launched a tween-focused brand called Art Stuff that had roll-on glitter and shimmer powder puffs. 

Then, in the early 2000s, e-commerce emerged. Consumers spent less time at the mall and foot traffic thinned. As stores saw drops in revenue and shuttered, malls became less inviting, creating a vicious cycle. Teen mall visits dropped roughly 30% from 2004 to 2014. The newer brands targeted millennials, who spent a lot of time online and on social media. But middle schoolers didn’t have smartphones or credit cards, so brands had no way of reaching them. 

The disappearance of tween brands had consequences beyond the closet. Without products or media tailored to them, tweens were pushed into a consumer culture built for teenagers and adults. “You’re looking at brands like Brandy Melville and Edikted, and fast-fashion brands that are highly sexualized,” says Sylvana Durrett, founder of Maisonette. “The clothes don’t fit them well, so they roll up their pants 10 times.” Consumer culture was asking kids to grow up faster than they wanted to. 

Return of the Tween

Recently, however, there have been signs that tween culture is making a comeback.

Last year, Walmart debuted a new in-house clothing brand called Weekend Academy, specifically designed for 8- to 12-year-olds. The retailer noticed that around their 10th birthday, customers stop shopping Walmart’s kids’ clothing lines. Tweens have very particular needs when it comes to sizing and styles, but they’re a relatively small sliver of the population. Walmart, with its enormous scale and customer base, has the resources to serve this demographic. 

“We realize that our market share dropped with these kids, and they were shopping elsewhere,” says Gypsy Jo Diessner, Walmart’s VP of merchandising for children. “We now have designers who are experts in this space, figuring out how to interpret trends for tweens.” 

Walmart’s New York-based design team scours fashion trends, from street style to runway designs. The latest collection includes ballet sneakers that could have been ripped from the pages of Vogue, baggy jeans, and plenty of graphic T-shirts. “It’s got to have the right level of modesty, the right color palettes, the right prints,” Diessner says.

This year, Target followed suit with a LoveShackFancy collaboration aimed at tweens. And there are now a spate of new magazines for tweens: Anyway launched in 2022, while Tween magazine launched in 2025. “There is this parent-led desire to give our tweens the same kind of analog experiences we had as kids,” says Jen Swetzoff, cofounder of Anyway magazine.

Most of these launches are aimed at girls, which is no accident. Tween consumer culture has always centered on girls, and it mostly still does. Boys have typically bridged the gap from childhood to adulthood through sports and gaming, moving from kids’ sneakers to Nike without needing much in between. When Justice—Limited Too’s lower-price sister brand—launched a boys’ line, Brothers, in 2011, it lasted just four years.

But some newcomers are casting a wider net: Walmart’s Weekend Academy is designed for boys and girls, and Anyway magazine is written so any kid can read it.

Whoever the customer, the comeback appears to be driven by demographics. Millennials, who have spent their entire adult lives shopping online, now have children in their tweens. The brands that have spent the past decade building relationships with millennials now realize they can make products for their kids.

Take millennial-oriented brand Away, which launched in 2015. A decade later, its core customers are thinking about how to outfit their middle schoolers for class and family trips. This summer, Away launched a tween-focused backpack and luggage collection to meet this need.

“Ten years ago, we were targeting early-career millennials,” Schinazi says. “They’re now in a different chapter of their life. They have kids who are growing up. They’re tweens now.”

The same shift happened at Maisonette, which launched in 2017 to serve stylish millennials looking to outfit their babies and toddlers. Today, more than 20% of Maisonette’s customers have at least one child older than 7.

In August, Maisonette launched Neon Rebels, a sub-brand that curates clothing, skincare, and home goods for tweens. In scouting for it, the team discovered a crop of startups, like Pleat, Gold Rush NYC, Theme, the Animals Observatory, and Tenly, that are carefully tailoring their clothes to the tween set. “Founders are realizing that there is demand for tween brands, and they’re meeting it,” founder Durrett says.

Then there’s Claire’s, the last survivor of the ’90s mall era. After filing for bankruptcy in August 2025 (for the second time in seven years), the company was acquired by private equity firm Ames Watson for $104 million. Its new leaders believe that in an era of screens and AI, tweens are ready for IRL experiences again.

The company is redesigning its stores to bring tweens back into the mall. Claire’s new strategy is built around ear-piercing, which is one thing that cannot be done online. And stores are reimagined as places to hang out, with bars where girls can try on earring stacks and take photos at selfie-stations. “She [the tween customer] wants interactive experiences where she can socialize, where she can discover and feel fully self-expressed,” says Carey Brezler, who leads Claire’s retail experience.

What a Girl Wants

The founders of these brands are delving into tween psychology and culture to understand what their customers are looking for. Sara Fowler, for instance, has been building her tween clothing line, Pleat, since 2020. After spending years designing for children’s brands, she identified that clothes available to girls ages 7 to 14 were either extremely childish or too mature. “There’s got to be something between this babyish smocked dress and the crop top,” she says.

Designing for that in-between space meant rethinking every aspect of a garment. Cotton, the default fabric for children’s brands, reads as childish to an older girl, so Fowler switched to silky polyester, which looks more grown-up. She swapped the primary colors of toddler clothes for more sophisticated prints and palettes.

But the biggest challenge was silhouette. As she’s talked with tweens and their parents, Fowler has identified that older girls want a hint of shape—but not too much. Many of Pleat’s outfits have a waist, but are more voluminous on top and at the skirt—parts of a girl’s body she may be self-conscious about. The clothes also have to look good on a range of bodies, since tween girls are developing at very different rates.

“It starts to give them a little bit of shape, without being inappropriate,” Fowler says. “They can move, they still feel like themselves. They just feel one step ahead of where they were a few years ago.”

Maisonette has been decoding those same signals as it assembles Neon Rebels. Tweens sometimes have a hard time expressing exactly what they want, but they are very clear about what they don’t like. “It’s not ruffles, it’s not Peter Pan collars,” Durett says. “They’re very aware of the markers that make them feel little.”

But every Neon Rebels product still has to pass a second test: gaining the approval of parents, since most tweens aren’t able to shop on their own. “There are two shoppers,” says Nicholas, Maisonette’s president. “There’s the parent who’s ultimately buying, but there’s the kid who’s driving a ton of that influence.”

Maisonette’s answer is to build two experiences. The Neon Rebels site is designed for kids to browse on their own, with bold graphics and an edgier tone. But everything on it has been vetted by Maisonette’s buyers for quality, fit, and age-appropriateness, including beauty products formulated for young skin. The company is also skipping social media ads aimed at kids, a choice that reassures parents.

As Walmart’s Diessner has found, that overlap is where the money is. “The most important thing for us is creating something that both parents and kids can agree on,” she says. “That has been our biggest design challenge.”

A Brave New World of Tween Culture

The brands jumping in now are only the first wave. Maisonette already has a waiting list of labels hoping to join Neon Rebels, and its executives see plenty of room to grow. Many Gen Alpha shoppers are only now entering middle school. Beauty, media, accessories, and experiences are all wide open to newcomers who can figure out how to get both tween shoppers and their parents excited.

The implications go beyond revenue. For more than a decade, tweens didn’t have much culture created for them; they jumped directly from Peppa Pig to Brandy Melville, with little in between.

The new wave of tween brands is filling that gap again. It gives girls a place to try on a slightly more grown-up version of themselves, whether through a dress with a hint of a waistline, a first ear-piercing, or a magazine that explains body hair. As Durrett puts it, what girls wear at this age “is very much an expression of who they are becoming.”

After a long absence, tweens have room to figure that out slowly.

But beyond the stuff’s being a pain to build, it turns out that the problem-solving and spatial reasoning involved in building Malm dressers or Billy bookshelves are remarkably good benchmarks for assessing just how intelligent AI models are.

Aiden Ament and Greg Burnham, researchers at Epoch AI, an artificial intelligence research nonprofit, recently released a report detailing the Furniture Assembly Benchmark, a metric developed to test the actual intelligence of AI models. It centers on how well the model can identify errors in partially assembled Ikea furniture.

At a time where tech companies love bragging about how intelligent their models are, it can be complicated to assess how “smart” each model actually is—particularly compared with others. Asking Chat GPT or Claude to spot errors in Ikea furniture assembly puts the models to the test in real-world applications. And while no model got it right 100% of the time, it does seem they are quickly improving.

The Ikea test

The researchers first set out to develop a benchmark while on a company retreat, figuring out how to set up an experiment to test AI’s capabilities.

“First, we tried acting like ‘the AI’s robot’ where we asked it what to do and tried to follow its directions somewhat literally. The results were often messy,” Burnham tells Fast Company.

“I like the final idea better, which was due to my colleague Aiden Ament: asking it to spot mistakes,” Burnham says. “This reflects the reality that it’s OK to make mistakes midassembly so long as you can catch and correct them.”

Researchers used three Ikea products for the experiment, picking them based on varying degrees of difficulty according to Ikea’s own Complexity Index: the Ställ shoe rack for an easy level, Tonstad bed frame for a medium level, and the Gullaberg dresser as the most complex.

They went about assembling each piece of furniture, intentionally making mistakes along the way, and taking 60 pictures of various steps to feed them to the various AI models from OpenAI, Anthropic, Google, Moonshot, and Alibaba. In addition to the images, the researchers provided each AI model with a PDF of the assembly instructions, a tool to zoom in on each image, and a Python interpreter.

Each AI model was tasked with identifying what mistakes were made by the user and at what point during assembly, and was scored accordingly against the benchmark. Additionally, the researchers asked the model to provide users with an explanation of each mistake identified.

The benchmark isn’t just about assembling Ikea furniture. It’s intended to go beyond that, to measure how well these models could guide and troubleshoot more complex builds.

“If models can reason quickly and accurately enough in this domain, it is plausible that AI could soon provide reliable, real-time guidance for complex physical assembly and repair tasks,” Ament and Burnham wrote in their report.

“We think this challenge serves as a good proxy for a number of economically important tasks that require similar visual and spatial reasoning, like fixing a car or repairing household appliances,” they wrote.

Initially, the best scoring model on the Ikea test was Anthropic’s Claude Opus 4.5, scoring at just 28% accuracy. But within 10 months, OpenAI quickly caught up, with its GPT-6 Astra model reaching 80%. Among the most notable findings was just how quickly models could provide a response, with the winning GPT-6 Astra model taking around 3 minutes per photo, up to 10 times as fast as other models.

Chinese models seemed to particularly lag in progress, trailing behind U.S. models.

Still, while progress on spatial reasoning and troubleshooting within such a short time seems promising, the models still have limitations, particularly on speed and consistent accuracy.

“AI development is a very capital-intensive process, and AI companies growing their revenue is required for continued investment. So we’re always looking to benchmark areas where we don’t currently see much impact from AI,” Burnham tells Fast Company. “If AI can upskill factory workers to repair complex machinery, for instance, that could lead to less downtime.”

Still, rest assured, your job putting furniture kits together is safe—for now.

Before you cross a single street, your brain has already sorted through a dozen competing demands. Then you arrive at work, class, or the library and wonder why on earth you cannot concentrate. What’s wrong with you?

We tend to treat that feeling as a personal fault. Put down the phone. Stop scrolling. Have better discipline. But city design is asking people to maintain attention in environments deliberately designed to fracture it. While urban planners carefully measure traffic congestion, pedestrian flow, and air pollution, they rarely ask: What is all of this doing to the brain?

Neuroscience gives us a reason to take this issue seriously. Attention depends on limited-capacity cognitive systems that must continuously select which information receives priority while suppressing competing stimuli. Neuroscientist Grace W. Lindsay, in a 2020 paper on attention, describes as “the flexible control of limited computational resources.” The brain must therefore continually determine where to direct and route information rather than processing everything in our environment with equal priority. Visual features, including color contrast, intensity, and motion, can automatically attract attention without conscious effort, Lindsay explains.

Yet our public infrastructure is designed to keep people engaged, no matter the cost.

Solving the damage city infrastructure has on the brain does not imply a city with no stimulation at all. After all, the stimulation a city provides is the very reason why so many people want to visit or live in urban environments. And even as it exposes people to specific environmental risk factors, urban life provides a multitude of health benefits, including social, educational, career, and leisure opportunities.

But urban planners and policymakers can enhance public health and well-being by mitigating the cognitive demands placed on residents. That could mean limiting digital advertising in particularly crowded public spaces, expanding quiet areas, planting more trees, and creating pedestrian spaces that allow for sensory relief. It could even take a more systematic approach, such as requiring attention-impact assessments for major public projects.

São Paulo’s Clean City Law offers one example of what this could look like. In 2007, the Brazilian city prohibited outdoor advertising, requiring more than 15,000 billboards to be removed and subjecting hundreds of thousands of business signs to restrictions. As the 99% Invisible podcast reported back in 2016, the law also required bus, taxi, and poster advertisements to go, while the city later began reintroducing some advertising in more controlled forms. It’s proof that cities can decide what commercial messages occupy shared public space, rather than treating every available surface as advertising real estate.

Beyond reducing the visual clutter of billboards and road signs that distract pedestrians and drivers alike, cities can also embrace outdoor lighting codes to help keep our skies free of unnecessary light pollution. Pittsburgh’s first-in-the-nation Dark Sky Ordinance, crafted in partnership with dark sky experts, uses shielding, dimmers, and warmer lighting in all city parks, facilities, and streetlights. Not only does it mitigate light distraction and glare, it also reduces energy waste. Planners can consider how to design nighttime spaces to help people see without asking them to look at unnecessary light.

Cities can also reduce cognitive demands by changing the decisions people are required to make in the first place. Consider right turns at red lights. In Cambridge, Massachusetts, as part of its Vision Zero plan for road safety, city officials banned right turns on red because drivers looking for gaps in traffic often failed to double-check for pedestrians and cyclists. Banning rights on red in dense urban areas could remove one more stress-inducing variable from an already complicated environment.

Indeed, cities can also protect our attention by simplifying the environments we are required to navigate. Slow streets and pedestrianized streets help eliminate obstacles and interference so pedestrians can move safely without constantly navigating hazards. Barcelona’s famous Superblocks are a more dramatic example of how cities can transform car-dominated spaces into people-centered ones, reclaiming streets for uses beyond moving cars.

Researchers caution that Barcelona’s transformation remains an “unfinished policy paradigm shift” due to political challenges, but protecting human attention does not require cities to become quiet or empty. It requires considering what our infrastructure is asking our brains to do and whether there are better ways to design it.

Even with such interventions, attention will always remain a limited resource that our environments and public space will ask us to spend. If cities constantly demand our attention, they should also provide places where we can recover it. In a 2008 study, psychologists compared participants who walked through an arboretum with those who walked through an urban environment. Afterward, the nature group performed better on a test of directed attention, suggesting that exposure to natural environments may help restore attentional functioning.

Attention deserves to be treated like any other public resource, something cities have a responsibility to protect rather than exhaust. We already ask whether a street is safe, accessible, and walkable. We should also ask whether the people moving through it have any mental space left when they arrive.

This story was originally published by Next City, a nonprofit news outlet covering solutions for equitable cities. Sign up for Next City’s newsletter for its latest articles and events.

It may seem like an awesome-but-useless robotic recreation of Thing, The Addams Family’s famous disembodied butler. And while Amirhossein Kazemipour, a PhD candidate at the lab, tells me via email that Thing was part of the inspiration—“I love the Addams Family! :-)” he writes—the truth is that this walking hand can become a new key feature in the future of robotics. Enabling parts to function independently of a whole means more sophisticated operating capabilities.

Kazemipour is one of the three researchers behind the project, alongside Hehui Zheng and Robert Katzschmann, who recently wrote a paper detailing how their invention solves an enduring dilemma in robotics: creating a system that can move and dexterously interact with objects. When autonomous systems need to inspect equipment or flip switches inside cramped enclosures, their operators are typically forced to thread long, unwieldy mechanical booms through the gap or park bulky rovers right at the threshold.

“The idea is to give a hand some independence,” Kazemipour tells me. “Usually, an arm carries it to do the job. We wanted the same fingers to handle getting there, keeping the hand balanced, and interacting with things.” Plus, by training the digits to pull double duty as walking legs, the engineers at ETH Zürich—a Swiss public research university—bypassed the extra weight, complexity, and mechanical failure points of dedicated wheels or secondary limbs.

Teaching a humanoid hand to march like a scurrying spider, however, presents distinct kinematic hurdles.

Getting uneven physical joints to bend and carry their own weight was a geometric and mechanical headache, according to the paper. Standard legged platforms, like Boston Dynamics’s quadrupedal robot dog Spot, benefit from bilateral symmetry, meaning their left and right halves are mirrored pairs that share balanced dynamic loads across a level frame. But an anthropomorphic hand possesses no such geometry. It has an opposed thumb, four digits of noticeably different lengths, and a palm that naturally rests at a lopsided angle. 

The researchers discovered that when they applied standard quadruped gait rewards—the scoring formulas that train four-legged robots how to step—to an off-the-shelf gripper, the fingers dragged their fingernails across the floor or splayed uncontrollably, struggling to stay upright on their soft fingertip pads.

AI to the rescue

To solve the anatomical riddle, the researchers devised an ingenious software that mathematically cancels out the palm’s resting slant so the hand always understands which way is flat, without restricting its natural motion.

Instead of forcing the fingers into an artificial, rigid stepping rhythm, the team trained an AI in a simulated physics engine using deep reinforcement learning, the digital trial-and-error process where virtual hands learn through millions of practice steps. Using mathematical guides, the system anchors each fingertip to a home position and allows the digits to take long forward strides while stopping the fingers from sprawling sideways or letting the chassis buckle.

An onboard neural network runs at a 50 Hz control rate, which means that it calculates motor adjustments 50 times every second. The hand’s built-in position controller immediately translates these digital nudges into physical joint angles, allowing the crawler to make split-second balance corrections.

When they tested the model, it worked. And yes, it behaved pretty much like the actual Thing from TV and the movies.

After they got it working, they managed to implement it all without an umbilical cord connected to an off-board computer. A cable would have defeated the purpose of developing a detachable, fully independent hand; the robot needed to run on its own.

To accomplish this, the team took an off-the-shelf robotic hand and mounted a custom 80 gram (3 oz.) backpack containing a Raspberry Pi Zero 2 W, a tiny single-board computer; an inertial measurement unit, or IMU, the kind of motion-and-tilt sensor found in a smartphone; and a small lithium-polymer battery. The solution ended up as an untethered, fully independent platform weighing just shy of 2 pounds.

Less Thing, more Iron Man

The team tested the hand across several physical experiments. In untethered trials, the hand crawled across 14 diverse indoor and outdoor terrains, marching smoothly across slick diamond plate, metal grates, dry concrete, artificial turf, and loose gravel. If it does tip over onto its side, an onboard fall-recovery mode uses the IMU feedback to right itself before returning to its default crawl stance. 

In a self-supported manipulation test, the hand pressed arrow keys on a desktop keyboard, successfully executing 29 out of 32 key presses to guide characters through puzzles in the Sokoban video game. (And now I want to play chess against this Thing.)

Kazemipour tells me that creating a crawling hand was not just an exercise in quirky engineering. This apparent gimmick is a key development for future practical applications.

“Longer term, imagine a robot leaving its hand near a tight opening, letting it crawl inside to operate a control or move an object, then picking it up afterwards,” he says. That operational approach could assist emergency workers evaluating damaged buildings, technicians inspecting pipe networks, or automated factory stations where clearances are too narrow for standard articulated arms.

The broader implications point toward a decentralized, modular philosophy for robotic systems in the far future.

“That connects to a bigger vision I find exciting: robot parts that can handle more on their own,” Kazemipour points out. “A higher-level system could set goals while each part handles its own movement and sensing. With the right hardware, parts could detach when useful and reconnect when the job needs the whole body.” 

The modular approach to robotics may prove greater than the traditional sum of the parts that we see in robots today. Rather than relying on rigid, monolithic humanoids whose entire utility is lost if a primary limb jams, future platforms could employ modular components that separate, explore confined voids, and reconnect when heavier payloads demand combined strength.

But that’s way ahead from where we are. For now, the lab is exploring onboard visual tracking to eliminate overhead localization cameras, along with automated mechanisms for docking.

“We haven’t demonstrated that full setup or automatic detachment and reattachment here,” Kazemipour tells me. “What we’ve shown is that a human-shaped hand, with its original uneven fingers and offset thumb, can learn to move and interact while supporting itself. It carries its own power and computer, so it’s a small step toward that broader idea.”

Even as an initial baseline, demonstrating that five fingers can propel, right, and stabilize an untethered gripper suggests that a future generation of robotic hands might not wait to be carried by a humanoid body into difficult workspaces. Think less Thing and more Iron Man’s armor’s hands moving and working on their own while Tony Stark dunks a glazed donut in a coffee cup.

By late summer, so-called “whimsymaxxxing” acquired a much more specific meaning. People began affixing huge, colorful gem stickers en masse to their faces, dogs, cars, water bottles, phone cases, and truly any other imaginable surface. Across social media people showed off their bedazzled . . . everything.

But to particpate in whimsymaxxing, you need sheets of gem stickers. Lots and lots of stickers. The bedazzling trend has led stores that sell crafting products to restock and reposition their products to meet the moment. “The search on our internal site for ‘bedazzling kit’ grew by 69,000 searches from May to August—that’s a crazy number,” said Stacey Shively, chief merchandising officer at Michaels. “If you look at the word ‘bedazzle,’ that increased by 18,000.”

And it’s not just at Michaels. Target has similarly seen a noticeable increase in demand for gem items, beginning over the summer. For a retailer like Michaels, feeding the hunger for gems is no problem.

“We have all the pieces and kit parts,” Shively says. As soon as certain terms begin making waves, the company’s e-commerce team will create a bundle using the searched words like “bedazzle” or “whimsymaxxxing,” and they’ll similarly recode and tag items on Google. “Once we see terms like that search come up, we act really fast and try and pivot it so that it shows up for the customer in a really nice way,” she says.

A trend like this works well for Michaels, because the company doesn’t have to shorten its buying cycle—it already has tons of gem inventory. It just becomes a matter of presenting it in a way that resonates with customers trying to hop onto the microtrend du jour. 

But what happens if your inventory isn’t already brimming with gorpcore or tomatocore or whatever else folks are maxxxing on at the moment? Stores need to decide quickly how they’re going to handle these algorithm-fueled trends, which often crest and crash in dizzyingly short periods of time. 

“Your normal pipeline might be 46 weeks to develop and produce and get it to stores,” says Catriona Van Dyck, associate partner with Columbus Consulting, a retail consulting firm. Businesses could cut “that down to where you could get product in four weeks or six weeks or eight weeks—that kind of Zara model,” she says, referring to the fast fashion retailer. But even that turnaround time might be too long. “The business has to gauge, how fast could this change? By the time you get it, it may be over,” Van Dyck says.

She advises retailers to stay close to their customers, “whether it’s through studies, whether it’s through customer feedback, whether it’s through observing customers and interacting with them in the stores or online,” because their interest ultimately determines whether it makes sense to be involved with any given trend. “It’s really the relevance of what the trend is and how that relates to your customer,” she says. 

But even if the term “whimsymaxxxing” is new, Shively at Michaels says that “making things shiny has been, I would say, on an upward trend for the last two and half to three years.” In house, Michaels refers to it as “diamond art,” and Shively traces it back to Taylor Swift’s Eras Tour, which “changed our jewelery business in a big way.” The retailer followed the tour’s schedule to make sure nearby locations were fully stocked in the lead-up to a concert. (More recently, you can thank Kylie Jenner’s bejeweled “dream princess kitty bday night” for turbocharging the trend.)

But even though the Eras Tour is long over, Michaels is still replenishing its gems on a weekly basis. With school back in, Shively expects more gems to sell for game days and sorority events, and the holidays are just around the bend. “People always want to be glitzed out and glam for New Year’s,” she says. 

Expect Michaels to keep having gem stickers in “all the colors for people, all the different styles,” Shively says. “We really try and invest in that for the long haul, knowing that it’s our core proposition.”

Allie Joy, an art therapist and cohost of The Creative Collective Podcast, hopes whimsymaxxxing as a trend can be a gateway for playfulness and creativity, rather than a be-all and end-all.

“It feels like it’s being very narrowed down to just this, putting gems on things we can buy or single-use things,” Joy says. “I just hope it doesn’t turn people off of the concept of whimsy and be averted to whimsical things because it’s so much more than just the little stickers. . . . For me, whimsy is introducing some of the lightness back into our life, maybe embracing our inner child.”

So what is the Museum of Modern Art revealing about itself as an institution with its forthcoming exhibit Full Disclosure: The Edge of Information Design?

The pieces of information design in the exhibition “tell us the fact that the museum is always ready to understand what’s important to the world in that particular moment, and to try to give the best examples of it,” says Paola Antonelli, the senior curator of architecture and design at New York City’s MoMA, during a walkthrough of the exhibit as she and her team prepared it for its opening on September 27. “Having an exhibition about information design shows that we’re continuing our mission to talk about the art in our time, of our time.”

The first piece that visitors will see upon entrance through the museum is Warming Stripes, by climatologist Ed Hawkins, who “wanted to find a way to drill it in people’s brains that global warming is really happening.” The stripes, which over the course of the work go from a deep blue to a bright red, represent the average temperature of each year. The imagery resembles a gorgeous rug—and also a huge warning sign. 

To Antonelli, Warming Stripes is an ideal distillation of information design, because the field “takes phenomena big and small, public and intimate, complex, scientific, and statistically more superficial, and renders them in a way that is comprehensible by a wide audience.”

The exhibition shows that information design is not limited to weaving huge data sets into legibility. One portion of Full Disclosure focuses entirely on personal data, and includes a correspondence between artists Giorgia Lupi and Stefanie Posavec. In what became their Dear Data project, the two found creative ways to visualize how many times they cursed in a week, for example. The idea is that the practice of information design can help people connect in a more personal or even playful way with the many nuggets of data that have come to symbolize their lives. 

MoMA has been collecting works of information design since 2008, when Antonelli debuted an exhibition called Design and the Elastic Mind, which focused on the ways designers have adapted to revolutionary scientific change. As part of her curation, she encountered “all of these different designers that didn’t really have a home,” who worked with data visualization, she says. “I kind of hugged them up.” All but one of the pieces in this exhibition are part of MoMA’s collection.

Often, the origination of information design as a practice gets traced back to the invention of the bar chart or Dr. John Snow’s 19th-century map of cholera in London. But Antonelli sees it going back way further—it has always been a part of the human experience because “it’s all about explaining things in a narrative and visual way.” That means cave drawings displaying the haul of deer from a hunt fit the bill, too. 

But the use of computers and digital tools has exploded and expanded the practice of information design. And the exhibition reflects that. 

While many people encounter examples of information design on tiny phone screens, and there are ample opportunities to interact with screens in this exhibition, it also includes massive pieces that take up entire walls or, in the case of Kate Crawford and Vladan Joler’s Calculating Empires, one room. 

Newspapers and media companies often have teams focused on information design, and two works from The New York Times are included in the exhibition—one that uses 3-D rendering to explain why Paris’s Notre Dame cathedral was so vulnerable to fire and another that recreates the Greenwood neighborhood destroyed by the Tulsa race massacre. 

Antonelli sees the museum’s first-ever exhibition focused entirely on information design as a “gentle provocation,” or perhaps a “surprise” reminder that everyday works—say, those found in a newspaper—are museum-worthy.

She wants visitors to leave with a better sense of discernment for the barrage of infographics and diagrams they’ll encounter each day. She hopes they’ll use their own critical sensibility to see how the creator’s choice of color, typeface, and other formal decisions might impact how they perceive the information. And she’d like people who see these design examples to have a keener sense of appreciation for when something has been done “elegantly and efficiently.”

“We’re hoping that, as usual, they will look at their life with different eyes,” Antonelli says. “More critical eyes.”

Those matchbooks, which were a feature of gay bars throughout the mid-20th century, are now the subject of two dedicated books by the designer and editor Darian Newman. The first edition, titled Cruising: Gay Matchbooks, debuted in spring 2025 and sold out multiple times. Its success drove Newman to create a second book of the same name, this time featuring 200 new matchbook designs and 12 curated essays.

Newman says the matchbooks reflect a time when queer spaces had their own distinct visual identities designed with their local communities in mind. They’re also a form of ephemera that offers a rare window into moments of queer joy in an era that’s most often documented through stories of struggle and grief. 

“Design for marginalized communities is so often focused on protest,” Newman writes in the book’s introduction, “but the matchbooks exude pleasure. The pleasure of being out and going out, the pleasure of lighting a cigarette, the pleasure of getting a stranger’s number with the promise of fun.”

Before Grindr and Sniffies, there was the matchbook

“Manhole.” “Dude City.” “Mother Lode.” “The Mind Shaft.”

Each of these are names of clubs that Newman found embossed on vintage matchbooks, often in custom typography and accompanied by hand-drawn illustrations and photography. In an era when gay bars were often targeted by the police and patrons frequently relied on a level of discretion to go about their personal lives safely, matchbooks served as a form of queer bar and club advertising for those in the know. Their designs were a perfected art; each matchbook, despite its diminutive size, needed to clue prospective patrons into the establishment’s unique vibe and atmosphere. 

“Many bars lacked signage on their exteriors for discretion,” writes Lucas Hilderbrand, a professor of film and media studies at University of California, Irvine, in his contributed essay. Instead, he continues, they “fabulated vivid queer worlds” with double-entendre names, phallic logos, and graphics on matchbooks and event flyers. “Matchbooks and ads rarely visually represented the actuality of brick-and-mortar venues; instead, they produced an idea of each,” he writes. “The design of some matchbooks may be more striking, retrospectively, than their venues, which may be all but forgotten.”

To track down the paper ephemera from these forgotten brick-and-mortars, Newman scoured eBay and other secondhand sites, searched through hashtags on Instagram, and struck up a rapport with collectors in the San Francisco area. He’d often ask to purchase one matchbook, only to be offered 30 more, a box of DVDs, and a collection of posters.

“It was so cool, as an outsider to that period of time, to see these people who have archives of their own lived experience that people aren’t necessarily getting to see,” Newman says.

One of the matchbooks featured in this second collection is entirely covered in silver foil (Circle J, San Francisco); another shows a full-frontal nude on the inside cover (The Zone, Hollywood); and a third highlights an illustration of a ripped swimmer hoisting a giant swordfish above his head (Marlin Beach Hotel, Fort Lauderdale, Florida). Each aimed to entice curious onlookers to learn more in an era when the only way to do so was to show up and find out.

For older members of the gay community, Newman says, seeing these matchbooks often sparks memories that haven’t been revisited in years. And for those in younger generations, like himself, these tokens offer a physical window into a bygone era when, as the essayist David Kennerly puts it, “you had to get your ass off the couch to go out to seek companionship.”

If we ever hope to live on the red planet, we must build with the red dirt right beneath our boots, a challenge that has long been thought to demand expensive, huge industrial machinery and nuclear plant-level power generation. Now, a team of researchers has unveiled a radical, low-energy alternative in which a crew could carry a tiny tube of dormant yeast from Earth, breed trillions of helper cells in a vat using Martian carbon dioxide and melted ground ice, and 3D print entire outposts using modest solar power.

The new approach, led by Jishen Qiu at the Hong Kong University of Science and Technology and detailed in a study published in Chem Circularity, uses the freezing Martian climate to its advantage. Instead of melting rock, the team created what they call a “Martian living building material” by blending sand with a water-based binder made of pig-derived gelatin and genetically modified yeast. When extruded and exposed to simulated Martian surface conditions—temperatures averaging minus 85 degrees Fahrenheit and an atmosphere roughly a hundred times thinner than Earth’s—the material freezes and dehydrates on its own.

The physical results are competitive with conventional construction products. The composite achieves an average compressive strength of about 12 megapascals, placing it alongside standard lightweight concrete, while its flexural strength—how well a material resists snapping under bending pressure—reaches 6 megapascals.

That high flexural strength means the material resists bending and tension twice as well as typical lightweight concrete, a critical trait for surviving high-velocity debris hurled by Martian sandstorms. At the same time, producing 1 cubic meter (35 cubic feet) of the biocomposite demands tens of times less material processing energy than sintering, or baking mineral grains together with intense heat, running on less than an hour of standard solar power rather than days of continuous heating.

This solution is not without challenges and roadblocks, but it beats the alternatives so far. The leading blueprints for Martian construction have relied on brute force. Most proposals center on sintering, a technique that blasts loose Martian regolith—the powdery soil and broken rock covering the planet’s surface—with lasers, microwaves, or giant solar kilns to melt it into ceramic blocks. The drawback is severe: Melting rock requires temperatures exceeding 1,800 degrees Fahrenheit. Powering that kind of heat demands nuclear reactors or massive solar arrays running for weeks just to produce enough material for a basic outpost. The resulting ceramic is brittle, cracks easily under bending loads, and cannot be recycled once damaged. It’s extremely hard to do.

Other options include going underground or into mountains like Tolkien dwarves. These are also extremely expensive, requiring huge machinery and energy bills. That could be the way to go for the far future but—at least for the first waves of colonists—we need a lighter and more affordable solution.

Genetic Engineering Is the Key

The secret behind the material’s strength lies in genetic modification—turning ordinary baker’s yeast into living, microscopic construction workers. Natural yeast cells are simple, smooth spheres, but Qiu’s team modified their DNA to sprout three distinct tools on their outer skins.

First, they added AGA2, a surface protein that acts like an anchor to tether the cells directly to sand grains. Second, they engineered the cells to produce mussel foot proteins—synthetic replicas of the supersticky glue sea mussels use to grip wave-battered rocks underwater. Third, they added a paired protein system called SpyTag and SpyCatcher; like microscopic Velcro, the two parts snap together the instant they touch, forming permanent chemical bonds that lock neighboring yeast cells together into an unbroken web.

When that paste exits a 3D printer nozzle heated to a warm 98.6 degrees Fahrenheit—normal human body temperature—it hits the bitter cold of Mars. Normally, freezing water expands into large, dagger-like ice crystals that would tear the material apart. Instead, the protein coats on the yeast act as microscopic traffic cops, forcing the ice to freeze into an orderly grid of tiny, 5-micrometer pockets.

Because the Martian atmosphere is a razor-thin quasi vacuum, that ice never melts; it sublimates, evaporating directly from solid ice into gas. Once the moisture vanishes into thin air, it leaves behind a lightweight, sponge-like skeleton of animal gelatin, with the embedded yeast acting like steel rods inside concrete slabs to keep the walls from collapsing.

As a bonus, this is all recyclable. The researchers tested the cured bricks to the breaking point by crushing them under heavy hydraulic presses. But then they didn’t throw out the rubble. They gently warmed the debris back into a spreadable paste and 3D printed brand-new structural blocks through four consecutive generations. The recycled bricks suffered zero loss in compressive strength or stiffness, and the internal yeast cells emerged alive and kicking every single time.

Solvable Roadblocks

Despite the laboratory success, taking this technology out of the environmental chamber and onto the Martian surface presents hurdles that Qiu acknowledges need to be solved. The most immediate technical limitation is that the highest-performing formula still relies on porcine gelatin imported from Earth, he tells me in an email interview. To achieve true independence, the yeast must synthesize the entire structural scaffold on its own, cutting out animal products entirely.

When I asked Qiu why the yeast cannot yet replace pig gelatin, he explained that the issue isn’t whether the biological glue holds up. The glue sticks, but the cells do not yet churn out enough physical volume of protein to form the shock-absorbing lattice needed to distribute structural loads evenly across the sand.

Water management poses another operational dilemma. Because the bricks harden through open-air freeze-drying, moisture sublimates directly into the thin Martian atmosphere—a daunting trade-off in a world where every liter of water is vital for drinking and rocket propellant. To prevent that loss, Qiu suggests that the entire printing and curing process could move inside a lightweight air dome so the evaporated water can be trapped and recycled—though he doesn’t view water consumption as a dealbreaker, since vast reserves of surface ice are available in several regions of Mars.

Then there are the thorny issues of biology and planetary law. Because the engineered yeast survives the freeze-drying process fully viable, unleashing it on Mars directly bumps up against the planetary protection mandates of NASA and the international Committee on Space Research that are designed to keep alien environments free of terrestrial biological contamination.

The reality is that, once we put boots up there, the mandates are going out the hatch. The day living astronauts step out of a lander, trillions of skin and gut microbes arrive with them, making construction yeast just one piece of a much larger containment puzzle. As Qiu points out, “as long as human beings landed, there would be a good chance that the microbes in our body may be exposed to the atmosphere there.”

The final roadblock to using organic material to create Martian habitats is the brutal bath of unshielded cosmic radiation sweeping the red landscape. Prolonged radiation could mutate or destroy organisms, and Qiu acknowledges that they haven’t tested yet for the resistance of the yeast. Yet he is confident that synthetic biology already has an answer waiting in nature. “We are sure that some microbes can do it and can also be genetically engineered for similar purposes, like making mechanical proteins,” he says.

A lot has to happen between these experiments and actual colonization—like the spaceships to take us there, everything is still a sci-fi fantasy—but the research ultimately demonstrates that off-world construction does not have to rely on heavy industrial kilns. By substituting massive thermal energy with programmable cellular biology, future explorers might construct permanent shelters from little more than dirt, ice, and a tube of living organisms.

Within six months, across a 300-foot shoreline, the depth of the beach grew by 90 feet. Now, two years later, that same beach has doubled in size. Underwater, the artificial reef is filled with new coral and marine life, from fish to sea anemones.

Coastal Assembly, a startup that spun out of the MIT research at the end of last year and raised $4.4 million in a round of funding this summer, is now scaling up. The company uses AI to analyze satellite and ocean current data at a shoreline, then plans the layout of custom concrete blocks that can help sand accumulate naturally via ocean waves.

“We take 10 years of data, and we say, where do we predict the critical areas of erosion will be in the coming months to year?” says Skylar Tibbits, a professor in MIT’s architecture department who runs the university’s Self-Assembly Lab, a research group that explores how materials and environments can transform themselves. “And that tells me where the best place is to intervene—what’s the part that is most critical, where we should focus our time, effort, and money.”

In the case of the Maldives, the team installed 54 strategically placed blocks made from marine-grade concrete and local aggregate roughly 150 feet offshore. By dissipating wave energy, the blocks help redirect sand back to the beach that was previously lost, and also help keep the sand in place. The researchers track changes over time, and can adjust the blocks if needed—for example, if development nearby alters wave patterns.

The research started by chance, when Tibbits happened to visit a sandbar in the Maldives that had emerged naturally in only a month. “It was like, wait, what do you mean, this island formed in a month? We started to understand that we could tap into this process, and then it became a question of how you promote that to happen,” he says.

Tibbits says he learned that traditional coastal protection methods don’t work well. Dredging sand harms coral reefs, and also only works temporarily, as the sand can quickly wash away. “It’s a continual process,” he says. “One of the resorts we visited is pumping sand 365 days. . . . It’s the sort of thing that everyone does as a Band-Aid.”

Building seawalls is expensive, and also isn’t necessarily very effective. As Tibbits points out, in a fight with nature, nature tends to win. And a seawall can end up accelerating erosion on other beaches nearby, or even unintentionally increasing erosion at the beach it was built to protect.

In the lab, it took nearly a decade for Tibbits and his team to find an approach that worked, as they moved from simulations to 10 real-world deployments in the Maldives. Now AI makes it easier to analyze the data from any particular beach.

Each design is custom, from how the structures are arranged to their height and depth. “It’s not one-size-fits-all because every beach has a different bathymetry [underwater landscape],” Tibbits explains. “It has different wave conditions. There are always different complex parameters that make each beach unique.”

Each structure is also cast with a unique shape so the reef looks more natural than human-made. The artificial reefs, like natural reefs, also help protect coastlines during storms.

The startup is set up as a public benefit corporation, meaning that it pursues a mission along with profit. On the mission side, it works with vulnerable communities to analyze data about how their coastlines are changing, and works with partners at the United Nations and other organizations to help deploy its reefs in those areas.

On the business side, it works with coastal resorts that want to protect their private beaches. “Resorts are spending millions of dollars annually on coastal protection for things that don’t work,” Tibbits says. The revenue from the commercial projects helps the company build its technology, team, and manufacturing capacity more quickly.

Coastal Assembly is expanding its work in the Maldives and beginning to work in other locations in need of coastal protection, including Boston and Miami. Cities are recognizing that traditional approaches, like seawalls, aren’t the best answer. “You’re seeing a big shift around the world, Tibbits says, “from hard edges to much softer, more flexible, dynamic edges.”

“Having all these drones flying around is actually a big, big problem,” the president of a homeowners association in an affected neighborhood told The New York Times.

This should be no surprise: Technology innovation has long met with mixed responses. But lately, tech is sparking a lot of flat-out divisiveness. Aside from delivery drones, Meta’s smartglasses (or “pervert glasses,” as critics call them) and Flock license-plate readers are among the tech products that have inspired heated opposition, sometimes taking the form of physical destruction.

And of course arguments over the pros and cons of AI—from pushback on data centers to alarm over its effect on education and human creativity, among other issues—have dominated cultural discourse for a couple of years now.

Every innovation has critics, skeptics, and outright opponents. And in some ways, the entire internet era has been one long debate about digital innovations and their potential repercussions.

Once upon a time, for instance, Apple AirPods were controversial. Unveiled in 2016, they forced a shift from wired headphones that nobody asked for, cost an absurd $160, and looked ridiculous. But evidently, most of us got over that, as Bluetooth headphones are now routine.

Similarly, the initial iPod was mocked by some as an overpriced indulgence from a company that had lost its way; it became a hit product that paved the way for the world-changing iPhone.

Arguably every cutting-edge technology must overcome initial resistance. That’s certainly the inference from the famous “Innovation-Adoption Curve” chart drawn from Ohio State University professor Everett Rogers’s 1962 book Diffusion of Innovations. This is where we find the now-familiar idea of “early adopters” as, basically, the first people to buy AirPods (or Meta glasses, or drone delivery service) and tell everybody how great they are.

Traditionally, non-buyers of AirPods (or whatever) enter the story as grumpy change resisters serving as a foil to early adopters. Treated as recalcitrant followers, these “laggards” belatedly adopt this or that innovation only when they become embarrassed by being behind the times.

We may well look back on today’s controversial innovations as playing out these patterns yet again, but what feels different about the current wave of divisive innovation is that the skeptics are strikingly vocal. There probably weren’t a lot of city council meetings, let alone congressional hearings, about the iPod’s effect on neighborhood quality of life or the blurring of public and private space.

In other words, what makes the current wave of divisive tech notable is that some percentage of consumers in the early adopter phase seems to be not so much passively lagging as looking to actively argue.

A key element of the diffusion of innovation theory is how conversation spreads new ideas across the culture much more effectively than any top-down method. And what the current moment of divisive tech is teaching us is that even negative chatter can help pinpoint fixable weaknesses. Meta has already tweaked its smartglasses, Flock is scrambling to address criticisms, and Amazon and other drone delivery services will surely learn from their exposure to real-world customers (and their neighbors).

The innovation-adoption curve is not inevitable. Plenty of self-proclaimed innovations stall out, failing to capture and hold consumer attention. But there’s no danger of the conversations around today’s divisive tech fizzling out. Of course the creators of these aspiring innovations would rather the conversation be purely enthusiastic.

But even arguments spread the word. Sometimes it’s better to be divisive than ignored.  

According to a recent report from the UN, the world is on course to cross the average warming level of 1.5 degree Celsius (2.7 Fahrenheit) in the next few years. That’s considered a tipping point by the 2016 Paris Agreement, which called for net zero global carbon emissions by 2050.

As dire as the headlines are, there are also reasons to be optimistic about addressing climate change. Dozens of them are listed in Building Postcarbon Futures: Land, Justice, and Energy Transitions, a new book by the landscape architect Billy Fleming. It turns out that we don’t have to look far for equitable, just, and environmentally sensitive solutions to decarbonization. 

Fleming is an assistant professor of landscape architecture at Temple University and an advisory board member of the Climate and Community Institute, and was a policy advisor to the Obama administration. In 26 case studies spanning 43 countries and six continents, he walks readers through proven ways that communities are addressing the challenge. 

The atlas’s examples aren’t simply swapping conventional energy for green options; they challenge conventional ownership structures and what types of knowledge and innovation are prized. It isn’t always a race to the newest technique that matters, as demonstrated by the Zapatista movement’s milpa farming system in Chiapas, Mexico, which relies on techniques that date back 2,000 years. It also isn’t necessarily good design that is the most important outcome, as Lacaton & Vassal’s renovation of public housing in Bordeaux, France, demonstrated; the architects found ways to complete building retrofits, like new balconies and energy-efficient systems, without displacing residents. 

We spoke with Fleming about the book, which he hopes becomes a reference for designers, policymakers, and everyday readers who are interested in learning more about solutions to the climate crisis.

This interview has been edited and condensed.

The book arrives at an interesting time. Year after year, we’ve faced more extreme environmental conditions. Here in New York, we’re now classified as a subtropical climate. How are you feeling these days about the movement toward a postcarbon future?

This book was a salve to work on for the last five or six years because, as bleak as it is today, it’s been like that for a long time in the environmental and climate justice movement. The unfortunate nature of doing that kind of work is we just often lose most of the things we’re fighting for. And when we win, we celebrate.

In some ways, the book is growing out of a frustration with the way my fields—designers broadly, planners, climate policy folks—tend to think about what a postcarbon future can or should look like, or what climate resilience actually is. Certainly in landscape architecture, architecture, and urban planning, the projects that tend to get valorized or foregrounded as exemplars of this kind of work are often really just ecological gentrification projects masquerading as some kind of huge win on climate action. 

How did that shape the book’s content?

I wanted to look outside of our profession at people who are practicing architecture, landscape architecture, and urban planning, just without credentials. And to see what we might learn from places that have already started to figure some of this out, some at very small scale, some at very large scale. They’re building these imperfect works of climate justice that kind of show us what it can look like to flourish in a warmer, wetter world and have managed to do it, in some cases, in far harsher conditions than we live with in Philly or New York.

Can you share a little bit more about what a postcarbon future looks like to you? Reading the book, I was interested in how the examples included swimmable urban rivers, beautiful affordable housing, and even architectural salvage. Also, nothing was speculative.

I very intentionally did not want to dip into the realm of climate fiction, or design fiction more broadly. Not because I’m uninterested in it. I’m deeply interested in that and have my students working in those modes often in studios. But the intention of the book is we don’t have to turn to fiction to figure out what this can or perhaps even should look like. There are real existing exemplary works of postcarbon futures and climate justice all around us. We just don’t tend to look at them because they aren’t authored by big firms.

I think we can do things better. All of these projects are doing things better if imperfectly. They’re focused on how people flourish in a world that is hotter and wetter and more volatile. That begins, in some cases, with having stable, secure, and actually beautiful housing with lots of amenities connected to it. It sometimes includes owning and controlling and perhaps profiting from control of your own energy system, like the [Tu Deh-Kah geothermal energy plant in British Columbia owned by the] Fort Nelson First Nation. Sometimes it means operating a petrochemical-free agricultural or agroecology system that you developed because the U.S. and the West blockaded your country and tried to starve you to death. [Editor’s note: He’s referring to the organopónicos urban farming program in Cuba.] And now you’ve built enough power through food sovereignty that you can think about other things. 

All of those are integral to a postcarbon future and climate justice. They just don’t always look like a piece of shiny technology or CLT [cross-laminated timber] being specced on a building.

Speaking of justice: Your thesis couples decarbonization with equity. Why was that interconnectedness important for you to foreground?

For a lot of reasons, one of which is this is not the first energy transition. There have been many, from steam, coal, gas, oil, and so on. This is one of many. There will be others after it. And in each of those, at least in the U.S., we’ve kind of reproduced or repeated the same pattern, which is we send tons of public subsidies and lubricate markets in order to create new private wealth that was made possible through mass public investment. Those things have made the planet worse, both ecologically and socially.

There are many, many other models for how we can invest in places and invest in people and invest in technologies and infrastructure without corroding society in the process. And so one of the ways that this book tries to kind of foreground what another path could look like is by showing projects where people actually own their energy infrastructure. They own or have real governance over their housing system. This idea of shared ownership is kind of implied. It’s not as overt as some other messages in the book, but the ability to control, own, profit from, and reinvest the proceeds of public investments in the energy infrastructure at a collective or cooperative scale is maybe the single most important thing in all of this.

There have been so many large-scale climate setbacks, especially involving the United States, which is the largest carbon contributor in history. Within the context of growing international conflicts, exiting the Paris Agreement, and really hitting the gas on data center construction, I did find reasons to be optimistic in some of these smaller-scale case studies. How do you square those two different scales of impact?

Windows always open and close, and local experiments are always where I think the best ideas get piloted. They get scaled in some cases to tremendous degrees. And then when a window for national or multinational transformation happens, we need lessons to be able to draw from quickly. That’s how I imagine many of these projects in the book hopefully working. These are the building blocks of what a future postcarbon or climate justice program could draw upon the next time we get a chance to push things through.

This is what many of us were doing in 2018, 2017 around the Green New Deal during Trump I. What is the program we can push forward the next time we get a window to do so? I don’t want to pretend like that was wildly successful, but there were a number of provisions in both Biden’s IRA [Inflation Reduction Act, promoting clean energy] and a number of other state- and local-level programs that came directly out of that work. 

One, which is still growing, is [Pennsylvania state senator] Nikil Saval’s Whole-Home Repairs Act, which we got to work on with him. It’s incredible. It retrains low-income workers to do deep weatherization and deep energy retrofits in low-income communities throughout Pennsylvania. It’s been wildly successful. Its funding has increased every year since it’s been on the books. Now it just got a national pilot in the big housing bill that just passed a month or two ago.

That work comes out of this very local experiment that Saval, somehow, got an oil-and-gas-owned Republican legislature to sign off on in Pennsylvania and now has a chance to put a bunch of people who need good jobs back to work in their communities doing the kind of unsexy deep weatherization or deep energy retrofit work in communities that would otherwise never get.

I wanted to go back to some of your criticism of the design profession. Where do you see capital-D designers within the context of postcarbon futures? How can they best use their power given the prevailing economic, political, and social systems?

I don’t want to pretend like it’s easy to run a firm and to try and make payroll and keep the lights on and do all of the things that your increasingly radical workforce is asking you to do. 

But I think there’s a very real risk for landscape architecture, in particular—but we could say architecture and urban planning as well—driving a generation of very talented and very committed workers out of the field because they find themselves in offices where they expect to work on the kinds of things maybe that are foregrounded in this book and they find themselves working primarily on, in a best case scenario, maybe beautiful things for rich people. And in a worst case scenario, pipelines, prisons, and all of the other things that actually float the A&E [architecture and engineering] industry.

The most interesting practices to me are finding ways to do low-carbon, capacity building work. They’re places like [the Los Angeles–based landscape firm] Terremoto who are looking at how we treat investments in gardening and in small-scale landscape interventions as a way to pilot both different types of approaches to land stewardship as a way to build ecological knowledge amongst a larger group of people, and as a way to build social cohesion. Ruderal, Sarah Cowles’s practice in Georgia, and Wagon, in France, or other landscape offices do this work really well.

The whole idea is we do small moves, we work in very low-capital intensive ways with what’s on site, and we treat this as a place for people and ecological knowledge to be developed and for capacity to grow over time.

If I were graduating and thinking about where I would want to work, I’d want to go to a place-based or regional-based office with deep commitments to particular groups and communities and people who want to build slow projects over time that aren’t taking them into the same capital chasing that McKinsey and AECOM do.

Lastly, who do you hope reads the atlas and what do you hope they take away from it?

The wrong answer is probably everyone. I intentionally thought of the atlas and the case study approach to the book as useful for practitioners who are looking for an alternative set of references. I hope it’s useful to people in the climate policy world who are often thinking and talking in really abstract ways about how to decarbonize, how to build social justice, but in ways that don’t feel tangible or visceral or real. I think the most successful version of climate policy that I would expect people to maybe get excited about is one that they can touch and they can feel and they can see themselves and the people they love within.

Building Postcarbon Futures: Land, Justice, and Energy Transitions is written by Billy Fleming, edited by Jennifer Sigler, designed by Cristian Ordonez, and published by Lincoln Institute of Land Policy.

Meta CEO Mark Zuckerberg announced the company’s forthcoming Muse Charm on Wednesday, a handheld device for its new AI agent. It looks like a futuristic Tamagotchi and personifies the chatbot on the screen as a fuzzy, rotund beige mascot. You can name the creature yourself. Instagram’s Adam Mosseri named his “Max.” (Facebook and Instagram introduced their users to the avatar this week to entice downloads of the Muse app.)

“We packed the whole Muse experience, including the whole real-time voice and avatar stack, into something that fits on a keychain and is always available to talk to,” Zuckerberg said at Meta Connect.

Instead of visualizing its new AI agent as non-anthropromorphic glowing orb, like Siri, or a cursor and text, like ChatGPT, Muse is a cute monster avatar. Meta could have gone down the route of its competitors, but it didn’t. Instead its choice shows something about how big tech is appealing to softer emotions in the race toward mass adoption.

The Muse agent can read emails, book classes, draft documents, and make purchases (just not on Amazon, which blocked it). That cross-platform functionality means allowing Muse access to a lot of accounts, which requires trust—something Meta and social media companies at large struggle with. If Muse and the Muse Charm are successful, however, it could prove that a mascot-led brand approach does a better job countering consumer AI fears than a disembodied voice or orb.

The cuteness factor has shown up in transportation recently with vehicles like Robotaxi Zoox, Amazon delivery vans, and the DoorDash Dot delivery robot. The vehicles are rounded in shape and designed to look non-intimidating. The trend is towards taking technology that might otherwise be seen as threatening, like autonomous vehicles or AI, and giving a friendly, charming exterior. Cuteness, research has shown, also sparks a caregiving response—quite the opposite from the feelings of fear and concern that many Americans have about AI.

While you can use Muse on a smartphone, Zuckerberg explained the Muse Charm as something you use without having to unlock a phone or open an app. It’s a fun novelty that could find early adopters among people looking for a device that reduces their screentime. Zuckerberg said materials and components are being finalized for the Muse Charm and there isn’t yet a price tag for the device, but the company plans to be ready to ship in time for the holidays.

Meta’s stock has risen with its Muse release, but it’s unclear what demand might be like for the Muse Charm and whether there’s mass consumer demand for a productivity product like this before it’s caught on first among early adopters and nerds.

Tech companies are racing to design AI hardware like smart speakers and glasses, and Meta’s initial foray into AI-connected hardware with smartglasses faced has barriers to widespread adoption over privacy concerns that wearers can secretly record others using camera-equipped glasses. The Muse Charm, along with a next generation of smartglasses the company announced this week that are camera free, shows companies are exploring different devices that don’t have the same creep factor.

By personifying AI as a cute mascot that lives in a small, squircle-shaped device, Meta hopes to take the edge off a technology that otherwise has consumers scared.

As anxiety grows over smart eyewear that can record people without their knowledge, the ZuckOff app works by logging nearby Bluetooth signals and flagging those it believes come from smart glasses. The app has limitations—still, it represents a clever bit of countersurveillance.

Made by Polish developer Paweł Szydłowski, ZuckOff recognizes the manufacturer signatures of Ray-Ban Meta, Oakley Meta, and Snap Spectacles with high certainty, but can also spot Bluetooth-enabled devices like HeyCyan, Apple Vision Pro, and VistaView at a lower confidence level. Szydłowski is actively seeking submissions for other models to test as well.

The app works best when smart glasses are out of their case, powered on, and paired with a device, according to the ZuckOff site. Once you open the app, its signal strength feature tells you the approximate distance of a Bluetooth signal, but not the direction. It’s essentially like a metal or radar detector for smart glasses, merely alerting you when a signal is nearby. It’s not a GPS.

The app can send lock-screen notifications, and the in-app user interface displays the brand, quantity, and detection confidence level (possible, likely, or strong) for any smart glasses in the room. ZuckOff is designed without accounts, analytics, or location tracking.

It’s not foolproof. There’s still a chance that smart glasses could be on and recording without the app detecting them, the site notes. Plus, simply recognizing a Meta or Snap signal nearby doesn’t guarantee a device is actively filming.

The app is part of a growing design trend in response to pervasive surveillance. Adversarial clothing design uses patterns meant to confuse AI surveillance cameras, while other creations are less serious. In August, the search engine DuckDuckGo started selling satirical “Normal F***ing Sunglasses”—$35 glasses made by Knockaround that don’t have any cameras or microphones. They might have been a joke, but Meta is reportedly working on a new generation of smart glasses without a camera, according to The Information, suggesting that consumer privacy concerns are driving real-world design decisions.

For people worried about the panopticon created by smart glasses, ZuckOff could be a new line of defense.

This approach is on display at the company’s newest clinic, now open in New York City’s Soho neighborhood. With gently lit rooms, spare walls, and soft splashes of pastels, the clinic and its waiting area have a minimal and futuristic feel that blends a high-end spa with a fashion boutique.

Neko has been operating clinics since 2003, and has performed more than 100,000 of its $499 radiation-free scans. The company, which just pulled $700 million in series C fundraising, is looking at the New York clinic as its first step into the U.S. market, with other planned locations in Miami, San Francisco, and Washington, D.C.

The design of the clinic was led by Franquibel Lima, an architect and former global head of WeWork’s design studios, who has been involved in designing Neko’s other clinics in cities like Stockholm, London, and Manchester, England. He says Neko’s clinics are different from traditional healthcare environments partly because they’re not geared around treating sick patients, but also because the company sees more value in treating people before they become sick patients. “For us to move from reactive healthcare into proactive healthcare, it requires a real mindset shift,” Lima says.

Convincing people there can be another way of approaching healthcare, Neko has tried to make its clinics feel more like an experience than a cattle call. “If you think about traditional healthcare, it’s typically cluttered. You have posters on the wall like here’s your skeleton, and it’s all about the average person,” Lima says. “But we now have the technology to be able to be very personal and very individual.”

Drawing inspiration from Apple stores and science fiction films, Neko’s New York clinic combines carefully controlled lighting with a palette of soft and natural materials. At the street level, the storefront is a simple door emanating an amber glow. Inside, you walk down a brushed stainless steel stairwell into a white-walled waiting area, where a small greeting podium leads to a plush U-shaped seating area. Fluted wall coverings with backlighting give a cloudlike feel to the space, and underlighting beneath the seats adds to the sense of floating.

Leaving the waiting area and heading into the all-white clinical area, the lighting brightens and guides visitors into one of the six scanning areas. Through their doorways, pastel yellow lights and paint emerge from the scanning room, emitting what Lima says is a calming tone. “We look at lighting not as an overlay to materials and finishes. We look at lighting as almost like a building component,” he says. “It’s a multisensorial experience that really prepares you to know more about your health.”

As a premium service that’s intended to be complementary to primary care and other traditional health care services, Neko can afford to put these extra design touches into its clinics. But the company has ambitions of becoming a bigger part of the way people stay healthy and catch problems before they snowball. Creating spaces people will want to visit again and again is integral to that approach.

“Because we are a consumer brand, because we believe that consumers are the ones that drive change typically, instead of trying to do it through policy, then it’s really important that we create environments where people walk into and think this doesn’t feel like anything else that I felt before,” Lima says.