The First Listen

The initial watch was genuinely impressive. (and I’m going to write the rest of this with the assumption that you have seen it.) Musk laid out a vision of orbital data centers powered by solar arrays five times more efficient than anything on Earth. AI computing is moving to space within 30 months. Robots building robots in an exponential manufacturing loop, he called an “infinite money glitch.” Starships launching once an hour, ferrying more compute into orbit annually than everything humanity has ever built on the ground.

It’s breathtaking. Say what you want about the man, but nobody else on Earth is connecting these dots (rockets, AI, manufacturing, energy) into a single integrated vision. When he talks about the bottleneck being turbine blade casting and electrical transformer backlogs, you realize he actually understands the hardware problems that most tech CEOs have never thought about.

I finished the first listen thinking, “This is the most important conversation about technology I’ve heard in years.”

The Second Listen

Then I let it stew overnight. And on the second pass, something shifted.

The vision was still there, but I started hearing something else underneath it. When Musk was pushed on AI alignment — how do you keep something vastly smarter than you under control? — Musk’s answer was essentially: a curious AI would find humans interesting enough to keep around. That’s not a safety plan. That’s a hope dressed up as logic. It’s Jurassic Park’s John Hammond saying the dinosaurs won’t escape because the engineers thought of everything.

When he talked about building his own chip fabs, he said, “I don’t know how to build a fab yet. I’ll figure it out.” Decades of institutional knowledge at TSMC, waved away in half a sentence. When pressed on timelines, every answer was 18 to 36 months. Digital human emulation? End of 2026. Orbital data centers cheaper than Earth? 30 months. A million Optimus robots? Just a few years out.

I’ve heard these cadences before. From him, and from others.

Spared No Expense

Most people know Jurassic Park from the movie. The book is darker and more interesting. Crichton wasn’t writing an adventure story. He was writing about what happens when scientific capability outruns the question of why.

In the novel, Ian Malcolm doesn’t just quip about chaos theory. Crichton structures each section around iterations of a dragon curve, a fractal that grows more complex with each pass, mirroring the park’s collapse. Malcolm delivers extended monologues from his hospital bed about the difference between science and discovery. He argues that the scientific establishment has no discipline, no self-restraint. It’s so focused on whether something can be done that the question of whether it should be done never even enters the room. Hammond isn’t a lovable grandfather in the book. He’s a businessman who never once asks why the world needs a dinosaur theme park. He just knows he can build one, and that’s enough for him. In the book, it kills him (spoiler alert! but it’s a book from 35 years ago).

That’s what I kept hearing on the second listen. Musk has the can. Orbital AI. Robots building robots. A new species of intelligence. But when Dwarkesh pressed him on why we should trust that superintelligent AI will keep us around, the answer was essentially: because we’d be interesting to it. Which is just another way of saying “trust me.” It’s Hammond insisting on the lysine contingency. Engineering the dinosaurs to be dependent on a supplement only the park could provide, so they’d die if they ever escaped, and this will keep them from surviving in the wild. A clever-sounding safeguard that collapses the moment reality stops cooperating with the model.

Crichton’s whole point was that “because we can” is not a reason. It’s the absence of one.

The Tool and the Why

Look, I’m not an REALLY an AI skeptic (maybe a little). AI has made me a lot more productive. I use it every day. It helps me write, research, and build things faster than I ever could alone. It’s a real tool that delivers real value.

But a tool only matters in the context of what you’re using it for. A hammer is productive if you’re building a house for your family. If you don’t know what you’re building, you’re just a guy holding a hammer.

Musk talked about “trillions of dollars of revenue” unlocked by digital human emulation. xAI is raising money at staggering valuations. OpenAI, Anthropic, and Google are in an arms race for market dominance over what they say will be the most transformative technology in human history. You don’t suppose there’s some incentive to hype that up, do you?

Productivity for what? Revenue for whom? When the people building the thing are also the ones telling you how world-changing the thing is, you have to ask what the why is. And if the answer is “trillions of dollars,” that’s not a why. It’s a price tag.

I’ve Seen This Movie Before

I’ve been through more hype cycles than I can count. The internet was going to eliminate geography. Dot-com companies were going to replace every brick-and-mortar business. Social media was going to democratize truth. Crypto was going to replace banks. VR was going to replace offices. The metaverse was going to replace reality.

And in each case, something real did happen, but it looked nothing like the hype predicted. The internet didn’t eliminate geography; it made us more aware of it. Social media didn’t democratize truth; it fractured it. Crypto didn’t replace banks; it created a new casino.

The fundamental changes snuck in sideways. Nobody predicted that the smartphone’s real impact would be turning every human into a photographer with a dopamine-delivery device in their pocket. Nobody predicted that the internet’s killer app would be arguing with strangers.

Walk into a house today and compare it to a house in the 1980s. Other than the screens — TVs, phones, tablets, the Ring doorbell — it’s remarkably similar. Same basic appliances. Same plumbing. Same HVAC. Same lawn out front. Human life at the physical level is stubbornly resistant to the disruptions that tech visionaries keep promising.

That doesn’t mean nothing changes. Jobs change. Industries change. Human behavior changes in ways that are sometimes profound and sometimes trivial. But the sweeping, civilization-redefining transformation that every hype cycle promises? It never quite arrives on schedule, and when it does arrive, it looks nothing like the pitch deck.

The Man in the Arena

I don’t want to be too cynical about this. Musk has earned the right to make big claims. He built reusable rockets when the entire aerospace industry said it was impossible. He scaled electric vehicles when every automaker was dragging its feet. SpaceX and Tesla are genuine, world-changing accomplishments built by a person who outworked and out-risked everyone around him.

But there’s something in this interview that goes beyond engineering ambition. The simulation theory tangents. The idea that consciousness might be the universe optimizing for something. The framing of AI not as a tool but as a successor species that we should hope finds us interesting enough to tolerate.

It sounds less like a business plan and more like a man searching for meaning.

Man’s Search for Meaning

Viktor Frankl survived Auschwitz and wrote a book about what he learned there. His central insight was that humans can endure almost anything if they have a why, a reason to keep going. But Frankl also warned about what happens when meaning is manufactured rather than discovered. When people can’t find genuine purpose, they fill the void with power, pleasure, or the frantic pursuit of achievement. He called it the “existential vacuum.”

Watching Musk in this interview, I kept thinking about that vacuum. Here’s a man who has solved problems that seemed unsolvable. Reusable rockets, done. Mass-market electric cars, done. The hardest engineering challenges of our era have seemingly been conquered. And now he’s reaching for orbital AI, humanoid robots, consciousness as a cosmic optimization function, simulation theory, and building a species smarter than us that he hopes will find us charming enough to keep around.

At some point, the engineering ambition starts to look like something else. Frankl would have recognized it. When you’ve built everything there is to build, what’s left? You build meaning itself. You build god. Not because the world needs it, but because you need a reason to get up in the morning that matches the scale of what you’ve already done.

The tragedy of Hammond wasn’t that he built the park. It was that he needed to build the park. The flea circus could work. People would LOVE it. But it wasn’t enough. It wasn’t real enough. He needed living, breathing dinosaurs. The world wasn’t asking for them. The flea circus had just stopped filling whatever hole was inside him.

Frankl might say that meaning can’t be engineered. It’s found in responsibility to other people, in love, in how you face suffering you can’t avoid. It’s small, specific, and unglamorous. It doesn’t scale. It definitely doesn’t fit on a pitch deck.

I know what my meaning is. It’s my family. My wife, my kids. That’s it. And having that answer makes everything else simple: what I work on, how I spend my time, what I’m willing to sacrifice, and what I’m not. I don’t need to build orbital data centers or solve consciousness to know why I got up this morning. Most people don’t. Most people already have their why sitting across from them at the dinner table.

And when the richest person on Earth is searching for meaning by building god, the rest of us should probably pay attention. What matters isn’t whether he succeeds. It’s what happens while he tries.

Where I Land

The interview is worth listening to. Musk is brilliant, and his ability to think across domains: energy, manufacturing, software, and aerospace is unmatched. The orbital compute thesis might even be right in the long run.

But maybe listen to it twice. The first time, let yourself be impressed. The second time, count the assumptions. Count how many answers are “18 to 36 months.” Listen for the alignment question getting answered with philosophy instead of engineering. And watch trillion-dollar revenue projections get tossed around like they’re inevitable.

And then go home and look at your house. It looks a lot like it did in 1990. The screens got bigger and flatter. You’ve got a robot vacuum bumping into chair legs. But the walls are still drywall, the roof is still shingles, and you still mow the lawn on Saturday.

The future always arrives. It just never looks like the pitch.

Did you listen to the Dwarkesh interview? What was your take — visionary roadmap or the world’s most expensive TED talk? Let me know in the comments.

Sources and Further Reading

• Dwarkesh Patel: Elon Musk Interview (February 2026)
• Dwarkesh Patel: Interview on YouTube
• Michael Crichton: Jurassic Park (novel, 1990)
• Viktor Frankl: Man’s Search for Meaning (1946)

But I’m not excited. And I’ve been trying to figure out why.

A Rocket Without a Mission

The Space Launch System wasn’t built to go somewhere. It was built to preserve jobs. That’s not cynicism—that’s the actual legislative history. When the Obama administration cancelled Constellation, Congress mandated a new heavy-lift rocket not because we needed one for a specific mission, but because NASA centers employed thousands of people who would otherwise lose their jobs.

We assembled a rocket from Space Shuttle parts and then spent years trying to figure out what to do with it. It looks like the kind of project you’d expect when you need to ensure that every member of Congress provides jobs in their district.

That’s backwards. Apollo worked because Kennedy set a goal and NASA built hardware to achieve it. Artemis has hardware searching for a purpose.

The “Moon to Mars” Treadmill

I’ve been hearing “Moon to Mars” my entire adult life. In 1969, NASA pitched to Nixon the idea of reaching Mars by 1981. He approved only the Shuttle. Bush 41 proposed a 30-year plan that died in 15 months. Bush 43 gave us Constellation. Now we have Artemis.

Every decade or so, a president announces we’re going to Mars. The price tag comes out. Congress balks. The program is scaled back, extended, or cancelled. Rinse and repeat.

I want to believe this time is different. But the pattern is hard to ignore.

Less Ambitious Than 1968

Artemis II is technically less ambitious than Apollo 8.

Apollo 8 went into lunar orbit. The crew performed engine burns that, if they failed, would have stranded them forever. It was a genuine test of the hardware that would take us to the surface.

Artemis II? A free-return trajectory. A loop around the Moon and back. It’s the mission profile NASA rejected in the 1960s as not worth the time or money.

And the SLS/Orion stack isn’t even the vehicle that will land astronauts on the Moon. That’s being contracted out to SpaceX and Blue Origin. What exactly are we demonstrating with this flight?

Yes, Orion has a glass cockpit and a modern flight computer. The life support is better. The heat shield is rated for faster reentry speeds. But none of that changes what the mission actually does. Apollo 8 committed to lunar orbit—a maneuver that required the crew to fire their engine on the far side of the Moon, out of radio contact, with no abort option. If that engine didn’t light, they were dead. Artemis II never enters orbit. The Moon’s gravity does most of the work. The crew loops around and comes home on a trajectory that’s essentially self-correcting. Fifty-seven years of progress, and we’re flying a less demanding mission profile.

The Privatization Question

I’m pro-commercial space. SpaceX has accomplished what NASA couldn’t—or wouldn’t—for decades. Reusable rockets. Dramatically lower launch costs. Rapid iteration. Starship is genuinely exciting in a way SLS never will be.

If NASA had simply said, “We’re getting out of the rocket business and letting SpaceX handle launch,” I’d be fine with that. Let commercial companies do what they do best.

But that’s not what’s happening. We’re spending billions on SLS—a rocket built from Shuttle parts, designed by Congress to preserve jobs—while also contracting out the actual lunar landers to SpaceX and Blue Origin. We’re getting the worst of both worlds: government pork and privatization, with neither done well.

Meanwhile, the 2026 budget proposal cuts $2.265 billion from space science. Mars Sample Return is cancelled. STEM outreach was eliminated entirely. The science that only NASA can do—the stuff with no commercial market—gets gutted to fund a jobs program disguised as exploration.

Commercial space is the future. But Artemis isn’t a thoughtful transition to that future. It’s a muddled compromise that satisfies no one.

The Numbers Don’t Lie

It all comes down to the engines.

The RS-25 that powers SLS is a 50-year-old design—literally the Space Shuttle Main Engine. NASA’s total contract for 24 new-production RS-25 engines comes to roughly $3.5 billion, or about $145 million per engine. Each SLS launch uses four of them. They’re expendable now, thrown into the ocean after one use. That’s $580 million in engines alone, discarded every flight.

The NASA Inspector General found that engine and booster contracts originally expected to cost $7 billion over 14 years have ballooned to at least $13.1 billion over 25 years.

SpaceX’s Raptor 3? In March 2025, Musk stated that the Raptor 3 costs “about four times less” than the original Raptor, which would put the per-engine cost at around $250,000. That’s not a typo. The Raptor 3 actually produces more thrust than the RS-25—about 617,000 pounds-force at sea level versus the RS-25’s 418,000—and costs roughly 580 times less.

For the price of the four RS-25 engines on a single SLS flight, SpaceX could build over 2,300 Raptor 3s. Enough to power a fleet of 59 Starship vehicles.

Launch Costs

The Trump administration’s FY2026 budget proposal called SLS “grossly expensive” at $4 billion per launch and proposed to phase it out after three flights. The NASA OIG called plans for a 50% cost reduction “highly unrealistic” and projected costs will remain above $2 billion per vehicle through the first 10 launches.

At current prices, you could fly 45 Starship missions for the cost of one SLS launch. Once Starship achieves full reusability, that ratio could climb to 400:1.

How Did SpaceX Do It?

The cost gap comes down to fundamentally different ways of building rockets.

Vertical integration. SpaceX builds almost everything in-house. No byzantine contractor networks. No cost-plus contracts incentivizing inefficiency.

High-rate manufacturing. At peak Raptor 2 production, SpaceX was building more than one engine per day—a rate confirmed by NASA officials. Even as the newer Raptor 3 ramps up production, SpaceX has already built over 75 of them. The RS-25 production line was shut down for years and had to be restarted at a high cost. Low-rate production means high per-unit costs.

Modern manufacturing. Forty percent of early Raptor components were 3D printed—turbopumps, injectors, parts that would traditionally require months of machining. This accelerates iteration and slashes costs.

Design for reusability. Starship is designed from the ground up to be fully reusable. SpaceX is targeting 1,000 flights per engine. SLS throws everything away except the Orion capsule.

Iterative development. SpaceX builds, tests, breaks, learns, repeats. They’ve blown up multiple Starships—and each failure teaches them something. NASA’s approach requires every component to work perfectly the first time, which means endless analysis, reviews, and delays.

The RS-25 is a genuinely magnificent piece of engineering. It’s also a museum piece being produced at artisanal rates for government-contract prices. Meanwhile, SpaceX is building rocket engines like Henry Ford built Model Ts.

However you feel about Elon Musk as a controversial figure, it’s hard to argue against his leadership touch for this sort of thing. He’s built a company that iterates faster, manufactures cheaper, and thinks bigger than the traditional aerospace industry ever could. That didn’t happen by accident.

This is what disruption looks like. And Artemis is on the wrong side of it.

What Apollo Actually Built

Apollo didn’t just plant flags and leave footprints. It fundamentally transformed American technology, creating entire industries that didn’t exist before.

The Birth of Silicon Valley

The Apollo Guidance Computer was one of the first computers to use integrated circuits at scale. In 1963, the Apollo program consumed 60 percent of the entire U.S. supply of integrated circuits. NASA didn’t invent the microchip, but it created the market that made microchips viable.

The people at Fairchild Semiconductor who built those chips—notably Robert Noyce and Gordon Moore—went on to found Intel. Moore wrote his famous law about chip density doubling while working on Apollo contracts. The area where Fairchild and its competitors clustered began calling itself “Silicon Valley” by the end of the decade.

Without Apollo’s demand for miniaturized, reliable electronics, the semiconductor industry might have taken decades longer to mature. Every smartphone, laptop, and server traces its lineage back to those early chips proven reliable by the demands of spaceflight.

Software Engineering as a Discipline

Margaret Hamilton, who led MIT’s Software Engineering Division for Apollo, literally invented the term “software engineering.” People laughed at her for it—at the time, programming wasn’t considered engineering. It was just something you did.

Her team wrote the code that saved Apollo 11. When the lunar module’s computer became overloaded three minutes before landing, Hamilton’s priority-driven architecture kept the critical tasks running while shedding the non-essential ones. The astronauts landed because the software was designed to fail gracefully.

Hamilton pioneered asynchronous software, priority scheduling, and human-in-the-loop decision-making—concepts that underpin every reliable software system today. She later received the Presidential Medal of Freedom for her work.

Fuel Cells and Clean Energy

Batteries were too heavy for a lunar mission. Solar panels of the 1960s weren’t efficient enough. So NASA funded the development of practical fuel cells—devices that generate electricity by combining hydrogen and oxygen, with water as the only byproduct.

The Apollo fuel cells didn’t just power the spacecraft; they provided drinking water for the astronauts. And every commercial fuel cell today—the ones powering buildings, buses, and backup systems—traces its intellectual property heritage to those three companies NASA funded in the early 1960s.

The facility in South Windsor, Connecticut, that built Apollo’s fuel cells continues to produce commercial fuel cells. They’re now made by HyAxiom Inc., still building on technical know-how developed for the Moon missions.

Materials That Changed Everything

After the Apollo 1 fire killed three astronauts, NASA developed fire-resistant fabrics that are now standard in firefighting equipment worldwide. The same materials protect buildings and save lives decades later.

The insulation developed for Apollo spacesuits—layered metalized Mylar sheets—became the “space blankets” found in every emergency kit and handed out at marathon finish lines. The same technology now appears in building insulation, MRI machines, and particle colliders.

NASA needed scratch-resistant helmet visors. They developed a coating, licensed it to Foster-Grant, and now virtually every pair of sunglasses and prescription lenses uses descendants of that technology.

The ablative heat shield that protected Apollo astronauts during reentry—AVCOAT, a quartz-fiber reinforced resin in a honeycomb matrix—pushed materials science forward by decades. Each of the 330,000 cells in that honeycomb was filled individually by hand, a six-month process that taught engineers how to manufacture at scales and precision levels previously impossible.

Project Management Itself

Apollo was so complex that traditional management approaches failed. NASA invented the matrix organization and systems engineering methodology to cope. The Program Evaluation and Review Technique (PERT), Phased Project Planning, and integrated systems management all emerged from the challenge of coordinating 400,000 workers across thousands of contractors.

Every large engineering project today—from skyscrapers to software platforms—uses management techniques pioneered because NASA had to figure out how to land humans on the Moon in less than a decade.

The Ripple Effects

Consider the scale:

• 1,800+ spinoff products documented by NASA as of 2015
• $7 return for every $1 invested, according to econometric studies
• 400,000 people employed at peak, including the first African-American co-op engineers hired through NASA’s diversity recruiting initiatives
• 20% increase in federal STEM funding during the Apollo era, creating a generation of scientists and engineers

Cordless power tools. Freeze-dried food. Water filtration systems. Memory foam. LASIK eye-tracking technology. Improved pacemakers. Digital imaging sensors.

Apollo was a technological Manhattan Project that reshaped American industry.

What Apollo Had That Artemis Doesn’t

Apollo had a why. Beat the Soviets. Prove American technological supremacy. Honor a martyred president’s vision.

What’s Artemis’s why? “Because we can” isn’t compelling enough to justify the cost. “Jobs” isn’t inspiring. “Stepping stone to Mars” rings hollow after 50 years of that same promise.

In 1968, Apollo 8 astronauts read from Genesis on Christmas Eve while orbiting the Moon. A quarter of the world’s population watched. People wrote letters saying the mission had saved a terrible year.

No one is going to say Artemis II saved 2026.

Where I Land

I love space exploration. I’ve dedicated years of my life to talking about it, thinking about it, dreaming about it. I want humans on the Moon, Mars, and beyond.

But I want it done right. I want a sustainable program with clear goals and public support. I want science prioritized alongside exploration. I want a “why” that matters.

Artemis, as currently constructed, doesn’t give me that. It gives me a jobs program masquerading as exploration, using outdated hardware, while cannibalizing the science programs that actually expand human knowledge.

I hope I’m wrong. I hope Artemis II launches successfully and reignites public passion for space. I hope the program evolves into something sustainable and inspiring.

But right now? I can’t get excited. And I don’t think I’m alone.

What do you think? Am I being too cynical, or does Artemis need a clearer purpose? If you find inaccuracies with any of my calculations, please let me know. I’m interested in getting it right. Easy to correct. Let me know in the comments.

Sources and Further Reading

Apollo Technology Impact:

• NASA: Going to the Moon Was Hard — But the Benefits Were Huge
• Smithsonian: Apollo Guidance Computer and the First Silicon Chips
• Computer History Museum: Silicon Chips Take Man to the Moon
• Smithsonian: Margaret Hamilton Led the NASA Software Team That Landed Astronauts on the Moon
• NASA Spinoff: NASA’s Moon Shot Launched Commercial Fuel Cell Industry
• Cambridge: Powering Apollo 11 – The Fuel Cell That Took Us to the Moon
• NPR: Space Spinoffs – The Technology To Reach The Moon Was Put To Use Back On Earth
• Apollo11Space: How Did the Apollo Program Impact the U.S. Economy?

SLS vs Starship Cost Analysis:

• Everyday Astronaut: SLS vs Starship – Why Do Both Programs Exist?
• NextBigFuture: RS-25 Engine Costs $145 Million Each
• Everyday Astronaut: Is SpaceX’s Raptor Engine the King of Rocket Engines?
• SpaceNews: Aerojet Rocketdyne Defends SLS Engine Contract Costs
• NASA OIG: SLS Booster and Engine Contracts (IG-23-015)
• NASA OIG: Management of the Artemis Missions (IG-22-003)
• NASA OIG: Transition to SLS Services Contract (IG-24-001)
• SpaceNews: New Contract Unlikely to Significantly Reduce SLS Costs
• Elon Musk: Raptor 3 Cost Statement (March 2025)
• Tesmanian: SpaceX Manufacturing One Raptor Engine Per Day
• Payload Research: Artemis Vehicle R&D Costs
• Spaceflight Now: Trump FY2026 Budget Proposal