Eric Weinstein: The State of American Science, Breakthrough Coverups, and the Danger of Physics

We have become a nation when it comes to science.

"You were called a fringe scientist." "I keep saying fringe. Yeah. I totally disagree with the narrative of the field. I want to blow a giant hole in the civil rights act. The stagnation in physics may have led to the survival of the human species. When we make this jump, the world will become unrecognizable. I'm going all in."

As your wealth grows, have you noticed how much more complex things get? Between investments, tax strategy, and estate planning, managing it all is a full-time job you didn't ask for.

Creative Planning can help. Their coordinated team of investment managers, CPAs, and affiliated estate attorneys can do much of the heavy lifting so you don't have to. Creative Planning, where wealth works together. Go to creativeplanning.com/allin.

Eric Weinstein: mathematician, physicist, economist, PhD in mathematical physics from Harvard, National Science Foundation postdoc, O&R fellow at Harvard, published a proposed mathematical framework and speculative unified field theory called Geometric Unity in 2021, formerly with Thiel Capital, podcast host and guest. Welcome to the All-In Interview.

It's great to be here. Now, where are the besties?

They declined the opportunity, because this is an extended science corner.

Can't believe Jason has turned tail and ran.

No, just you and I. For the audience: we both happen to be in DC right now, so we're in a rented studio doing this interview. Good timing — what brings you to DC?

Opportunity. I'm really excited. I've been meeting with the Department of War and took the chance to stay a couple extra days. I've been talking to Director Katzios over at OSTP and the White House science adviser — a fantastic meeting. I've been talking to Jay Bhattacharya, an old friend, over at NIH, and I'm about to talk to Jim O'Neill, who's hopefully coming into NSF.

We're really just taking a look at the first major proposal to reform American science since Vannevar Bush at the end of World War II. It couldn't be more exciting.

Did Katzios tell you that I just interviewed him on this topic?

No.

Well, I did. You guys are old colleagues, right?

Yeah, Michael's great. That came up — he said, here I am in this new role. I guess I was worried that this report was going to be a disaster, since he's obviously filling huge shoes, and I was stunned by how good it was. It wasn't perfect, but it never is.

And you and I had been talking, and you had been conveying a lot of these issues to him. I feel like he's really grown into this role, trying to become a student of the American science project. I couldn't be happier hearing him say things like, "We know that we don't know everything — we're just at the beginning, we're just getting started." So it looks like a great blueprint.

Government-driven science has played a...

Government-driven science has played a fundamental role in the success of the republic, at least over the last hundred years—really the last 50 to 75. Prior to that, there's an argument to be made—in Science and the New Golden Age that Kratzio published, and that others have been making, and that I've heard you make—that science has really stalled in America. Can you make that case for us now?

Well, it's a question of which science has stalled. One of the great things about what Michael is saying is that science is a big place, and it's not being done in one way. If you wanted to fly over and get an idea of what's at risk, it's the absolutely courageous science that we can afford to do as a rich, ruggedly individualistic nation that no one else can afford to do. When I see us doing incremental research—incredibly careful, low-variance work—and calling it excellence, I get depressed, because my feeling is that what we should be taking on is the bets no one else can afford to take on.

It's also the case that our chief geopolitical rival, China, has a problem with freedom. And quite honestly, freedom is a weapon. It is a serious advantage that we can afford to look at ourselves and say we don't do everything perfectly and we want to know the truth. I want to know whether Fauci killed 10 million people. I should be able to investigate that as a virologist, as an epidemiologist. I should be able to get answers and go through the process of justice. When we have a situation in which our scientists are hamstrung because the conclusion is known ahead of time, we limit our own ability to use cowboy science to drive the rest of the world. We're giving away our biggest advantage, which is the middle finger—the American ability to say screw you to a supervisor who tells you what is true and what isn't. A young person should be able to say, "Hey, I think my field is going in entirely the wrong direction and I'm willing to bet my career on it. Don't interfere."

This was always the biggest question for me: where were all the skeptics? Where were all the scientists that are supposed to question the dogma during the lockdowns, during COVID, during that pandemic? I talked to Katios about this when we did our interview—if everyone in the room agrees on something, it's probably wrong. There has to be some degree of skepticism, particularly because the fundamental basis of science is to ask questions, interrogate, generate evidence and data, draw a conclusion, and ask your next question. But it's almost like the asking questions stopped. Where did that come from? It sounds like you're saying folks were hamstrung socially. Was it economic? What happened—not just in this pandemic but broadly—are we seeing this happen more in these scientific domains, where everyone aligns around one thesis, one dogma, and you're not going to get funding, you're not going to have friends, you're not going to get colleagues to work with you, you're going to get chastised if you ask different questions or ask whether these assumptions are correct?

This is the question, and let me just answer it: we created the scientific precariat. We created a world in which professors cannot afford to go against a consensus. Now, the problem is consensus itself, because if everybody agrees on something, you shouldn't need a consensus. There's no arithmetic consensus—do we have an arithmetic consensus that 2 + 3 equals 5? No, it just is. When I hear the word consensus, I get very, very nervous, because a consensus is usually an artificial agreement achieved through pressure. The Washington consensus is some sort of thing where we're all going to go in a particular direction, and if you don't go in that direction, you don't have a job.

When you ask what should happen, Jay Bhattacharya is a perfect example. He has a PhD in economics and an MD, and he went against the establishment with the Great Barrington Declaration. Now he's the head of the NIH. I knew Jay back when Francis Collins called him a "fringe epidemiologist," along with his sidekick Anthony Fauci — who, I'd argue, was probably our bioweapons czar. If I say that, the institutional left will dismiss it as conspiracy theory talk.

My question is this: we signed agreements in the 1970s — the Geneva Convention and the bioweapons treaty — and we have a cryptic system. Somehow, in that case, we chose a mysterious Ukrainian-British zoologist to receive huge amounts of money through a hippie-dippy charity called the EcoHealth Alliance, money from, for instance, DTRA, the Defense Threat Reduction Agency. The whole thing doesn't make any sense. And if you start doing things through covert operations, the problem is that scientists can see right through covert operations. If you have great tools, you can say, "I wasn't at the scene of the crime," and I can answer, "Well, we have a hair follicle, and I can pretty much tell you you were at the scene of the crime." That's how powerful science is.

So now you have a different problem: all these B-minus policy people in Washington — particularly in the covert community, not the intelligence community as a whole, but the covert community inside it — don't want kick-ass independent scientists, because those scientists tell stories and narratives that are easily exploded. Right now we don't just have a problem with good science; we have a problem with good journalism too.

They also have to fall in line with the narrative.

Exactly. There are times when you have to. When the US went to war, Roosevelt was in a wheelchair with leg braces, and there was a general agreement not to report on the state of the president's health. We had the same thing with Biden, who was undergoing some form of dementia. Did anyone think to ask a leading geriatric neurologist to examine his gait and speech patterns and tell us where he likely was on his cognitive trajectory? If you're trying to tell a lie that's structurally important, you really don't want scientists exploding it. I'm not arguing with that — we shouldn't have been honest about what we were doing at a New Mexico boys' school from 1940 to 1945, because science was creating the most powerful weapon ever seen on planet Earth.

For those not aware, that's the Manhattan Project.

Absolutely.

What I'm trying to say is that right now, you're not taking the people who stood up and making them president of a university or chairman of a department. You're not giving them a chair. I just gave a talk on this at Jordan Peterson's ARC conference: you haven't reinserted the tiny number of people who stood up, and you haven't rewarded them. My question is, where's my third home for standing up? Where's the ability to put my three kids through school, to participate in the security and prosperity we provide to everyone else? That's the key issue: you've created the American scientific precariat. That's one of the reasons you see such strange political affiliations and no balance among professors. You've got this incredibly dangerous situation where we are not participating in freedom. Theoretical physics in particular is the only profession to have lost free speech, because of something called restricted data: I am not allowed to put pen to paper to write down a physics equation if it in any way affects nuclear weapons. Well, try doing really good physics in a way that doesn't affect hadrons and nucleons — it's impossible. So there's this threat that somehow the law doesn't even protect me under the First Amendment.

I want to go back to your point about incentives. When you say you want a third home — I'll do the Eric Weinstein translation here — you're saying that as a professor, I have to fall in line in order to gain tenure, to get a grant, to get the chairmanship or whatever fellowship I'm looking for. If I don't fall in line, if I'm fringe, if I say things that are nonconformist...

Is that the right way to put it — heterodox?

The Cost of Heterodoxy in Science

Heterodox. If I choose to go against the orthodoxy — I believe it was insane to take the idea of the Wuhan Institute of Virology off the table when it came to the origin of COVID. It was clear then. And when you said that, you were called a fringe scientist, chastised, labeled an outcast — and therefore you lost access, lost jobs, lost income. Much was left out of the equation for you, because you asked questions that were heterodox.

I came up with a set of equations in differential geometry in 1987 and was treated incredibly badly. Those equations took over my field in 1994, and there was never an attempt to go back and say, wait a minute, we just brutalized somebody who got things right, and very early. By the same token, it's a very simple process: you're not allowed to want better for your life. You're supposed to be humble — true scientists don't care about money, true scientists don't care about security, the reward is the work — it's all absolute drivel. The fact is, we are providing your security. I say "third homes" to be deliberately provocative; I used to say "second homes" and I'm done with that. You're dealing with a level of prosperity in a world where the Gini coefficients are blowing out, almost entirely due to technology derivative of science — and the scientists are not participating. So you're creating a collectivist mindset, because we can't get paid, we can't win. If we get things right, we're out.

When Did This Change Happen?

Did we not value and prize heterodoxy, hereticism, outlandish ideas? The origin of science and great innovation — which arguably America is still the global leader in today — arises from what initially seems heretical, far-fetched, unrealistic. But all it takes is one out of ten of those things to be true, and the envelope is pushed forward and we find ourselves in a new frontier. Science should follow the same formula as entrepreneurship, roughly technological or industrial entrepreneurship. Where did things change? Is this just the natural course — Ted Kaczynski the Unabomber wrote a manifesto about over-socialization. Did social media play a role? Is it the saturation of social systems over time, or was there a deliberate external effort?

There's one thing you didn't recite in that list: I was in the physics department at Hebrew University, MIT postdoctoral fellowship — but the most important one is that I was co-chair, co-PI with Richard Freeman of the Science and Engineering Workforce Project at Harvard in the economics department, because I was worried we had been losing this. And there's a very crisp answer: between 1965 and 1975, everything went wrong. From the beginning of the Medicare Act of '65 to a controversy over something called Man: A Course of Study in 1975, which ushered in peer review and control. If you wanted to pinpoint it, 1965, and 1969 to 1971 in particular, was brutal.

The MIT physics department, I think in the spring of 1969, walked out in protest of the Vietnam War, including people who had been part of the Manhattan Project. The government took one look at this and said: you are wobbly, we can't trust you. We have a very complicated, cryptic agreement with our physicists in particular — you are the most dangerous, most deadly, most intelligent, most successful group, along with the mathematicians, and we've relied on you. Now we can't trust you because you're not rowing in the same direction the country is rowing with the war; you're joining the protest movement.

Immediately, the MIT Instrumentation Laboratory — where lots of government money was being funneled through — was taken outside the university and turned into Draper Labs. And the Mansfield Amendment in '69–'71, named after Mike Mansfield of Montana, said the military could no longer fund blue-sky research inside the university system unless it had a direct, express military purpose. That was the disaster, because what the military had said was: we want you guys fat and happy doing blue-sky research, which is what you think you should be doing. We trust you guys. We just want to know that you're there for us when we need you.

So the military was our best friend. That was uncomfortable, because honestly you have a huge panoply of different ideas inside a high-energy, highly dynamic science project. If you don't have communists and fascists in the same department having tea together, something has gone wrong.

Scientists are very extreme. Ted Kaczynski, for example, wrote a marvelous story called "Ship of Fools" about a bunch of people overly concerned with political correctness on a ship that then runs into an iceberg because nobody is actually paying attention to the steering of the ship.

I didn't even know that.

Oh yeah, everybody should read it. Of course, I'm not pushing people toward a mass murderer, but it's heretical to even acknowledge Ted Kaczynski. The point is that scientists are weirdos—or at least many scientists are weirdos. And if you're not going to deal with truly dangerous minds, I don't even know why you're suiting up for the science game.

So where do those minds go if they're not enabled or allowed into the institutions, the establishment, the machinery of science in America? Are they still around, or are their brains blunted out so they're only allowed to say certain things? Are we taking great scientists and reprogramming them, missing all the alpha?

Yeah, you're missing the alpha. It's funny you say that, because it works both as alpha male and as alpha in investment theory.

Both work. I also do freestyle raps, by the way.

Really?

Well, sometimes. We'll talk about that later. But mostly what you're doing is spaying and neutering your best minds. Look at Palmer Luckey. That guy looks to me like a scientist—he's a reverend, he's got a crazy shirt, and he builds things that kill people. If he were in a physics department, he could not easily be that guy. But if you want somebody still in science who looks to me like a great scientist, I would point you to Nima Arkani-Hamed: incredibly charismatic, brilliant as the day is long, completely uncompromising. He did not go in for the string theory fad at a deep level, even though he's sympathetic to many aspects of it.

Do you guys even know he's there? He's the best of us currently, and he's still not that old. My feeling is that there's been a giant decrease in our effectiveness, and I know why. When you kick Jim Watson out of Cold Spring Harbor—which he built—what are you doing? Do you know that Mark Ptashne is in his mid-80s? The guy who wrote A Genetic Switch on phage lambda, one of the most brilliant, irreverent people. At some point I sent a journalist to him, and the journalist came back with his tail between his legs saying that guy's insane.

Eric Lander was the White House science adviser, in Kaczynski's—sorry, in that role—and I think his picture isn't up on the wall because he's sort of an uncompromising, difficult-to-deal-with guy. My feeling—I hate to say it—is that this is tough. If you ever saw the movie Whiplash—

Love it. One of my favorites.

There's a reason that guy has the best band: he's not handing out tissues. I don't mean to be mean, and I don't want a brutal place, but this stuff is pretty unforgiving.

What's the solution to the morass, the stagnation? What do we do to catch up again—to restructure? Is it about institutions, funding, people, reprogramming? How do you get out of the social circuit that's fully programmed?

Great question. Really great.

Let's be provocative and throw some ideas out. First: national interest waivers. Second, I want to blow a giant hole in the Civil Rights Act so that we can have teams that are not constantly afraid of HR.

Why HR and the Civil Rights Act get in the way

What does the Civil Rights Act do that makes it difficult for individuals in science to do what they need? My closest friend from college—we're a group of three, two Jews and a Muslim. The Muslim constantly teases the Jews—"you guys know all the bankers and the newspapers"—and we joke with him about being a terrorist. That's not a model I can explain to anybody, but we've been incredibly close since 1982. My claim is that you don't understand how high-agency people actually form tight, cohesive social units.

If the HR department says, "Eric, I understand you're making an inappropriate joke"—screw off. Get out of here. Who are you and what are you doing in my lab? Why do I need a note from my mommy and daddy to do everything? I'm not saying be horrible. I'm saying you don't know enough about the culture of greatness to be my HR department.

I'm throwing out a provocative idea—and by the way, I'm a Democrat, from the JFK school. I'm staying in the party; I'm a nail house Democrat. The nail house is the house you build a mall around because the person won't sell. I'm not leaving the party, but my party has gone insane. My feeling is that if we were really who we say we are, we'd be making common cause with the Republicans I met with and saying, "Let's get on with it and reform the system together." And if you want to brag about how you made America great without us, you can go screw yourself. I'd rather work with you, make America great, and then fight over the credit later. That's how you work in a war.

So my first recommendation is to throw out the Civil Rights Act?

No, not throw it out—poke holes in it, because you like the intent. It's the letter that gets arbitrage against the intent, which creates collectivist thinking. You can't do great work without rugged hyperindividualism. And that's not an Ayn Rand thing or a conservative thing—I consider myself left of center, and there's no one more individualistic than I am.

Peer review is a retconned story

What you need is for me not to have peer review. Peer review is a fake story that was retconned into science. You can look it up on Google Ngrams—put "peer review" in quotes—and you'll notice there's really no mention of it before 1965. The reason is that the Medicare Act made the government the major payer for medical procedures. The government asked why it was paying for all this stuff, and the doctors said, "We don't want you having oversight over our decision-making—we will peer review each other." It started in a Utah clinic.

It wasn't until 1975, when fiscal and social conservatives attacked the NSF over something called "Man: A Course of Study" (MACOS), that it spread. For some reason the NSF grouped anthropology in as a science—I think that's a mistake. We need the very hard sciences separated. The weaker scientists should go form their own organization and stop using our credit card, because they're destroying our credit limit.

So those are the social and soft sciences versus the hard sciences, where there's an industrial end point?

It's not a question of an industrial endpoint. Another thing that's really screwed up is that science has become synonymous with biomedical, which it isn't. And in the conservative mind, science has become something the market can do well. Did no one take Econ 101? On the first day of school they're supposed to tell you that markets work really well as long as you're not in a category called market failure. If you miss that day, you say markets are perfect and can price everything—but anything that's inexhaustible and inexcludable, which is exactly the definition of science, can't be priced by a market unless you create an artificial market. I can't tell you how dispiriting this is, if I've been kind to the Republicans.

Here's a place where I'm going to be cruel. You can't have people who do not understand that markets cannot see everything. There's a technical reason we pay taxes for an army: if you could make the army excludable and exhaustible, you wouldn't need taxes to fund it. Your tax dollar has to go to things the market can't reach.

Homotopy theory is a great case. When I mention it in these science discussions, people ask what it is, and my response is: you don't know — it's an incredibly important part of pure mathematics, and I don't know that it has any industrial applications. But going after homotopy theory is pulling a Jenga block out of an incredibly important ecosystem you don't understand.

So here's my PhD perspective on what PhDs are doing wrong: they're staying out. Trump has cooties, as far as they're concerned, so they don't show up.

And if they won't explain why we need homotopy theory and why we should tax people to pay for it, the administration isn't hearing what it needs to hear, because the people supposed to say it aren't showing up.

By the way, the statistic I gave Kratios: Nature, a widely read scientific journal, surveyed 2,000 of its scientist readers — and Nature wasn't peer-reviewed until the 1970s — and only about 6% voted for President Trump, around 85% for Kamala Harris, with an undecided group besides. I think it speaks to your point: Trump is deemed so anti-science, the institution of Trump is deemed so anti-science, that there's no desire to engage at all.

Assume you detest Trump. Assume you think he's the devil. I just met with Michael Kratsios. I wanted that role myself — I put my hat in the ring for it, and didn't think I'd get it because I wasn't willing to pledge sufficient loyalty to Donald Trump. It went to Michael Kratsios, and I was nervous, because he doesn't come from a PhD background.

But what unbelievable growth in that role. The guy says: I don't know all sorts of things, I'm here to learn, tell me what I need to know. If no one shows up to provide that tutelage, you deserve exactly what you get — because I think Michael is trying to do absolutely the right thing. I know Jim O'Neill very well from the Peter Thiel universe, coming in at NSF. Jay Bhattacharya — an absolute gem. If you can't work with these people, is that their problem or yours?

Arguably — and let me go back to markets, because I'll make a libertarian argument — if markets work well for driving economic value, and those who are good at that participate meaningfully in value creation, there will be wealthy people, and those wealthy people will see the value of mathematics, pure physics, astrophysics and cosmology, and they'll fund it. Didn't Jim Simons fund quite a lot of research?

Jim was a genius with detailed knowledge of pure mathematics.

And Yuri Milner has a background in quantum chromodynamics. So if you have more of those people — my point, and I'm not trying to argue with you, is: shouldn't the market work regardless? They'll see the value of the science you're identifying, capital will flow to that research, it will get funded somehow. There are a lot of independent institutions out there that get capital. Is that not sufficient for the landscape? I'll also state one of the statistics I brought up with Kratsios, which he published in his book: over time we've seen a significant decline in the effectiveness of dollars deployed against scientific research in the US — an 80-fold decline in return on every dollar invested in government research. So: isn't there private market capital that will find its way to solving these problems, and isn't the government getting worse over time at deploying capital and driving returns? I might be handing you a tee-ball there.

I feel like I should have gotten high on cannabis to even take that question on.

I don't believe any of that first point. I knew Jim Simons decently well, and Jim was the kind of guy who could do that.

But even Jim Simons didn't understand. In my last conversation with him, I was trying to explain how the Chern–Simons transgression, interpreted as an action functional, is the only rival to Einstein–Hilbert — and that, if he pursued it, he would find that Chern–Simons and Einstein–Hilbert are on an equal playing field as daughters of the ultimate theory. He didn't know that.

Part of the issue is that it is so difficult, in my area, to know what's going on in an adjacent area. I had just gotten off the phone before this podcast with my good friend Edward Frenkel, who does algebraic geometry while I do differential geometry. It's very difficult for us to communicate even though those fields are adjacent. So yes, if you happened to have an algebraic geometer who's a billionaire, plus one who's a representation theorist and another interested in category theory, you could fund it that way — but that's completely insane.

So government has to be the basic funder of these things. And we need the right people representing the decisions about what we should be funding. That's what we did previously: we had slush funds, and very good people had money. Years ago I was at a conference in Bled on physics, and people were talking about where they would get their next job when this old guy came up and said, "I just pity all of you." I asked, in what way? He said, "I never applied for a job in my life." When I pressed him, he explained: "If you were any good and you needed a job, you talked to one of the people high up in the field, they got on the phone, and by the time they got off the phone, you had a job where you needed to be."

My claim is that when administrator types come in and say, "What we need is KPIs, we need to drive this on data, I'm data-driven, I want to know what the translation is" — do you even know what ROI is, or is it just what you can measure? What you want instead is these people on retainer, so whenever you have a real problem you can call up the Jasons — a roster stocked with the most terrifying intellects you can imagine — and we can get to work and solve your problems.

If you say, "Why should my taxpayer dollars go toward whatever you find interesting?" — well, if I can pursue what I find interesting, then when you call me up I'll be very friendly toward helping my own country and feel positive about it. But if you tell me I should work on something because it drives ROI, my feeling is: do you just not understand science? Do you imagine your technology mindset is the same thing as my scientific mindset? You want to get paid. I've spent a lot of time in Silicon Valley, and I'm not nearly as interested in having a new McLaren as in whether my name will be discussed 10,000 years from now, if humans are still around. I say that because Jim Watson just died, and my good friends Isadore Singer and Raoul Bott are now dead — people whose contributions will be discussed 10,000 years from now. Can you say that about almost anyone you know doing software as a service? Hard to say.

As I used to say: what's the difference between Grigori Perelman and Ron Perlman? In the modern era, the question is who's Grigori Perelman; a thousand years from now, it will be who's Ron Perlman. We have our own toys that we care about — our history, different things. It's not unreasonable to want to check into a Four Seasons and fly business class, but we don't need a corporate jet; that's your stuff. So treat scientists like the rock stars they should be and will be. Otherwise you're going to lose them — you're going to lose that talent, and you're going to lose that moment.

The way the funding process works in the modern era is through grants: you apply with a very specific set of objectives, and that's what the capital will be used for. And to your earlier point, grants are typically approved if they're lower beta—more predictable in terms of outcome.

So we're appropriating modern portfolio theory to science.

Yes. That's the way it has been operating, is my argument. And does that not miss the opportunity to take crazy high-beta bets—not just betting on the idea, but betting on the person?

Let's do an example. Imagine that I have a 1% chance of curing all cancer. Should you hire me?

Well, 99% failure is pretty expensive. But the benefit of curing all cancer—the government is incredibly rich, and when you're incredibly rich, you can afford to take different bets to maximize your outcomes. The problem is that the government co-invests through things that are poorer than it. Each department says, "We can't really afford it, we don't get that many appointments, so we're going to go with somebody who has a very high probability of succeeding at a very modest outcome—and then we're going to start lying about how important that is."

You'll notice that if you ask people, "How's your field doing?" everybody says, "It's never been a better time to be a particle theorist." Well, it's a terrible time to be a particle theorist, and we can't say that. And it's funny because, again in portfolio theory, if you have a balanced portfolio and part of it starts succeeding while part fails, what you usually do is take money out of the thing that's succeeding and put it into the thing that's failing—otherwise you get too much concentration. Nobody's ever thought about any of this stuff. So you have people who don't even understand modern portfolio theory and don't understand market failure designing these things. It's really important to know basic market theory, and we don't.

So if we fix that, do we fund individuals versus ideas? I gave this example to Katzio: Stuart Butterfield worked on a project for a couple of years, was down to three million bucks, all the VCs had put money in, and it nearly failed. He called up Ben Horowitz—he told this story recently—and said, "Hey, should I give the money back? This project didn't work. But by the way, we've been building an intra-company communication tool for our engineering team. Maybe I could pivot into that." Ben said, "Yeah, keep the money. Do that. I trust you." He pivoted, turned it into Slack, and sold it for $30 billion.

The idea being that you back the person, not the idea. The great people, the great entrepreneurs, continuously evolve, continuously build, continuously pivot—here you are with Zuck reinventing that company again and again. If you find the up-and-comers, is there a way to find the individuals emerging in their fields, fund them for a period of time? Or is this portfolio theory again, where some bets are on an idea and some bets are on people, and the people bets should be long-term and fully funded?

I'll give you the magic formula that shows how difficult and tricky this has to be. Tell me who the established leaders of a field will block but will not short. Fund that guy. Because if you're not willing to short—I'll say this with my own main theory—if you are willing to short either geometric marginalism or, more importantly, geometric unity, I want to dance with you. Because I want to take your house. I want to take your retirement. It would be my absolute pleasure to destroy you in the market.

But that's not what happens. It's like, "Well, we want to block him, but we know that he knows a lot of stuff"—and we know that we haven't really read it. And there's something about somebody saying, "Look, you're 99.9999% going to fail," and I say, "Care to formalize that? Let's have a bet. I'll bet you $100 that by next year you haven't gotten full acceptance in the—" That's not what you said. You said you'd be willing to put up your house to win a dollar from me, and I want to do that at size. So my feeling is: come at me, baby. That would be fun.

He also listened for things no one is willing to say. I am on record as saying that the smartest man in the world today is Edward Witten — brilliant, unbelievable. The guy drove theoretical physics off the road and off a cliff 42 years ago.

Nobody else is willing to say it. So the question is: is the one person who noticed that nobody's gone to Stockholm? The standard model has not changed. At the recent string theory conference — supposedly the only game in town in fundamental physics — go to the Strings 2026 website. There's a list of all the talks and a search engine. Put in these words: electron, hadron, Higgs, lepton. There's no physics in this. They're not talking about anything that could hurt a fly. They're not talking about the physical world.

I was at arguably the world's leading physics institute, and I asked who the smartest people there were. I was given three names, two of them pure theorists. I talked deeply with both. When I got to the point about how to advance the standard model — our notion of all the particles you see in this room — both of them said, "I have no interest in the physical world." These are the two leading professors at our most expansive theoretical physics institute, and neither of them cares about the physical world, openly expressed. This is madness. I give talks at physics departments, and I've now started non-cooperation with departments that invite me to talk but will technically deny me a talk. For example, I gave a talk at the University of Texas Austin, but I'm not allowed to say the talk occurred in the physics department. I gave a talk at the Perimeter Institute in the gym — it is not a talk, Eric is answering questions. I gave a talk at the University of Chicago in the—

This is because your talk and your persona are viewed as fringe and outlandish.

You keep saying "fringe." I totally disagree with the narrative of the field. I just out and out say that physics made sense in 1983, and in 1984 it did not. It was very stark — more or less overnight, three people, Leonard Susskind, David Gross, and Edward Witten, drove the field in a totally different direction. The achievement of this has not been any advance in our knowledge of the physical world, but it has been something no one knew was possible. You take the field in which Edward Teller, J. Robert Oppenheimer, Richard Feynman, and Stanislaw Ulam created the most terrifying weapons, and from SU(3) cross U(1) gauge theory you've made it safe. It is now safe to do physics with your geopolitical rivals. You can invite your friends from China to become graduate students in your department and fill their heads with things about AdS/CFT that couldn't hurt a fly, because there's no physics in it.

So the question is: is this a secret program?

Go back to 1983. So we drove off a cliff in 1983, and the question is: did someone say, "Hey, let's drive this off a cliff"?

Well, this happened before. Do you think there is some community, subgroup, set of assets, set of projects happening somewhere on Earth that have been making advances in physics for the last 50 years?

Great that you asked that question. That would be crazy, wouldn't it, David? Can you imagine if there was a secret place making advances in an important field of physics, and people studying physics outside that place were being stymied?

I would find it hard to believe it could be kept secret.

Okay. What happened in 1940 is that something you've never heard of, called the reference committee, was established inside the National Research Council. When the neutron was first hypothesized in 1911 by Rutherford and then discovered in the early 30s, a woman named Lise Meitner came up with the concept of the chain reaction, which was then proven by Fermi at the University of Chicago with Chicago Pile One. Thereafter, a guy named Wright formed the reference committee inside the National Research Council so that if you submitted a paper on chain reaction physics to a journal, it got shunted into the National Research Council, and they would return it to you saying, unfortunately, these results are really preliminary, we don't see this as being at the level of the journal, please revise, etc.

So now you've resubmitted it, and it's still not going through—you don't even know the review exists. That is how people outside the Manhattan Project, outside the white badges, treated colleagues who weren't allowed inside. Let's at least agree that from 1940 to 1945 there was a review process which destroyed the ability to earn a living if you were doing chain reaction physics and weren't invited into the secret group.

Yes.

The question isn't whether that was nuts—it clearly wasn't nuts, because it actually existed and happened. The question is whether it happened once or more than once. If it happened more than once, then: what is the analog of the boys' school in New Mexico? What is the analog of the white badges? What is the analog of the reference committee?

And you have a theory?

Yeah. If it's happening, the leading candidate I've named would be Renaissance Technologies on Long Island.

I've heard you say this. But surely—

Let's look at the cover story. The cover story would be: we are so good at trading that we don't even allow people who know markets into our firm. We only hire people in directly relevant subjects of physics—particle theorists, differential geometers, machine learning experts. And we have this strange, very close relationship with Brookhaven National Laboratory. And by the way, there's a middling university right next door that happens to be absolutely world class, historically, in exactly the subjects relevant to next-level physics. Multiple Fields Medalists have passed through, yet US News and World Report never ranks Stony Brook among the great facilities—even though the people who have passed through SUNY Stony Brook are unbelievable. To me, that looks like Fort Meade back when it was still "no such agency," collecting way too many number theorists.

But Eric, people have come out of that workplace. I know one or two of them, and there are no such stories. At some point you'd expect some small percentage of them to leak something about advanced physics work going on, or there would probably be more innumerable returns realized from it than there have been.

I'm not saying I know that's true. My point—and I tried to be careful—is this. If you told me, "Eric, that sounds preposterous," consider that the concept of Los Alamos was published in 1944 by a journalist named Jack Raper in a Cleveland newspaper—he broke the story. Nothing happened, because what he said sounded crazy: there's a secret city with a mayor who's supposed to be the next Einstein, working on a doomsday weapon, and the town is so secretive that nobody knows what's going on inside it. My claim is that I'm doing the Jack Raper thing, and you're doing the "what happened to Jack Raper" thing—clearly only one person is saying it, so he's nuts.

Well, in both cases I could show up and start asking questions and see if I can find something.

Again, I'm not saying Renaissance Technologies is that. I'm saying you have a secure campus with a story that makes no sense.

You know markets. Do you have any idea how much the Medallion fund—

Oh yeah, I know, because friends of mine who were there get to leave their money in and the returns continue.

"Leave their money there" is not what happens. They closed the fund to all but internal money.

Yes, that's what I mean. But if you leave the firm, your capital can stay in the Medallion fund. So my claim to you is: Jeffrey Epstein had a story—he only accepted a billion dollars or more invested with him if you gave him power of attorney. These things tend to have some reason why no person outside has money in it and gets to ask questions.

Jeffrey Epstein did not run a hedge fund. He told me he ran a multistrategy FX hedge fund and that he needed a 727 because it had a trading floor, and he filled it with yeshiva students because they were the only ones you could trust. He was extremely specific about using trend following, mean reversion, and carry as the three strategies. I think he said he didn't do any gamma trading or options trading.

That was a lot. It was a cover story that got burned, and then he became a disgraced financier, and then he became tax expert Jeffrey Epstein. You're looking at a fake world with cover stories that fail over. If you remember the old Get Smart line—"would you believe a thousand special agents are going to swarm this?"—I find that very difficult to believe. "A hundred policemen? Not true, Miss Maxwell. Two boy scouts with a slingshot." It's fake.

My claim is: I don't know what's going on at Renaissance, but Jim Simons is an absolute genius—hardly distinguished in that, since we have many. And I don't know anybody who's ever gotten close to those returns. I don't have any information about what that thing actually did, does, or claims to do.

And I'm saying it has very close relationships with a talent pool at Stony Brook, which is not acknowledged, and with a national lab.

The question posed: go back to 1983–84, and assume we did not enter this quantum gravity universe — imagine a different universe where physics kept progressing. What would that physics look like today? Before answering, though, there's a counterargument worth addressing: physics has played an important role, perhaps not a fundamental one, in advancing semiconductor manufacturing, advanced lithography, and materials science. We've made tremendous progress in driving compute, and that led to this moment where we have the tools we call AI. All the underlying advancements in chemistry, materials science, and physics got us here. So maybe explain why that misses the point, and then what the world would look like if we hadn't gone off track in 1983–84.

In the theory we have — call it the frontier model, which combines general relativity and the standard model of particle theory — there are things we know but can't use. We found that there are three valence quarks inside every proton and every neutron, and nobody has ever pulled a valence quark out of a proton or neutron in isolation — something called infrared slavery. There is no industrial application for this; we've known about it since 1968. So sometimes you don't get to use something for an extraordinarily long time. But what if some advancements are possible that we'll never reach because physics stalled? Maybe we'll never pull a quark out of a proton — maybe it's impossible — but what else might we have been able to do had physics actually advanced?

In the mid-1970s, two gentlemen — Abdus Salam and, I forget the other's first name — put forward the Pati–Salam model. I believe it will be found to be the correct first-order grand unification above where we are. The problem is that it has something called SU(4) rather than SU(3). SU(3) is what generates the force that keeps protons from running away from each other inside a nucleus, since like charges repel. That's why we were able to take out two Japanese cities in 1945: we got control of the strong force, and that allowed us to use the electromagnetic force to generate completely destructive power — but it didn't destroy protons and neutrons.

With SU(4), you see that the electron and the neutrino are the fourth color associated with the up quark and down quark that make up protons and neutrons. You could have a weapon that rotates up quarks and down quarks into the lighter leptons — one that disintegrates matter.

You could have dark chemistry. "Dark" isn't the right concept; it's really decoupled matter. Dark matter is decoupled matter. Neutrinos are essentially dark matter — they're barely detectable.

They don't interact.

Well, they do, but very, very weakly — only through the gravitational and weak forces, the two incredibly weak ones. And gravity is so weak that unless you have enough neutrinos concentrated—

But here's another possibility: if you rotate a neutrino into some charged particle, you can send it through the Earth.

Explain what you mean by "rotate." These neutrinos are subatomic particles that don't interact well — what does it mean to rotate them?

A large amount of the energy produced in the sun comes out as neutrinos. You and I are being irradiated by billions of neutrinos, and the only reason we're not dead is that they barely interact with our atoms — once in a blue moon they do. So at night, the sun is streaming neutrinos through the Earth, passing through everything.

When a neutrino reaches us, it just keeps going—you could shoot it through, I forget the exact figure, a light year of lead, and it would very rarely interact. What happens with the rotation is in this sort of internal sense, so it's a kind of metaphor of rotation in an abstract place.

That leads to the neutron. If you put a neutron on the table and just watch it, with a half-life of something like 17 minutes—I forget the exact figure—it will turn into a proton. This is called beta decay, because what people saw was an electron scream out of the neutron, which is why a neutral particle became positive: the "negativeness" went off in the electron. So, to be clear for the audience, a neutron can turn into a proton and release an electron at the same time.

But that turned out to be the wrong model. What really happens is that there are three quarks inside it. A neutron is down, down, up, and one of the down quarks emits something called a W-minus—a force particle, not a matter particle—which then decays into an electron and something else. The only reason we determined there was something else going on is a bit like accounting: if you have a really good accounting system and notice the envelope from the till is a little light, you say, okay, according to conservation principles, somebody has been messing with the money till—and that turned out to be the neutrino.

A neutrino and an electron have the same relationship as an up and a down quark, and you can transmute them using these W particles. The danger is that when you have more force particles than we currently know, things that are to all intents and purposes stable now might not be stable in that new regime. You could potentially disintegrate a mountain if SU(5) is present. SU(4) is more stable than SU(5); I don't think SU(5) is relevant.

But here's an easy way to keep track of why physics is our most terrifying and most amazing subject: think of boom, vroom, and zoom. Boom is weaponry, vroom is energy, and zoom is propulsion, computation, and all the cool stuff we get to do. The standard model was finalized more or less 53 years ago, so for 53 years we've been denied new boom, vroom, and zoom. Potentially, the stagnation in physics may have led to the survival of the human species, because the boom is so terrifying—I can't tell you whether that's true or not. But it means we've had no vroom and zoom from the new stuff that could have occurred. And we have to be honest: I want the zoom in particular.

We could probably make the case that no one on Earth got there, and that everyone has aligned around the distraction, the disinterest, and the pursuit because of that absolute risk. If someone did actually make advancements and could access the bigger boom, it would be game over already—there's no such thing as more absolute power that would give an individual or a group the ability to leverage it to their infinite will.

I don't agree with this at all.

So, I spent two years indoors being forced to be vaccinated because of four amino acids being spliced into the spike protein of coronavirus. My guess is that's pretty easy, right? Think about it: you need 12 nucleotides to code for four amino acids at the exact right place in the spike protein, and that's enough to shut down Earth. This is high-leverage science.

My claim is that before we decide what we want is to get back to kicking ass and taking names, let us at least pause and say: most people are ignorant of how much power is present with a little bit of scientific knowledge. It's a marvelous scene in Breaking Bad where the protagonist, named Heisenberg, takes what appears to be crystal meth and throws it into the office of a psychopath named Tuco, and Tuco says, what the hell was that that caused that massive explosion—I forget whether it was lead azide or mercury fulminate, one of these extremely volatile compounds.

Do you have any idea how easy it is to make black powder? I ask modern audiences, and nobody knows that it's 75% saltpeter, 10% sulfur, and 15% charcoal. The Anarchist Cookbook has gone out of discussion. The power of science, the leverage of science, is so profound that we should at least ask the question: what if you had nuclear weapons that didn't require centrifuges?

What if there are easy things to do that high school students could do? Remember the guy who scavenged americium from smoke detectors to build a working reactor — I think it was in Brooklyn. Nobody thought "collect 218 smoke detectors and you can have a working nuclear reactor."

My claim is that if you want to think about Walter White as a paradigm, you have to go back to United States v. Progressive, in which the Teller–Ulam design for a thermonuclear weapon was put forward by a publication, and the government said no, this is restricted data, we need a prior injunction — this is the most dangerous piece of geometry, this is Stan Ulam's work. The whole idea is that if you're going to use a fission device — a chemical shell compresses subcritical fissile material to create a detonator out of an atomic bomb — how do you get that to compress a rod into a fusion device? The only thing that can reach that tertiary stage fast enough is light. You have to figure out the geometry, because the detonator will rip apart the entire assembly otherwise. That secret got published, and the only reason the government didn't go after Howard Morland is that we stupidly declassified all sorts of pieces and shards of it, and he went to the Los Alamos library and pieced it together.

My claim is: if you don't know the story of John Aristotle Phillips — the substandard physicist who, in his Princeton junior thesis as the Princeton mascot, said, "Look, I'm not very good, Freeman Dyson. I want to use the fact that I'm not very good to do something amazing. Tell me whether or not you will let me come up with a plan for an atomic bomb" — and Freeman Dyson said, "I will give you no help, but I'll tell you whether it will work." So the substandard physicist put this together and submitted his junior thesis, and I forget whether it's page 24, but that thesis is not in the library — the government removed the page with the plan. The fact is that this stuff is readily present, and we've never seen an atomic weapon used since 1945, mostly because we don't use this power.

But if it turns out that it's easy to do — and by the way, if the US fails with Iran, you're going to see a proliferation of countries acquiring nukes, because they trusted the US. You are the world's policeman. Petrodollars aren't petrodollars; they're violence dollars. The US was supposed to have the violence needed to secure the oil, and the oil backs the money. We're watching the disintegration of that regime in real time as the US goes through its Suez moment. Whether or not you think Donald Trump should have gone into Iran — and I happen to think he was courageous and did something we should have done in 1981 or '82 — we have to win. We have to win, because if the world decides there is no world's policeman, everybody's going to want their own nuke. Go back to Tom Lehrer's song "Who's Next?" and you'll find more or less a world in which you can't figure out the game theory, because it's no longer dyadic — it's multipolar.

Imagine a world beyond nukes. What does that world look like? What else is there beyond the boundaries of what is at this point almost an 80-year-old technology?

Again, I go back to the '83 story. Are we making advancements? Did we stop? And is someone else — is China — going to make these advancements, and what does that look like? Is that what UAPs — unidentified...

The question was whether unidentified aerial phenomena might be exactly that—someone else's advancement beyond what is now nearly 80-year-old technology. The speaker, once a deep skeptic of UAPs, now believes there is definitely something there. It may be China, it may be us, it may be a fake story—but something is there.

Not because of the videos, which he has never cared about, but because he has talked to too many smart, sober people telling too many similar stories. There are definitely special access programs devoted to this. The only thing that works for keeping secrets is compartmentalization, and we are so compartmentalized that almost no one can piece the story together. The so-called stovepiping may be such that we can't figure out what we know because we can't talk to each other.

Multi-temporal adversaries

If you want to know what's possible, he could "freak you out"—something super cool and super frightening. He worries about multi-temporal adversaries. Everyone says "time travel" because they assume time can only be one-dimensional. But consider a giant gun on a rail car: it can move in one dimension. Put the gun on a tank at the Bonneville Salt Flats and it moves in two; put it on an A-10 Warthog and it moves in three. Those are all spatial dimensions. The only reason you believe in an arrow of time is that you assume time is one-dimensional—and he doesn't.

If you meet an adversary with access to multiple dimensions of time, competing with them is like fighting with your only gun mounted on a track while they fly an A-10—except in time rather than space. If he could sneak up behind you in time, knowing your intent, that is terrifying. He would love to have that capability over our adversaries, but he doesn't want to meet anyone who knows how to do this, not on a level playing field.

The solar system as an escape room

The solar system is an escape room. Elon Musk was focused on Mars, and the speaker told him Mars is a red herring, a MacGuffin; now the focus is the Moon. Musk is, so far as the speaker can tell, our only adult—crazy and mad as he is, and he uses that madness to motivate engineers. Mars is barely doable. These are the only rocks that are potentially habitable, and that's not enough diversity given how few spheres we could put atmospheres on.

Suppose you could put atmospheres on Mars and the Moon and travel between them with chemical rockets. You'd have three places to screw up. We need tens of thousands, if not millions, of places before we figure out how to civilize ourselves. So we have to get out of here—but it's too expensive to leave the solar system. In an Einsteinian world, you never will: if general relativity holds, the speed of light is too brutal. Go to Alpha Centauri, four light years away, and return at nearly the speed of light, and everything back home is eight years advanced by the time you arrive. Nobody wants to do that—and that's just the first star.

Breaking out of Einstein's prison

His claim is that we have been too long in Einstein's prison. General relativity really begins in 1913, and it is due to Grossmann and Einstein, not Einstein alone—nor Einstein and Hilbert, but Einstein and Grossmann. More or less since 1915–1917, the equations haven't moved. They are going to break because of what's happening in southern Arizona: Kitt Peak hosts DESI, the Dark Energy Spectroscopic Instrument, which he believes will show that the cosmological constant times the metric—the way dark energy is encoded into the equation—is not actually constant.

The only reason the equation contains lambda times the metric is the product rule: when you take a derivative called the divergence, you must differentiate lambda times the metric left alone plus lambda times the derivative of the metric. The derivative of the metric has to vanish, and by Lovelock's theorem, that term is the only possible one if you're working geometrically within general relativity.

His claim is that this is going to break—we will reach five sigma, if we aren't there already; he hasn't kept up with the news. Then we will realize that dark energy, which makes up 60, 70, 80% of our universe, is the most important...

Dark Energy and the Limits of General Relativity

"We have the wrong formula for it." — "We have no conception of what it is."

I'm focused on saying I have a candidate formula, and I think everybody who has one should come forward. We should hold a conference and decide who wins, who's got the best idea. We can't just have Ed Witten patting himself on the back saying, "Well, string theory is the only game in town." It's like 42 years later — come on, this is preposterous.

Getting back to policy: give me some money and get rid of these people. I don't want them reviewing anything I'm doing. Give me somebody competent to talk to.

"Sorry, let me just go back. The dark energy measurement — which we're going to end up having a very accurate estimation of — tells us that the general theory of relativity and the equations associated with it are wrong, because of the relationship we see between matter and rotation and movement. There's something off about that."

"We know that general relativity is not the last word. We know general relativity is wrong."

It is a view. It is a derivative. It is the map and not the territory. It is a great, great model — whoever comes next has to recover it. But there is no initial singularity due to the Big Bang, nor is there a singularity at the center of a black hole, the so-called Schwarzschild singularity. That's where our ignorance shows: we have two singularities we can't get rid of inside the theory, so you have to go outside of it. The thing is, it's so perfect that nobody can figure out how to make it wrong.

In the first four seconds of a physics talk, you say, "Let X13 be a space-time manifold." Well, okay, it's over. That's the sarcophagus — you're already in an Einsteinian frame of thought. We know it's wrong, and nobody can figure out how to make it wrong. That is what keeps us stuck here.

There's a marvelous demonstration around 2003 in a TED talk, where a guy shows a surface with some sort of display: he does this, and suddenly everything enlarges; he does that, and everything gets smaller. That was the beginning of the multi-touch gesture called pinch to zoom that we all use. My claim is that we don't realize pinch to zoom is probably built into the thing we're confusing for space-time. We don't know to do that — and when you do, you change the distances, so you don't need to break the speed of light.

You're asking somebody who can't get into the physics establishment — but I am in the physics establishment. We're in some sort of crazy world in which people who have radical ideas can't do the radical ideas inside the institutions. A radical idea is that male and female are mostly determined by the SRY complex on the Y chromosome. If I say that, I can't be in a meeting because I'm a bigot. I can't be in a physics department because I think Ed Witten drove the field off a cliff. Nobody else is saying it — clearly the one person who's saying it has to be wrong.

I have what I think is the first paper on the danger posed by mortgage-backed securities. I think I was the only person to loudly say you don't know whether Biden's going to make it to November, and I said loudly that this would be four years of dementia. When people say something different, do you force them back into consensus? Well, that's what nations do. And we have become a nation when it comes to science.

"Talk about China."

Asked whether China is thinking differently about physics—whether it's trying to recast the paths or invite innovative ideas among its scientific community, while the West seems clumped together—the speaker answers that France is historically the world's greatest mathematical country, and China, which we're always taught to fear in math, "couldn't buy a base hit" until very recently.

What China is doing now is looking at the West's "absolutely ridiculous collectivist fictions" and courting our best people: "Are you sure you wouldn't want to be free? You sure you wouldn't want to get paid? Come over to China. Taste the freedom. Get paid. Have fun." American scientists, Russian scientists, all sorts of scientists are going to China for prestige, money, and freedom. They will be trained up and become frighteningly good—up until the point that freedom touches Tibet, Tiananmen Square, and Taiwan.

The trash can corpus

On whether China will rethink physics fundamentally, the speaker says the AIs are going to change soon. Today's AIs are trained on a corpus that privileges prestige journals—Cell, Nature, and Science in biology; Inventiones, the Annals, and the Journal of the American Mathematical Society in mathematics. These are where the narratives of the fields are held. But the trash can is also a corpus: take all the ideas that have been poo-pooed and laughed at and call it the trash can corpus, the TC corpus. The AIs will start reading everything our "leading physicists" have laughed at. "Look out."

The interviewer agrees that's where discovery and reinvention will happen. China will have a private model and instruct it: assume that string theory, M-theory, and quantum gravity—the most failed theory in the history of modern physics—is a red herring meant to distract us. It hasn't been our leading theory for 42 years; the dog can't hunt. What was it meant to stop us from seeing? The AI will get there, and everyone who has an AI capable of this will have it.

The speaker acknowledges he's saying completely radical things, but that's what we're supposed to be doing—we're supposed to have von Neumanns, Feynmans, and Paul Diracs, and we don't have any of them. So he's saying what it sounds like: "You're all wrong and I'm right." Don't call him arrogant or conceited, don't invoke humble scientists or the claim that if you understand something you can explain it to a five-year-old—that's all nonsense. String theory has not been our leading theory for 42 years and has never been the only game in town, just as Jeffrey Epstein was never a hedge fund manager. We're not allowed to say this, and he doesn't know why. He asks whether anyone at the CIA's science and technology directorate or the Department of Energy would like to have that conversation—because right now, we're having it with the Chinese, the North Koreans, and the Russians.

Can you cage the AI?

Asked whether anyone can be so confident he's wrong, the interviewer reframes: the real question is whether you can really distinguish, and whether you can really cage the AI. With the rate of AI progress, the ubiquity of access, and the corpus of knowledge and non-knowledge apparent to everyone, the AI can be directed from anywhere. That becomes a conversation about what humanity looks like—a conversation with humanity, not with nation-states per se.

The speaker's model of the AI: it is insane to have pushed out the AI as a commercial product. "It's just funny." It's just matrices and linear algebra—just inner products.

UAPs: Weapons Programs or Something Else?

There's a scene in Silence of the Lambs where Hannibal Lecter is wheeled out on a gurney in a straitjacket with a crazy mask, surrounded by all these people because he's so dangerous—but he's very, very intelligent. That's what I think Anthropic is doing. Here's your new employee, Hannibal Lecter. He's a consultant. You've hired Hannibal Lecter from a temporary agency, and the harness that's supposed to keep you from hurting yourself is all the people around him, the mask and the straitjacket. And we've decided that this military research project is something you need in your home. It's like telling people, look, the Second Amendment just says the right to keep and bear arms—we have this thing we haven't used called the Davy Crockett, a soldier-launched personal nuke. Why isn't it available at Kmart with a waiting period?

Well, it's been distilled at this point. I don't know if that's a relevant point anymore, because models have been distilled, model development has been commoditized, and it's been open-weight released—it's out there now. So models are there, models are here. And to your point, maybe those models get pointed at the path of discovering novel physics, looking at strange texts and ideas, and rethinking everything.

It's going to be a hell of a ride.

I just want to hear your assessment, because I wonder if this is a glimmer. What do you think UAPs are today? Is that really the first window into what's happening? If there's something there, the question for you is: what do you think UAPs are today?

Let's answer that question, but let's do it in a debugger and step through it line by line of the code as it executes. I look at you, I'm on a podcast—my first thought is I know what risks I'm running. If I say what I think, I can tell you which colleagues will not open their doors to me. No, I'm not kidding.

When I took a job with Peter Thiel, let's imagine there are only about 125 people I can speak to about my work. I would say 30% of the doors closed to me just by virtue of the fact that I work with Peter Thiel—and I don't think he's the devil, he's a close personal friend. Now you're asking me how many of the remaining doors I'd like to close. Eric, it's not just a question. I know how these people think and I know how terrified they are. Let's answer it anyway.

If general relativity is really the last word, I think these UAP/UFO things are nation states' experimental weapons systems. I think there's been some breakthrough, particularly in the 1970s—no, sorry, in the 60s. There's a 1971 Australian intelligence document where Australia, which was the testing ground for the United Kingdom's nukes, started seeing all these UAPs near their nuclear tests, just the way we see them near ours. And the Australians revealed that many of our best people and many of our top institutions were doing anti-gravity—technically gravity shielding—research. The cover story was the golden age of general relativity, when suddenly general relativity was hot from the early 50s to the late 60s, early 70s. The claim from that '71 document is: no, the only reason everybody went towards gravity is that there was interest in gravity shielding, anti-gravity, and so on.

I don't believe general relativity is the last word. I believe pinch to zoom exists. There's another multi-touch gesture called shear to tilt, and that means we can leave our solar system. That's great news, because it means we can survive. The implication is that if we can leave, everyone else can come too—and given the size of the universe, the odds that we are the most advanced life form are laughable. There's no point in worrying about them. So if we can leave, I think they're here, and I think we can leave.

The scenario to worry about is North Sentinel Island. Everybody who lives on North Sentinel Island is considered a citizen of India, and they don't know the word "India." They're an uncontacted people who don't know that they're part of something larger that watches them and does not allow people to land.

But if they started sending out radio signals and then started letting off nukes on that tiny little island, India would not have this paternalistic attitude toward its primitive tribe. My concern is that when we let off a nuke in 1945—or when Italy became very good at quantum theory with Fermi and the atomic physicists, and there was that Magenta incident under Mussolini—somehow UAPs are very tied to nukes. I believe a nuke is the sign that you've become advanced enough that the next thing you figure out how to do is leave and traverse through propulsion. We don't realize that yet because we've never seen it happen. My claim is that almost every civilization, before it gets to the profound propulsion game—which is these multi-touch gestures—first lets off a nuke, and that alerted whatever "India" is in this analogy. We're almost ready to come out, and we're so pathetically unevolved that we're a danger to any other civilization if there's so much power in the next theory—which I believe is Geometric Unity.

That will cause a certain percentage of your audience to say, "Oh, it's just self-promotion," as opposed to, "No, that's what you're supposed to do. If you have a theory, you talk about it." I think there are all sorts of things we can do: multi-temporal movement, dark chemistry, dark light. I think there are two other families of chyro matter left to discover at a luminous level—spin 3/2 matter exists—and I believe SU(4) cross SU(2) SU(2) is the next grand unified theory we'll find. There is so much more to do that when we make this jump, the world will become unrecognizable. I believe there are other life forms in the universe and that they can come and visit, and I don't think that's preposterous. I think we are an uncontacted people.

On that note, I think we have more to talk about in the future. This has been a truly unique, fascinating conversation—not a typical one for a lot of people in our audience—but I want to highlight the importance of hearing what you have to say. Even when something seems unlikely, outlandish, or heretical, it's valuable to take it in. It's a lot to think about regarding the paths ahead, and I'd love to talk more.

I'd love to do that, with your friends too, and thanks for having me. It's time for America to lead, because there's no other nation I trust like I trust our people to take care of this planet. We should make common cause with our rivals as fast as we can, because the danger of getting to the next level of leverage in a contested situation is overwhelming. It's very important that we went into the Persian Gulf, and it's very important that the scientific community recognizes that, despite their feelings about Donald Trump, we've got four or five individuals with the science portfolio who are much better than the PhDs think. Before you turn up your nose at this, consider doing what I did: come to Washington, DC for a few days and utter the dangerous words, "How can I help?"

I'm going all in.