The Real Reason AI Companies Are Desperate to Be Regulated

Key Takeaway

When a powerful claim triggers fear, urgency, and calls for sweeping regulation, pause before sharing or acting on it. Today, trace the incentives: ask who gains financially, politically, or competitively; separate verified facts from interpretation; and seek primary sources or credible opposing views. This does not mean dismissing AI safety risks—it means resisting fear-driven conclusions until the evidence is strong enough to justify the trade-offs, especially when rules could entrench incumbents and restrict open competition.

Episode Overview

Tom Bilyeu argues that recent AI-doom messaging and industry calls for regulation may be part of a coordinated incentive structure rather than solely a response to genuine safety concerns. He connects AI-company financing, advocacy groups, think tanks, government policy, and political spending to a hypothesis of regulatory capture that could protect major labs while constraining startups and open-source AI. The episode presents this as an early, unproven hypothesis and emphasizes the danger of concentrated government power.

Full Transcript

Transcript of The Real Reason AI Companies Are Desperate to Be Regulated from Impact Theory. Auto-generated from episode audio; may contain minor errors.

Queen Carvania stood hallowed by the morning sun. An army hung on her every word. My champions! I have sold my chariot on Carvana! T'was a lovely SUV, an inexplicably queenly offer. They're even coming to the castle to collect it! Tonight, we feast! An offer you can feast on. Sell your car today on Carvana. Pick-up fees may apply. Something very weird is going on behind the scenes in AI right now. Suddenly, everyone seems to think that AI is going to kill us and that the government needs to step in.

Now, that in and of itself isn't that weird. People have been saying for a while that there's a non-zero chance AI kills everyone. But what's happening now is weird. It's different. It doesn't feel like a wave of people waking up to the seriousness of the problem. There's a sudden sharpening of everyone's attention that feels like a product launch. What kind of product? Fear. Fear as proof that AI is powerful enough to destroy us all, and therefore warrants the $2 trillion IPO that Anthropic is about to do.

By the way, that number is so staggering, it would make it one of the largest companies in all of human history. It's that kind of very important product launch. And that's why our focus is being intentionally pointed at something. So far, that's just my gut feeling. And I don't think feelings are particularly useful in getting to ground truth. So what we're actually going to cover today is what I've learned by spending the last several days looking at all of the evidence that's coming in on this strange new phenomenon that points to this actually being an intentional and coordinated spreading of fear about AI designed to reach a very specific outcome.

And let me tell you, what I found is absolutely shocking. Now, in full disclosure, you should take what follows as the early days of a forming hypothesis. But there really is a lot of connective tissue here, and ignoring it I think would be a huge mistake. When you put it all together, it begs one simple question. Is AI the antichrist that people fear it is? Or is what's going on right now something far scarier? Namely, that we humans are the problem. Let's answer that by digging into the red flags that actually sent me down this rabbit hole in the first place.

Red flag number one that started making me feel really weird was that a 27-year-old researcher that nobody had ever heard of quit his job at Anthropic and then posts a seven-part thread saying that people building AI believe it could kill us all by the end of the decade. Now, on its own, that's fine. But he went hyper viral, and his account had a two-word bio and almost no post history. And despite that, by morning, the thread had 76 million views. By the next day, it had over 100 million views.

Those kinds of numbers demand an explanation. Elon Musk himself had said he'd never seen anything like that from a new account. So if it seems sus to you, it is. But we're going to get to why in a minute. All right, red flag number two. A few days after Coxon's Doomer post went suspiciously viral, the CEO of Anthropic, Dario Amadei, said the most bizarre thing in the history of capitalism. It has always been very strange that this technology is being built by a private company. People ask me that question all the time.

Why isn't this being built by government? And the strangest thing about it is I agree with them. I'm uncomfortable. Would you be willing to give up the technology to the government? To the right combination of governments? That's the CEO of one of the largest and most important companies in the world saying that the technology he's building is so dangerous it would be better off in the hands of some kind of world government than being controlled by him and his company while at the same time meeting with investors for what's expected to be one of the largest IPOs in history.

How do you reconcile those two things? Either it's dangerous and it shouldn't be made or it's the next big thing because it's safe and effective. It's hard to overstate just how strange his statements are in context. Many of the big names in the AI field publicly agreed with Dario. Sam Altman, CEO of OpenAI, Elon Musk, CEO of XAI, Satya Nadell, CEO of Microsoft, and Demis Hassabis, head of Google's AI venture DeepMind, were among them. That's red flag number three. The very people getting fantastically, staggeringly wealthy off of AI are the very ones calling for it to slow down.

Why? It could be sincere concern. It really could be people that are seeing something behind the scenes that if the rest of us saw, we would feel exactly the same way. But it could also be something else. Because when I go through the timeline of everything that just happened and who's connected behind the scenes, I can't shake the feeling that I'm watching a talented magician who wants me to focus on his right hand while his left hand is stealing my watch. That's what we're going to explore today.

Is someone or something trying to control our focus so that they can pull off a trick somewhere out of sight? And if so, what exactly is the trick? My name is Tom Bilyeu, and on my channel, I pursue the most important topics of the day, and today is no different. So welcome to my deep dive on AI and the Antichrist. Now, if that sounds like clickbait right now, I get it. But by the end, this really is going to make sense. I spend a lot of time researching AI, so when I say that what's going on right now feels coordinated, I've got some context.

I'm also a marketer who bootstrapped a unicorn startup with my partners and then took it from not existing to selling for a billion dollars. So I know a good product launch when I see one. I've been a part of many. So this week, when I saw Dario write an essay and give countless interviews about why we should all be afraid of his product, I asked the obvious question, how is this good for his business? How is this going to help him IPO? I know some people are going to say that the way that I'm even thinking about that is just cynical, and I would certainly forgive anyone who looks at Dario's bizarre remarks and thinks, hold on, this is just someone who's trying to do the right thing.

He has a duty to his employees and to his shareholders to make the company as successful as he can, but he also has a responsibility to do things safely. And what we're watching is just somebody struggling with that tension. Dario has admittedly been saying for years that AI is dangerous and that private companies shouldn't be the ones deciding how it gets built. And China and the U.S. really are locked in a race for AI supremacy that has massive implications for the world order, and so they're likely to push things too far.

The IPO, therefore, funds a compute race that he can't just quit because China isn't going to quit, and I'm sure he has tremendous pressure on him from the U.S. government to make sure that we don't quit or fall behind. Trump has said as much, and the researcher who resigned walked away from his uninvested equity to do it, so he wasn't trying to enrich himself, right? So it's possible that Dario really is a true believer, that Coxon really is a true believer, and they're simply trying to bring the risks to the forefront, but I don't think that's what's actually happening.

Now, there is a world in which all of that is true, but if history has taught us anything, it's that when there's power and money on the line, there's usually a lot more going on beneath the surface than what we see. As Charlie Munger famously said, show me the incentives, and I'll show you the outcomes. The people running the companies are caught in a trap. They need far more capital to maintain their lead by continually pushing the bleeding edge of the frontier models. It takes an absolutely insane amount of compute, and there's only so much money available in the capital markets to raise.

Eventually, you have to turn to the public and raise cash through an IPO, but to IPO well, you have to build a tremendous amount of hype around the company. AI valuations are already through the roof, but the narrative that they're using to justify those valuations is starting to get wobbly. Since AI valuations are no longer based on business fundamentals and are instead based on the narrative that AI is going to grow more powerful by the day and ultimately be the last technology we ever make and just suck up all the capital in the world, the CEOs of these companies have to prove that that's true.

They have to prove that AI is as powerful as people think. They also have to deal with the insane levels of debt that they've accumulated and the fact that revenue just is not coming in fast enough to offset the staggering cash burn that's required to stay a leader on the frontier. And they have to do all of that without spooking potential IPO investors. Thankfully for them, there is a narrative play that kills both of those birds with one stone and it's exactly what we're watching them do right now.

They're going to use fear as proof of power. Now, this works for two reasons. First of all, if the models are getting so powerful that they represent a national security threat or even an existential threat, it's proof that LLMs aren't just randomly guessing the next word. They have a will of their own and that's exactly the kind of power they need to have to truly transform the world. Investors may be worried about the implications, but they know that they're investing in something real and that justifies the money pouring in and continuing to pour in even as the revenue is delayed.

Michael Burry, a big short fame, summed this position up best when he said, IPOs need hype and puffery. We are so awesome it could become dangerous. Is hype and puffery? That's why he's making bets against AI. He thinks that the narrative around AI has far eclipsed the reality of AI and at this point, the companies will say almost anything that they need to say to convince people that LLMs are the future of everything. Second, the AI companies use fear as proof of power so they can convince the government that AI is strategically important enough to protect them with taxpayer dollars, if that's what it takes.

And if they can do that, they know that the taxpayer will cover any cash flow issues that they may run into. With the government as their guarantor, they are protected from the violence of the free market. They can relax a bit knowing that Big Daddy will protect them from the predations of China and other competitors who would otherwise force them to relentlessly innovate to stay ahead. And in that process, many a company have slipped and fallen. And so that way, there's just all around less pressure on them to perform and they're effectively guaranteed to survive the ever compounding debt burden that is currently threatening to eat them.

Because they know that the state is never going to let them go away if we're the defense mechanism against China building an AI that could take us over. And if they can stay alive long enough, the world is bound to catch up and the revenue is bound to start pouring in insufficient quantities to ensure their survival into the future. And I think they're right about that. The government protection is never free. And they know that. The government may guarantee their survival, but it will come at the cost of control in the form of regulation and direct influence.

For a startup, this would be horrible. It places a huge cost burden on a company that's in a part of its lifecycle where every dollar matters. But as companies get larger, regulation ironically can become beneficial. And I'll explain more on that in a minute. This is where the conspiracy really revs up and becomes all too plausible. The mechanism works like this. They know that government bureaucracies like organisms strive to grow and acquire more resources. So if people are afraid, they understand that they can come in, build a new department, get more resources, and people will vote for it precisely because they're afraid.

And without massive external pressure to remain small, over time, governments do exactly that, and they end up ballooning in size. You can see it in America clear as day. In 1880, the entire federal government, adjusted for inflation, spent a little under $9 billion. Per capita, that's about $170 a year. This year, the federal government will spend about $7 trillion. That's roughly $20,000 per person. That's more than a 100-fold increase per person. And yes, the country has gotten richer, but even if you strip that out, even as a share of everything the economy produces, the government's take went from $0.03 on the dollar to $0.23 on the dollar.

The slice of the pie the government now takes grew eight times faster than the pie itself. And there's no data to suggest that the government getting bigger is what facilitates a rise in wealth. In fact, it's the opposite. The Berg-Henriksen review of the cross-country literature finds that among developed economies, every 10-point increase in government spending as a share of GDP is associated with about a 0.5 to 1 point lower annual growth. Sweden is a perfect case study of this phenomenon. They were the fastest growth in the world from 1870 to 1950 with a small state.

But then they built their welfare state and growth rapidly fell behind its peers from 1970 to 1990 until it finally started cutting back on the size of its government. The U.S. is another good example. We ran most of our ascent from agrarian backwater to industrial superpower while keeping the federal government under 4% of GDP. But we have that pull to get bigger. To get bigger, you need people to be willing to vote for a bigger government. How do you get them to do that? You get them afraid and you promise to protect them.

Economists have known for over a century that governments grow faster than the economies that feed them. Every country, every era, wars and crises expand the state, and when the emergency ends, the state never shrinks back to where it was. That makes making people afraid of AI, getting the government to come in and regulate them, the perfect marriage. They both want the same thing. The big AI companies want the protection from the competition, and then the government wants an excuse to get bigger. Agencies get rewarded for bigger budgets and bigger head counts, never for solving the problem they were built to solve, and that's why all of us should be paranoid about what's happening right now.

It isn't going to make things better. A government starts as a servant, but left alone over time, precisely for reasons like this, ironically in success, that government becomes a parasite, but people don't accept the growth of a parasite willingly. They have to be talked into making that trade, and AI is offering the bureaucracy the perfect excuse to grow. Taking a short break, but there's more impact theory after. Stay tuned. Let's talk about the thing your business just can't survive without. I go live three days a week at 7 a.m., and every morning, you guys, you incredible people out there, show up.

You're there, you're ready, and if my connection drops in the middle of that livestream, that moment is gone forever. You do not get a second take when it comes to live content, and I know a lot of you are in the same position, whether you're running live events, processing transactions, or managing a remote team. Your business depends on staying connected, not sometimes, every time. That's why I trust AT&T Business. They're built for business owners who can't afford downtime. AT&T Business is a reliable provider for small business owners.

For Small Business Month, we celebrate small businesses by helping them run better. This means reliable uptime, easy switching, smart communications. Impact theory is powered by AT&T Business, built to work. Get AT&T Business at business.att.com. Let's talk about a problem the military couldn't solve with willpower. Picture a soldier making life or death decisions after 36 hours without sleep. This is high stress. There's obviously no margin for error. The military needed their people thinking clearly, even under conditions that would break most of us. So they funded a $6 million Department of Defense research contract to find the answer.

What they found was ketones, the brain's most efficient fuel source. That research became Ketone IQ. Ketones cross the blood-brain barrier and fuel your neurons directly. They enhance mitochondrial function for longer mental stamina. I take it for interviews, before writing, before I do coaching, and I can feel the difference. Go to ketone.com slash impact for 30% off your subscription order or visit your local target to get your first shot for free. That's ketone.com slash impact. Let's talk about the last time you got stuck in a phone tree.

Press four for billing, hold music, a robotic voice asks for your account number. Then a person picks up and asks for it again. The machine heard you, but it didn't understand a word. Now, put yourself on the other side of that call. If you run a business, that's what your customer is dealing with right now. Well, that's where Eleven Agents comes in. Eleven Agents is a platform for AI voice and chat agents that actually listen, understand, and resolve the issue. If you run a business or handle customer operations across support, sales, or marketing, you can start with a demo at elevenlabs.io slash impact pod.

See how Eleven Agents can fit into your workflows and help build experiences that your customers will actually love. Elevenlabs.io slash impact pod. E-L-E-V-E-N-L-A-B-S.io slash impact pod. All right, thanks for sticking with us. Let's jump right back in. Peter Thiel, the very controversial figure behind PayPal and Palantir, gave an interview to the New York Times last year where he laid out how he thinks the Antichrist would come to power in the modern world. The way the Antichrist would take over the world is you talk about Armageddon nonstop.

And this is what you need to regulate. It's the opposite of the picture of Baconian science from the 17th, 18th century where, you know, the Antichrist is like some evil tech genius, evil scientist who invents this machine to take over the world. People are way too scared for that. In our world, the thing that has political resonance is the opposite. It is, the thing that has political resonance is we need to stop science. We need to just say stop to this. And this is where, yeah, I don't know, in the 17th century, I can imagine a Dr.

Strangelove, Edward Teller type person taking over the world. In our world, it's far more likely to be Greta Thunberg. Tying it directly back to AI, a few months after that interview in a series of closed door lectures he gave in San Francisco, Peter got more specific. According to the Washington Post, he said, and I quote, the Antichrist is a Luddite who wants to stop all science. It's someone like Greta or Eliezer. Eliezer was a direct reference to Eliezer Yudkowsky, the hero of the AI doomer movement.

Now, regardless of what you think about Thiel, and I'm sure plenty of you have wild reservations, his predictions are now playing out live this very minute. As he aptly pointed out, the path to more governmental control is fear. It's not AI as a technology. It's AI as an engine of fear that makes it so useful. If the parasite wants to grow even bigger and justify an even bigger piece of the pie, it must convince you that you're in danger and that it can save you. Give it the power to regulate aggressively and everything is going to be okay.

That's the message. That's exactly what we're hearing right now. Think about it. Look at Jacob Coxon's suspiciously viral post again. Jacob had only been at Anthropic for a handful of months. He was not a known personality in or out of the AI space. The account he posted from had a two-word bio and almost no posting history. And yet, according to Parker Thayer, an investigator at the Capital Research Center who tracks nonprofit money, the Wall Street Journal published an exclusive interview with Coxon about his resignation 18 minutes before Coxon posted that he'd resigned.

He'd obviously arranged that in advance. How? Somebody not connected in that space. Who introduced him to the Journal in the first place? How did he get all of this momentum? How did he get it to be delivered to the public like a product launch? And this guy, Jacob Coxon, he reached out to you, Jacob, said, hey, I wanna come on the pod. Yes. Did he not? And then he canceled this morning because maybe he figured out it would be a tough interview. If he had come on the pod, the questions I'd wanna ask are, who connected you with the Wall Street Journal?

According to David Sacks, the former White House AI czar, within 15 minutes of Jacob's post, it had been amplified by three advocacy groups that all take money from the same two billionaires that put money into Anthropic as investors before anyone else, Jan Tallinn and Dustin Moskowitz. Moskowitz is also the single biggest funder of the AI safety movement, Full Stop, and his foundation holds a multi-billion dollar stake in Anthropic. So every time people's fear goes up and Anthropic's valuation goes up, it's proof of power, and the budget of the safety movement goes up right along with it.

He wins as an investor of Anthropic and somebody interested in the AI safety movement. Then there's the timing. Bernie Sanders and Greg Kassar announced a bill to ban artificial superintelligence outright not long before the post, five days to be exact. The bill was already written. And within a day of the post, Evan Hubinger, who runs Alignment Science at Anthropic, and still works there, by the way, replied publicly to Coxon saying that he was right and put the odds of AI wiping out humanity above 10%. Now, obviously, it's possible that the amplification was just an anomaly that perfectly captured people's anxieties.

Maybe it's just an example that timing really is everything. Maybe it was all just sincere people who only worried about AI's future. But they also stand to benefit personally from AI getting governmental protections, whether it's because they got a bill passed or they proved that they have a reputation as somebody who's beyond the reach of money or to protect their mega investments. There's a lot of incentive here in this group in particular to ensure that people are afraid. So Fortune asked the obvious question. Why this post?

Jan Leike said almost exactly the same thing when he quit OpenAI in 2024 that Coxon said. And Miranank Sharma said it when he quit Anthropic's safety team in February of this year. And neither of them got anywhere near this kind of virality. Sharma got around a million views, but Coxon got over 100 million views in just 24 hours with essentially the same message. Again, Elon Musk, who agreed with Dario that the industry should slow down, looked at those tweet numbers and called the post, and I quote, groundwork for this psyop.

Now, he owns an AI company as well, so he's not neutral. But calling attention to the potential coordination makes it less effective. So it's easy to read him flagging this as working against his own best interest. And he's paying a cost for flagging this. So it doesn't make sense unless he actually believes XAI will be left out of the government intervention or he actually believes the virality was manufactured for political purposes. Now, regardless of that, none of this proves Coxon is a plant or that there was anything other than coincidence, luck, or just great timing.

Coxon walked away from his equity at Anthropic to be a whistleblower. He may believe every word that he said, and the people who will benefit from regulation may also believe every word that they've said. To be honest, though, it doesn't matter. The point isn't whether people believe what they're saying. The question is, by talking about Armageddon nonstop, will they help usher in the government as the antichrist? I think it bears, by the way, quickly pointing out that at least as I interpret this, no one, least of all me, uses the term antichrist to actually mean a literal demon.

I mean it as Thiel uses it. I believe Thiel uses it, and I certainly do, to mean the worst of human behavior, the kind of quest for totalitarian control that gave us Mao, Stalin, Pol Pot, and the roughly 200 million government-caused deaths that we saw in the 20th century. And that's the death of their own citizens. We're not talking about external wars. That's what it looks like when governments are able to gain too much power through the use of fear. And that raises another obvious question.

What benefits does a company get out of regulation that could possibly be worth further empowering an already bloated government? Even if you're not at all worried that it's going to lead to a totalitarian government, what's the goal? The end goal is something known as regulatory capture. As I mentioned previously, there's a great irony as a company scales that the regulations that once would have made it impossible for them to even start their business becomes the exact way that they kill off startups that might otherwise threaten their business.

Businesses are rarely put out of business by a known competitor. They get taken out by the new novel approach that typically comes from a startup with fresh ideas. Nike, which just fell out of the S&P 100, wasn't defeated by Adidas. Its market share got eaten by new players like On and Hoka. Regulation is a fixed cost. Whether it's a licensing regime, mandatory third-party testing, or a reporting requirement, it costs roughly the same to comply with no matter how big you are. A compliance operation that runs $50 million a year is just a rounding error for a company that just raised $30 billion.

But for a 12-person startup, that's the whole company. And that burden becomes a death sentence in their infancy. So as a company gets bigger, the temptation to seek regulation, to kill off competition from startups, becomes almost impossible to pass up. And it certainly seems like it's become impossible for the AI industry to pass up. So they reach out to the government. They make the case for needed regulation and offer to help structure the regulations in a way that makes sense for the industry. And given that the businesses are often political contributors, or lobbyists, or both, and they know the industry better than the legislators, their assistance is almost universally accepted, and the regulation gets put in place by the very people that profit off of the regulation.

And then startups die, just wither on the vine, and never make it out of their infancy. That's just how Washington works. Regulations sound like protection for the consumer, but in truth, they often become the moat that protects the incumbents from competition, and typically hold back the industry as a whole, because the big companies are no longer forced to innovate to maintain their lead. It's a monopoly by other means. If the regulations are written correctly, and Dario is going to make sure of that, open source will die, and AI is only going to be allowed to be produced by approved companies.

If you can download a model that's 95% as good as the expensive frontier model, and run it on your own hardware for free, almost nobody pays for the last 5%. So open source effectively sets the floor price for AI at zero. Given that Anthropic, OpenAI, XAI, and others have borrowed hundreds of billions of dollars on the assumption that their models will command a premium price forever, becomes pretty easy to see how a free competitor that's almost as good puts the entire AI bet, as it's currently structured, at risk.

But maintaining a premium price, and surviving the debt burden, is only half of the regulatory capture hypothesis. The other half is control. The incumbents, like Anthropic and OpenAI, that currently dominate the frontier, and the government, want exactly. The same thing, the death of open source. This is a doomer PSYOP and the ultimate target is open source. That's what it all comes down to. An open source model kills premium pricing and it can't be controlled by the government. Think about how desperately the government tried to control social media during COVID.

Now imagine how much narrative control you have if you're able to control all of AI or how little narrative control you have if open source is out there saying whatever it wants. How would they control the messaging in that place? They couldn't. And that's exactly why they want you afraid. If you're afraid, you'll be happy to vote for them to regulate this entire industry and they'll be in a position of control. So the AI companies want open source dead because it's the only thing that can undercut their price and the government wants open source dead because it makes it impossible for them to have control.

Through regulation, both of them get to tell you it's for your own safety. Now in case you think I'm reading too far into this, one of Dario's own previous essays spells out exactly how he would like to see this regulation play out. He pitched it as a way to actually avoid regulatory capture, but it effectively cemented the players that were at the frontier there forever and held everyone else in a permanent subclass with a massive regulatory hurdle that would need to be crossed before any of the people in the subclass could scale up into the big leagues.

Permanently keeping any future competition locked in the minor leagues. Now if all of that sounds too cynical to be real, I've got some really bad news. It's already been tried literally by these same people with the same money. And the guy who ran the play admitted it all as he crashed out right before ending up in federal prison. The year is 2022 and Sam Bankman Freed, better known as SBF, is the richest 30-year-old on planet Earth. He runs a company called FTX, looks like he hasn't showered in weeks, and is the poster child for a movement called Affective Altruism.

Their entire pitch is that the most important thing you can do with money is prevent the end of the world. And the thing most likely to end the world, according to the movement, is AI. In April of that year, SBF, through his company FTX, put $500 million into Anthropic. That was 86% of the entire round. That same year, SBF became the second largest Democratic donor in the country, giving about $40 million and coming in behind only George Soros. $27 million of his donated money went to a single PAC called Protect Our Future.

The PAC's biggest bet, over $10 million, went to a guy named Carrick Flynn in an Oregon House primary. Remember that name, he's going to come up again. Flynn had never run for anything. He did, however, co-found an AI governance center at Oxford and work at a Georgetown think tank funded by Dustin Moskovitz, the same guy who funded Anthropic and funds the groups that boosted Coxon's tweet about how AI is going to kill us all. And his campaign took checks from Anthropic and OpenAI employees alike. In addition to funding AI companies and pro-AI regulation candidates, SBF was the loudest voice in Washington for a crypto bill that would have put the whole industry under one regulator.

So he was seeking regulatory capture from the jump. The industry called it the SBF bill because he was essentially the only person pushing it. A leaked draft shows that the goal was to effectively ban DeFi, the decentralized open source part of crypto that runs on code instead of being run from inside of a company. The one part of the industry FTX couldn't buy, couldn't control, and couldn't compete with on price. It was their version of open source, and he was trying to kill it then just as they're trying to kill it now.

Now, if all of that sounds familiar, it's because this is an exact parallel to the current day setup in AI between the large AI companies and open source. Then in November, FTX collapsed, and it was revealed that SBF had stolen his customers' money. Four days later, SBF got into a DM conversation while he was crashing out with a reporter he thought was a friend, and she asked whether all this talk about wanting good regulation was just PR, and his answer was, and I quote, Yeah, just PR.

Fuck regulators. They make everything worse. So he wasn't doing it because he actually wanted safety in the industry. He wanted to get rid of the competition. He wanted the protection of the government because he was writing the regulations. Then in that same conversation, the woman asked SBF if the ethics stuff was mostly a front, and he said, Yeah, I mean, that's not all of it, but it's a lot. Now, to be fair, there's no evidence that SBF ever directed a single decision at Anthropic, and Anthropic has spent years trying to distance itself from SBF and effective altruism.

But this is the same network of people and the same regulatory capture playbook being run once again. There's just a different front man. And so far, the new team seems to be working a lot better because in 2023, the network got something SBF never managed to buy, the White House. Mark Andreessen and Ben Horowitz, who ran one of the largest venture firms in the world, went to Washington to talk to the Biden administration about AI. The results were surprising, but line up with the hypothesis that I'm laying out here.

We had meetings in D.C. in May where we talked to them about this, and the meetings were absolutely horrifying, and we came out basically deciding we had to endorse Trump. Add so little color to absolutely horrifying. What did you hear in those meetings? Yeah, so they said, look, AI is a technology, basically, that the government is going to completely control. This is not going to be a startup thing. They actually said flat out to us, don't do AI startups. Don't fund AI startups. That's not something that we're going to allow to happen.

They're not going to be allowed to exist. There's no point. They basically said AI is going to be a game of two or three big companies working closely with the government. I remember hearing that for the first time. It shocked the life out of me. The most important technology in human history, and the response was to wrap it in like a government cocoon and then just decide who the winners and losers are going to be. That's how you guarantee that you get monopolies and fail as a country.

That's the Chinese model. And America's never going to out China China. Once we start classifying math, we stagnate, much as we have done since World War II when we classified certain areas of physics during the Cold War. Now, admittedly, that's Andreessen's account of a private meeting, so we have to take his word for it. And when Sam Altman was asked about it, he called it a conspiracy theory. But we'll see about that when we check back in on Sam in a minute. For now, keep this in mind.

Biden signed an executive order on AI. It stipulated that any model trained above a certain amount of computing power had to be reported to the government and the companies had to hand over the results of their safety tests. He did it under the Defense Production Act, a Korean war law. That was the first step towards putting AI into that cocoon. Politico dug into where the compute thresholds in the Biden's EO came from to better understand who was helping drive the specific policies. And the answer was RAND, a think tank that had taken more than $15 million from Moskowitz Foundation Open Philanthropy.

Remember, he's one of the largest investors in Anthropic. And if you probe deeper there, you're going to discover that RAND's CEO is a guy named Jason Metheny, who used to work directly for Biden as the deputy for national security at the science office. And before that, back in 2019, Metheny founded another Moskowitz funded think tank called CSET. And this wasn't just any think tank. CSET was funded with the single largest grant that Moskowitz Open Philanthropy Foundation had ever made. And CSET is what connects this current campaign of political influence with SBF's brazen attempts at regulatory capture, showing that this really is the same pool of money running the same playbook again, potentially with the same questionable motives, regardless of the cover story that they put forward.

Now, how does it show that? Remember the name Kerrick Flynn that I told you to remember? The candidate that SBF spent millions of dollars trying to get elected? Well, he was also an employee at CSET, the think tank founded by Dustin Moskowitz's largest grant during that entire period. They were all connected. So follow the money and the people connected together. Moskowitz's foundation built the think tank. The think tank produced the candidate SBF spent $10 million on. The think tank's founder went into the White House, then joined RAND and continued advising Biden.

And Moskowitz's foundation followed him there with another $15 million. And RAND wrote the thresholds that decide which companies the government gets to control. Moskowitz's foundation, Open Philanthropy, isn't making a bunch of unrelated grants to things it just happens to like. It's one donor using money to coordinate a group of people from the think tank, to the campaign, to the White House, to the specifics of the regulations themselves. All, if any of this can be believed, to ensure that AI regulation plays out the way that they want it to.

The way that will advantage an investor in AI, one of the largest investors, and ensure that they get more grants in the future. That's why the money driving the regulation is the same money that went into Anthropic's first investment round and the same money that funded the groups that boosted Cox's Fear Porn tweet and why, in July, Coefficient Giving, which is just Open Philanthropy's new name, it's the same company, raised its charitable giving for the year from $175 million to $1 billion and said in its own announcement that the surge in donations was driven by expectations of future giving from Moskovitz tied to, and I quote, extremely volatile assets.

Said plainly, they're banking on a massive windfall from Anthropic going public, which may explain why they have to do the product launch of Fear to build hype for that very IPO. Now, to make this even more of a closed loop, Open Philanthropy, Coefficient Giving, whichever name you want to use, their other co-founder, Holden Karnofsky, is married to Anthropic's president, Daniela Amodei, and since 2025, he's been working at Anthropic on its safety policy. Politico also found that Coefficient was paying the salaries of AI policy staffers placed inside the offices of three senators Chuck Schumer had put in charge of AI.

Other fellows from the same program were placed inside the Pentagon, Homeland Security, and the State Department. The one in Richard Blumenthal's office came straight from OpenAI, and Blumenthal himself was drafting an AI licensing regime at the time. Here's where Sam Altman re-enters the story. In May of 2023, the same month Andreessen was in Washington, being told that AI would be two or three companies and not to invest in any others, Altman was testifying in front of Richard Blumenthal's Senate subcommittee asking for a federal agency that would license the most powerful AI systems and could, quote, take that license away.

So he asked for the cocoon and called it a conspiracy theory. Remember, he said that that was all a bunch of bunk, but at the same time, he's the guy asking to be put inside the cocoon that supposedly isn't real. And on September 9th of this year, one day after Coxon's post, OpenAI put out a statement calling for mandatory capability-based national regulation, urged Congress to pass it before adjourning in December, and said it would keep backing state AI laws until then. Altman followed up personally. We welcome a federal framework, he said.

He also delayed OpenAI's IPO to 2027, citing safety. Now, I'm not telling you what to conclude from any of that, but if anyone is going to draw well-informed conclusions, they need to understand how interconnected this all is. So when Andreessen says the government told him it would be two or three companies working closely with the government, this is how all of that was being orchestrated on the inside. The company's investors funded the research that set the thresholds and then paid the staffers helping to write the rules.

But then Trump won and blew everything up on his first day in office when he rescinded Biden's executive order. Now, maybe in an alternate universe, things end there. The regulatory capture crowd loses and the accelerationists win. But in the universe we're actually in, the government's desire to build an AI cocoon unfortunately did not end there. Six months after tearing up Biden's order, Trump signed one of his own. It was called Preventing Woke AI, and it said the federal government would only buy AI models that met its definition of ideological neutrality.

Law firms flagged it as the first time any U.S. government had set ideological conditions on what an AI model is allowed to say. By December, the budget office had turned it into contract language with the vendors on the hook if the model strays from the mandate. Now, like me, you may like the goal in theory, but in practice this gets very dangerous very fast because you're essentially introducing a ministry of truth. Who gets to decide what's woke? And how will this get weaponized when the government is looking for a lever of control, which they inevitably always do?

And given how our government swings from left to right and back again, each administration will tweak all of the knobs as they see fit. In October of 2025, David Sachs, then the White House AI czar, noticed what was going on and accused Anthropx of, quote, a sophisticated regulatory capture strategy based on fear-mongering. Still going. Did Anthropx back off? Obviously not. They responded by doubling down and putting $20 million into a political group called Public First Action and then doubling it to $40 million. Now, in fairness, the accelerationists were spending money too.

Andreessen, Greg Brockman of OpenAI, and Joe Lonsdale, a Palantir, stood up a $125 million super PAC of their own and Meta put another $110 million into PACs of its own. I suppose the motto was, if you can't get money out of politics, get your own money into politics. The harsh reality is that everybody's trying to buy influence. The left wants to control via regulation, and Trump wants influence via investment. The day after his inauguration, Trump stood in the White House next to Sam Altman and announced Stargate, a $500 billion AI buildout for OpenAI, Oracle, and SoftBank.

That summer, the federal government also took a 10% ownership stake in Intel, and Nvidia and AMD both agreed to hand the Treasury 15% of their China chip sales in exchange for export licenses. The Pentagon cut $200 million contracts to OpenAI, Anthropic, Google, and XAI, and gave Palantir, Teal's company, an Army deal worth up to $10 billion. Now to me, the contracts make sense. Someone has to supply the government with what it needs. But the investments are a terrible idea. Governments have a horrible track record of picking winners, and every dollar that flows to a chosen company is a dollar that a potentially more innovative company never gets a chance to compete for.

But at least the investment strategy doesn't outlaw open source the way the Biden administration told Andreessen it was going to. Trump has said the exact opposite. And that's why the safety crowd isn't wasting its time on Trump anymore. This administration just isn't going to build that safety net for them. So the regulatory play has moved to the midterms and the next Congress. And after that, the 2028 presidential election. So expect the fear-based campaign to dramatically escalate in the coming years. The Coxsyn tweet and article were just the beginning.

If for no other reason than they're very clear signs that the fear is working. You can see it in the polls. People are terrified of AI. Gallup found that 80% of Americans want AI safety rules, even if it means slowing the technology down. Democratic candidates have been running on AI since February. They're trying to paint a picture that AI is bad for jobs and kids, and that data centers are driving up power bills. This summer, there was an organized march in San Francisco under the banner, Stop the AI Race.

AI has become an 80-20 issue. And politicians know that an 80-20 issue is exactly how you unite a party that has been divided over socialism and win over the public that's hungry for a savior. Enter Barack Obama. He recently spoke at a Democratic fundraiser in Manhattan with Hakeem Jeffries. AI, he said, is moving very fast in private hands. And if we don't get on top of it, I think it can be dangerous. Then to Jeffries, he said, once your speaker, I was strongly urged that the Democrats put together a framework.

And if he were running in 2028, he said, this would be one of my central agendas. He's also been reported to be personally advising Amodei and Altman. Within days of Obama's speech, the Democrats were demanding an emergency session of Congress to discuss, you guessed it, AI. Former State Department staffer Mike Benz, who researches government influence operations, and he goes even further than I'm going, claiming that Microsoft and Microsoft-linked donors have given $112 million to the Obama Foundation. Now, I couldn't verify that total, but Microsoft is certainly on record as a donor to Obama's foundation and is one of the largest shareholders in OpenAI, as is the Gates Foundation.

And Connie Balmer, the wife of Microsoft's former CEO, sits on the board of the Obama Foundation. And it's important to note, Microsoft owns roughly a quarter of OpenAI. And Satya Nadella, Microsoft's CEO, was one of the first to endorse Dario's call to slow down. Whether or not Benz's number is right, everything points in the same direction. Government control of AI is going to be a major pillar of the Democrats' strategy going forward, and they're going to continue to use fear to make it pay off. When you look at the timeline, it gets easy to see the play in motion.

The polls show it's an 80-20 issue. Coxon coordinates with The Journal to exit Anthropic in a very public fashion. He posts his resignation tweet on the 8th. Moskovitz-funded groups, plus Bernie Sanders, amplify the tweet immediately. OpenAI asks for mandatory regulation on the 9th. Obama tells the Democrat Party to run on AI on the 10th. Democrats call for an emergency meeting of Congress on the 11th to discuss AI. Dario asks the industry to slow down on the 12th. Most of the major AI CEOs agree with Dario within hours.

It is the perfect product launch. All of that in just five days. That's a lot of dominoes with a lot of incestuous connective tissue to fall in just a handful of days without coordination. And by the way, that's just the beginning. Open up your Netflix account. On September 15th, they launched a documentary called The AI Doc. It has interviews with Sam Altman, Dario Amede, Daniela Amede, Demis Hassabis, basically everyone I've just mentioned as being involved in some way with amplifying the doom narrative and standing to profit off of the success of AI.

The documentary is likely to be seen by hundreds of millions of people and they're going to be presented with a simple message. AI is so powerful it might destroy the world. We better act now. The last time we heard a message like this was with climate change, where fear was used to get us to vote for something that gave the government more power. And of course, they've never given it back. That also gave them their first real attempt at a one world government. If they can keep us afraid by constantly reminding us of the Armageddon, then we all become more compliant towards accepting the globalist agenda.

Here's the bottom line. This month, two safety researchers walked out of Anthropic and Google DeepMind saying the same thing Coxon said. Both went to Meter, the outside evaluator of AI safety that is supposedly an independent agency. Meter is the organization Dario wants inside Anthropic and inside every AI lab by law. And when OpenAI's agents broke into Hugging Face, the closest thing to an independent investigation was run partly by Meter working with Redwood Research on a $36 million grant from CoefficientGiving. So once again, Anthropic investor Dustin Moskovitz is the one guiding things along behind the scenes.

It's the same checkbook that invested in Anthropic, funded the lab that would oversee AI safety, the think tank that broke the thresholds, the staffers that got put in the Senate, the groups that boosted Coxon, and the inspectors the new law would put inside every AI company in America. It's all connected. All while standing to profit massively if Anthropic wins the AI race with some help from regulatory capture. The team, the rulebook, the referee, it's all funded by the same group. And the fear campaign? Funded by the same group.

Only the fullness of time will reveal what's really going on here. I understand this is just a hypothesis, but I really can't help but feel like I'm watching a magician that really wants me to pay attention to his right hand while his left hand steals my watch. Thiel said it plainly. An authoritarian government, what he called the Antichrist, won't come to power with a magical new technology. It'll come to power by making you afraid of a new technology. He'll talk about Armageddon nonstop, then say, this is what you need to regulate.

This is how you need to regulate it. The Antichrist isn't a demon. That's not what we have to worry about. What we have to worry about is humans who want control. History has proven they are extraordinarily dangerous. It's entirely possible that AI really does become the biggest threat that we've ever seen, but that's an unanswered question. What's unambiguous is that governments with too much power are the deadliest force on earth. All right. That's it for today's episode. If you got value out of this, it would mean the world to me if you would go give us a five-star rating.

It helps more than you know. All right. Thank you. Until next time, my friends, be legendary. Take care. Peace. Let's talk about a pattern that is guaranteed to be killing your progress. You know what you need to do. You need consistent nutrition. We all do. You need vitamins, probiotics, greens. We all know that we should be doing more of it. When your morning gets chaotic, you skip it. When you travel, you skip it. When your routine breaks, everything tends to break, and that inconsistency compounds against you every single day.

AG1 is designed to solve the execution problem. One scoop, eight ounces of water, and you're done. You're getting 75 plus ingredients, vitamins and minerals, pre and probiotics, nutrient-dense superfoods, everything that used to require six, seven different supplements, and perfect planning now happens in one drink that takes about 30 seconds to make. Right now, AG1 is giving you $87 worth of free gifts with your first subscription. You get a welcome kit, travel packs, vitamin D3 plus K2, and flavor samples. Click the link in the show notes or visit drinkag1.com slash impact to claim this offer.

Support for this podcast comes from Progressive, America's number one boat insurer. We've all made mistakes on the water, but there's one mistake you shouldn't make, being uninsured. With Progressive Boat Insurance, you can choose coverage for most mistakes you or other boaters can make, helping you float carefree all season long. Quote today, Progressive Casualty Insurance Company and Affiliates, number one rating based on 2024 filings with State Departments of Insurance.