OpenAI 的身份危机、数据中心大战、伊朗消息推动市场上涨、Mamdani 的首项税收、Swalwell 出局

All right, everybody. Welcome back to the number one podcast in the world. We've got the core four here, and, dare I say, the king of Adams. The king of Adams, yes. Captain Travis Kalanick is here. How are you doing, brother?

I'm pretty good. Pretty good. I'm sitting here doing the podcast just next door to David.

Ah, yes, there you go. Don't reveal our locations. Please don't dox.

Put my address out there, dude. No, that's true.

Don't worry, Mamdani did. He's outside your house right now asking them to foreclose on you.

3.9% or something? Yeah, he decided there's no rich people left in New York, so he's looking for other cities to tax.

Is it 3.9% a year? Is it per year, or is it—

I don't know if the percentage has been released yet, but the speculation I've seen is 3.9%, and I don't think that's final. But yeah, it's a pied-à-terre tax, so if you have a second home, it's every year.

Wow. And, by the way, it's for any home over $5 million. There are no homes under $5 million in Manhattan. This is not a rich-person tax. If you're within 15 miles of Midtown Manhattan, you're paying an extra tax.

I don't know.

But only if it's a pied-à-terre, J Cal. What it means is that the most elastic part of the market is what they're targeting for this tax. In other words, people who don't live in New York, who just have it as a second or third home, could buy that property anywhere.

Yeah, they are now being taxed the most. So what do you think that's going to do? It's going to have a massive impact on demand for second homes in New York, which will crash the whole market.

Yes. Congratulations, Mamdani. But in a weird way, that'll be good for housing affordability in New York.

Well, that's sort of the claim, but I don't think it'll be good for it because there'll be no incentive to build more.

Yeah. All units matter. Every time you add units, people upgrade, and it's not like these are going to be low-income housing. A penthouse on 57th Street or in Gramercy isn't low-income housing. You'd have to break it into 7 units. It makes no sense.

By the way, I don't know if you guys saw the video—not to get too serious—but he's doxxing a certain billionaire who owns a certain place, and he's literally pointing at his home.

No, I said that to you, Jason. He's not doxxing him because everybody's known for years that Ken Griffin bought that place. Everybody knows that address. Everybody knows that unit. We all knew it. It was marketed widely. I don't think that's really doxxing. He doesn't live there, and everybody knew he owned it. It would be very different if it was a place where somebody was keeping their primary residence, and you didn't know, and they stood in front of the house. That, I would agree with you. I think this one is a little bit more tenuous.

Okay, fair enough. But what I will tell you is that it's a dog whistle. Crazy people—and this thing's been seen by 30, 40, 50 million people now—it's a dog whistle.

That's true. That's true.

That's the next UnitedHealthcare CEO. And in the week that a Molotov cocktail and a bullet get shot into Sam Altman's house, it's deadly serious. If you reversed it—I always reverse it—what if a Republican sat outside of Bernie Sanders' second or third home and said, "This is his pied-à-terre. This is his summer home. We should add taxes to it"? I'm sure you could look up Bernie Sanders' home pretty easily. So just before you point at people's homes and say, "This is the villain," be careful, folks, because then if something does happen to that person, like what happened to Sam Altman this week, you can feel however you want about him, but nobody deserves to have their house firebombed or shot at. Period, full stop. I just think that if you were a person who was thinking about buying a pied-à-terre in New York, there's no way you would do it now.

Because you don't know what the tax rate's going to be, and it's going to keep going up. After a decade, with interest and inflation, you've effectively almost doubled the price of your unit. So if you're going to buy a $10 million unit, you're probably then looking at a $20 million purchase price just to break even after about 10 or 11 years. That's crazy. The math doesn't work anymore.

I just think that it's going to kill the demand for, you know, maybe people who already have a home in New York like Ken Griffin, they probably are just going to suck it up and pay the tax and keep whatever they have, but if you were a person who was thinking about buying a pied-à-terre in New York, you there's no way you would do it now.

What is the downstream effect of these individuals not coming to New York, going to a Knicks game, or going to a restaurant?

No, they spend money. They spend money in New York. They have their birthday parties in New York. They do all kinds of things in New York. I'm not a big fan of the pied-à-terre culture. I own one house. I don't like this whole multiple-houses-you-flitter-flutter-everywhere thing. It ruined London. When you look at London—

You vacation? Chamath, where do you vacation?

I go to a hotel.

Okay. So my point is this: I think the way that it works is that it's not just pied-à-terres. If you own a home that you fully rent out, but that is not the primary residence of the person who is renting it, it's also a problem. So this is any—even if you had a 2-week rental, a 1-week rental, a 30-day rental, or even a 5-year rental, and it was somebody else's second home, the owner still gets hit. The rentals of folks who want to come to that city or who don't want to own in that city, but use it as a second place—

The second-place thing goes away. So maybe this is just good for hotels.

I think it's good for hotels, but if you look at London, it's probably the best example, where there were a lot of people who used London as a place to store assets. Real estate became the primary way in which they would do that. It hollowed out parts of London. It's not as if they were unlivable; it's just that they were unlived in. Nobody was there. You'd drive around Chelsea and certain parts of London, and it was like a ghost town on a Thursday or Friday night. That is an issue. It's land banking. It's like your Bitcoin is hollowing out a neighborhood. There is something there.

But you know what? The whole housing thing is complete crap and utter bullshit, because if you move to a place like Austin, or if you're in Nevada or Florida, you see what happens when you allow people to build units. In Austin, for 3 years in a row, rents and housing prices have gone down while net migration has gone up.

But that's been good, right, for Austin? You guys would say? Net good?

Incredible. Yeah, amazing. Austin has roughly doubled as a city over the past decade, and yet the rent for one- or two-bedroom apartments has gone down. That's incredible.

So, in other words, if you let people build to satisfy the demand, you won't have this problem. And who's stopping the building? It's Democratic cities. It's NIMBY people. In a Republican town, they're actually building units for affordability. So you have one group saying they care about affordability, and they're doing nothing about it and stopping it. And in another place, they're like, "We're just going to let you build because it's your right to build, because it's your land. Go." That's the approach in Texas. It's your land; you have the right to develop it. Go.

By the way, the high end of the market in London has basically turned over and collapsed.

Sacks, to your point, they introduced the stamp duty, which I think is equivalent to this tax that the New York City mayor is proposing. If you look at London as a guide, the real estate market just bit it at the high end. And I don't think that's going to be good for that city or even the UK as a whole. There are a handful of cities where people do what you say, which is park money there. The reason why they do that is because they believe that the city has the rule of law and is a unique, world-class city that's going to keep appreciating. If you have new management that doesn't really believe in the rule of law and keeps imposing all these arbitrary taxes, that money is going to flee and find other kinds of investments. People aren't going to park their money there. That has to be a bad thing for the city.

It's like, who cares if the top floor of that building where Ken Griffin lives is owned by 1 guy who isn't there that much? It's not going to affect the city that much. But having all these billionaires from all over the world decide to park money in that American city has to be good for the US, just like it's good for the UK. And if you give that up, then, again, the money will just go somewhere else.

Well, to your point, the other thing is that they're already paying taxes on the property. And because they're not there very often, they're not using city services. So they're paying taxes on the property, and they're not using city services. They're profitable to the city.

Think about a developer who's underwriting some new project. The fact that a whale like Ken Griffin is willing to overpay, in the sense of price per foot, for that top floor might make that whole project pencil. This is what contributes to the vitality of New York. There's constantly development going on. There are constantly cranes. So you take out that part of the market that, in effect, was price-insensitive and was subsidizing all these projects, and I think development's going to dry up to a large degree.

Mm-hmm. He must be paying $3 million or $4 million in taxes every year and getting no services, to your point, Travis. Good point. That's all profit for the city.

London also did something else, which is that they essentially crippled what's called non-dom status, which is the big tax arbitrage if you're moving or parking assets in London.

And to your point, Sacks, what did all the rich people do? They just redirected themselves to Zurich, Lugano, and Milan. They took advantage of more hospitable tax policies in other places.

Now, what the people in Britain would tell you—I actually met with the UK government yesterday—is that they don't see a measurable impact yet that's meaningful enough for them to think it's a five-alarm fire. So the real question is: Is it more of a slow bleed and a slow melt, and at some point it's just hollowed out and very hard to reverse? There's not going to be a cataclysmic, acute moment where people say, “Oh my God, we need to reverse these policies.” It doesn't seem like that's in the offing.

Los Angeles did something a little different but similar: It introduced that 5% mansion tax. It's on all the areas except for Beverly Hills, and San Francisco has that, too. Yeah, San Francisco got that, too. I think it's actually 6% in San Francisco.

It's 6% over $25 million, and it starts at $5 million. So there's a transfer tax—an excess transfer tax on properties above $5 million—that scales up above $25 million to an extra 6% on top of your brokerage commission. So when you sell a house in San Francisco, you're paying 13% now?

Which is why you look at the transaction volume in the Los Angeles real estate market, and it's just completely dried up. People aren't doing house flipping anymore, that kind of stuff, because the transaction costs are too high.

But it just shows—I mean, again, this tax was imposed retroactively. For example, I had my house in SF, and then they just take a couple of rooms of it. We talked about this on the show when it happened. My point is just: Your property is not safe in blue states.

Wealthy people who have a choice of where to park their money are going to increasingly realize that, and they're not going to buy—I think real estate in blue states is dangerous because the political class thinks that they can take a chunk of it. The wealthy people are going to react to that, and they're going to move their money elsewhere.

All right, let's go to topic number 1. OpenAI is apparently suffering from a bit of an identity crisis. On Sunday, OpenAI's Chief Revenue Officer, Denise Dresser, sent a 4-page memo to employees. Obviously, it leaked immediately—probably the point of it.

She called out Anthropic. She said their $30 billion run rate is cap, inflated by $8 billion due to a revenue share and some accounting with AI model providers. Chamath, you pointed that out over the last couple of weeks. She also said Anthropic's story is built on, quote, “fear, restriction, and the idea that a small group of elites should control AI.” Obviously, she's a fan of the pod.

She also laid out OpenAI's pivot, and she said they're going hard after business customers and want to win the agent-platform layer. If you remember, they hired the architect of the open-source project OpenClaw. They didn't acquire OpenClaw, so Peter Steinberger is working at OpenAI. Cynical people said, “Hey, maybe they want his next set of innovations to go inside OpenAI's products as opposed to the open-source one.” I kind of agree with that directionally.

There's obviously Perplexity Computer, doing really well. They quadrupled their revenue. I was over at xAI earlier this week with Elon, and I can tell you he's got some very cool stuff coming. This new model, Spud, is coming from OpenAI. Here's your Polymarket: 75% chance Spud is released next week.

Additionally, on Tuesday—2 days after this memo came out—we'll go to the panel in just a second. Obviously, these memos go directly to the press, so they're obviously trying to undercut Anthropic's valuation, and they feel that's a threat. That's my take.

The FT cited anonymous OpenAI investors who are frustrated with the company's lack of focus. Here's the anonymous quote: “You have ChatGPT, a 1 billion-user business growing 50% to 100% a year. What are you doing talking about enterprise and code? It's a deeply unfocused company.”

And we talked about this: ChatGPT's market share is going down as the number of users is going up because Gemini and Claude are gaining significantly. Meta just last week released its first proprietary model. Apple doesn't have a product in the market yet, but it does have a lot of users.

So here's your generative AI website traffic. Let's start, Chamath, with you. Your thoughts on OpenAI: Should they be pivoting straight into business and developers and getting focused on that, or should they stay focused on the consumer, where they are the verb?

As it relates to complex, long-horizon coding tasks, what I can tell you from my team at 8090 is Codex is generally better than Anthropic. So what happens is, for more day-to-day work, I think it's more reliable to use the Claude ensemble of models. But when you're dealing with something that's very tricky and complicated and a little more long-horizon, Codex is really functional and really good.

If you're looking at it through the lens of OpenAI, what they're probably saying is, “Hey, hold on a second. If we allocate our resources and just double down and crush consumer, that's probably $3 trillion or $4 trillion of enterprise value. And then if we slowly refocus the company with the rest of the resources and double down on Codex and do something meaningful in enterprise, we can probably capture $2 trillion or $3 trillion there.”

Now, all of a sudden, you can paint a picture for a $7 trillion, $8 trillion, or $9 trillion market cap in the fullness of time—not tomorrow, obviously. But Codex is really good, and there's a business to be had in both.

You have to separate the 2 businesses. You can't have a lot of overlap because there's too much context switching. You have to let the consumer team run, and then isolate the enterprise team and let them do what they think is right.

Travis, there's a big debate going on among investors. The FT piece also questioned the $850 billion valuation of OpenAI. Secondary markets have now priced Anthropic higher than OpenAI for the first time. This is the flippening that people predicted.

One investor said OpenAI would need to IPO at a valuation of $1.2 trillion for the last round to make any sense. But there are currently no buyers at the $850 billion valuation that OpenAI just closed, according to Bloomberg. Travis, how do you handicap this race between these 2 leading frontier models?

Growth is king right now in this world, in this segment. Growth is the whole damn thing. If Anthropic is growing faster than OpenAI by a significant clip, investors right now are going to play it forward.

You start to get network effects around compute, network effects around the number of tokens you're pushing out for various customers, enterprise or consumer. Ultimately, it's not great today, but it's about how that plays into reinforcement learning and things getting smarter over time.

There's so much upside to volume and scale that if they're growing faster at the same size—even if OpenAI is still growing—but if they are at half the growth rate, a third the growth rate, or a fifth the growth rate, I would be worried.

And you saw this at Uber specifically when you accelerated away from competitors like Lyft and DoorDash, yeah?

Yeah, you had to. Network effects were the whole thing at the end of the day, and network effects were based on scale. So, yes, I'm coming from my very specific experience.

But if you believe there's a network effect from the scale of data that you have, the scale of customers, and the revenue—the cash that's coming in that you redeploy into compute—and say there's a network effect from large compute, I'd be very worried if I'm OpenAI and I'm seeing somebody growing faster at the same size.

So, Friedberg, when you look at this race between these 2 giants, maybe your thoughts on the flywheel as it relates, as Travis is pointing out, to advantages in compute, reinforcement learning, and also the ability to fundraise.

One of the great things that Travis and the team did was, as they were pulling away, they just sucked all the oxygen out of the room by using capital as a weapon. So your thoughts, Dave, on this high-stakes game? Because there's also a point at which—and I'll just end on this—there's a point at which you could run off the cliff. You raise so much money and deploy it so fast, and the revenue doesn't catch up to it, and then you go public and the markets don't believe the story. So your thoughts, Friedberg.

I don't know the financials of the 2 companies well enough. Obviously, Sam has no problem raising money.

Didn't he just close something like a $150 billion round or something? How much was it?

$122 billion. The largest round ever raised in any market, I think, private or public.

Public, yeah. That doesn't seem to be an issue.

What I've noticed is just the pace of innovation at Anthropic is, in my experience, unprecedented. Their release cadence is extraordinary. They've basically supplanted OpenClaw already with this release they did a few days ago, and then today the new Opus model got dropped.

So there's something about the momentum—not necessarily just in user growth, but in how they're operating this business—that just seems to be head and shoulders above everyone else in the cadence of upgrades.

If I look back 6 months ago, I think we were pretty heavy on Cursor and Gemini, and now I think we're probably 90% Anthropic in just the last 6 months in my organization. There's something very powerful about the flywheel they have going on.

Yeah, but here's where I go, just real quick, Dave: Mad respect on a $120 billion or $130 billion raise. I mean, this is obviously next level, but you can use capital and investment to acquire scale and network effects associated with it.

But if somebody is getting that scale with revenue—and let's call it contribution margin, contribution profit—efficiency will outstrip subsidy. And it will. You can't just keep raising $100 billion things forever. That's where the train will stop. And if Anthropic is funding theirs through revenue and other folks are funding it through investment, that's a short-term solve. But in the long run, whoever is scaling their actual usage and system and ultimately generating contribution profit that then soaks up the need for investment, that's a very scary machine if you're competing against it.

Which is exactly, Sacks, what the legacy Magnificent 7 are doing. You have massive profits from Meta's core business and Google's core business that are being redeployed into infrastructure. Even Tesla, which has a lot of profits, and SpaceX, which has a lot of cash on hand, are building out Colossus and other assets, including—Elon is working on building a fab, as folks have been talking about.

So, Sacks, is there a chance for the legacy companies, the Magnificent 7, to compete in this? Or are we looking at OpenAI and Anthropic as 1 and 2, and then everybody else can fight for 3rd place and the bronze, as it were?

Well, I think Google's clearly in the mix. DeepMind has an outstanding team, and I think Elon's still in the mix with xAI. And then you've got Meta, which also has the resources. They seem to be further behind, but they're going to compete.

Look, let me just go back to the central premise here. I agree that there's some valid criticism that OpenAI has been unfocused and should be moving forward more focused in what they do. I have no idea, for example, what they're doing buying up podcasts. That's not us, so I don't know what that was about.

If you were going to buy a tech podcast for a few hundred million dollars, I mean, we were here. We're here. We're here. Dario, reach out.

Small potatoes for you guys, though. Small potatoes.

Yeah, because they can't afford you guys. They can't afford you guys.

Yeah, they thought we were too expensive. Little did they know—we would have sold out. Punch the ticket, Dario.

But look, this other part of the criticism of OpenAI—that they shouldn't do enterprise—is totally misguided. One of the reasons why they should have been more focused is to do more enterprise and get enterprise more correct. Now, why do I say that? To Travis's point about growth rates, it's true that OpenAI and Anthropic, as of the beginning of Q2—so, let's say 2 weeks ago—were both around $30 billion of revenue. And that memo from that OpenAI employee was right that if you compare them on an apples-to-apples basis, then Anthropic is about 20% less because they are including revenue made by their channel partners.

But that doesn't matter. What matters is the growth rate, again, to Travis's point. Let me just put some numbers around this. OpenAI's growth rate has been around 3–4× a year. Anthropic's growth rate has been around 10× a year. So, they went from, let's call it, $1 billion to $10 billion of ARR last year, and by the end of Q1 this year, they're already at $30 billion of—again, let's call it—their revenue.

They're on their way, as Brad Gerstner was saying on our podcast, I think in the last couple of weeks, to ending this year at $80–100 billion, at least on the current trajectory. And so you can plot their revenue on a logarithmic graph. No one's ever seen anything like this before, where every unit on the y-axis is another X—it's 10×—and it's a straight line.

Yeah, that's crazy. It's crazy, right? So, if it's taking Anthropic, let's say, 1 year to 10×, and it's taking OpenAI 2 years to achieve a 10×, then it's obvious which one is going to win. Now, what is the reason for this? It's because Anthropic was very focused on enterprise, specifically coding. And what you're seeing is that businesses are willing to pay for coding—code tokens on a metered basis, let's call it, like electricity. The more they use, the more they're willing to pay. And their usage just continues to scale and scale.

Consumer is completely different. Consumer is the thing that OpenAI prioritized. Consumers have a lower willingness to pay. Maybe only 3 or 4% of them are willing to convert to premium in the first place. And what they want is a $20-a-month, all-you-can-eat subscription. So, the revenue simply doesn't scale the same way that enterprise does. And so, if you want to tap into the scalable revenue source in the market right now, you have to go after enterprise.

Again, where I would agree with the criticism of OpenAI is maybe they should have been more focused, but they need to be more focused specifically to pursue coding and enterprise. And if they don't catch up soon, to your point, then you could see Anthropic taking a lead here that, let's say, over the next 1 or 2 years could be insurmountable.

Just by the way, let me say one thing. Even though Anthropic's revenue has followed this graph, this exponential graph, very predictably, it can't do that forever, right? Let's say it does get to maybe $100 billion this year. Can it really get to $1 trillion in revenue the year after that? It seems hard to believe, right?

And the reason is because as you hit new levels of scale, you encounter new problems. You're simply going to run out of compute or electricity, data centers, infrastructure. There are physical limits, or limits in the physical world, that you're going to hit.

And there's already some evidence that Anthropic is hitting some of those limits. Users were complaining, for example, that Claude was thinking less. Did you guys see this? A typical Claude prompt seemed to have cut down on the thinking time by about ⅔. Now, I saw someone tweeting today that they just released Opus 4.7, replacing Opus 4.6, and the thinking is back. But maybe they're charging more for that. Hard to say.

They're going to hit some sort of physical limits, and I do wonder if, over the next year, Anthropic will reconsider whether its support for all this doomer NIMBYism was the right call. It kind of made sense for them from a business standpoint when their competitors were building data centers and they were just getting compute from the hyperscalers. But now that they're, I think, going to have to move into the game of building their own data centers, they might regret salting the earth for data centers all over the country. And I wonder if that'll be the natural limit of their growth: they'll be hoisted on their own petard of doomer NIMBYism.

And they're also using the coding platform to build Anthropic itself better. So that in and of itself is a reason to nail coding. You get the double whammy: You can make a better product, and you can get paid for it.

Yes and no. Hold on, because if you look at Twitter, people are panning Claude Desktop, and what they said is, “This is all a bunch of vibe-coded slop.” I think we have to remember, if you keep these agents on task and they're properly guardrailed, this stuff is a force multiplier.

The problem is nobody knows how to do this really well yet. Nobody has built real products at scale largely using agents yet. Nobody knows how to give an example of how an org structure should be redefined. Nobody knows how to budget properly. We had the CTO of Uber say, “I give up. I've hit my token budget.”

The problems I see are twofold. The first is just to build on Sacks's point. All of these frontier labs have a very serious issue, which is that both OpenAI and Anthropic are growing so fast that they're at a point now where they need their own infrastructure. It's kind of like when you first start building any kind of company: It's just much easier to rent capacity from the hyperscalers.

And it's a dependency.

Yeah. But then it becomes a dependency. Exactly. And now, when you're so big, it's actually a strategically huge mistake not to have your own compute supply. Why? Because if you look at who's leading, the frontier labs are leading.

And, Sacks, to your point, you mentioned this on X. There was all this doomerism, but maybe it was tied to compute capacity, because when Bedrock opened up more capacity for Anthropic, all the doomerism went away. You're left wondering: Is it really tied to just the fact that they were trying to throttle usage?

So, if you're a frontier lab, you don't want to have to go through Amazon, GCP, and Azure and hold a tin cup for access and capacity. What you'd much rather do is go straight to your customer. On the other side, if you're Google, Microsoft, or Meta and you have all this compute—because I saw a stat this week that the hyperscalers control 60% of all the compute—the game theory there is, if you kneecap the frontier labs, it'll give you some chance to catch up. And it gives you time to catch up.

You guys remember, in social networking, when Friendster was the cat's meow?

Yeah, the cat's meow.

Remember what the biggest problem with Friendster was?

Friendster was slow as a dog.

Yes. Yes. And what happened? MySpace came in and took all the share. Then Facebook came in, and we took all their share.

So, there is a way where you can handicap and kneecap these companies by throttling compute access to them. So, A, they are forced to now get in the game, which is weird because, look, OpenAI has tried to displace some of the Stargate spend. I don't see any path except that they're going to have to do it themselves, and Anthropic will have to do it themselves.

But then, separately, the other problem is when you change the subscription model in enterprise and you say, “Hey, we're not going to subsidize any more tokens,” what's going to happen is all these token budgets are going to go crazy. And what Friedberg said is going to happen, where he's like, “Hey, guys, why are you spending all this money? What are you making?”

And you inspect the code and you're like, “What is this slop?” And you're not going to add 30, 40, 50% opex to produce nothing.

So, I think that's an open question, and that question will become more amplified over the next year as they push the costs off of them. As Travis said, no more subsidy from the capital. You have to grow into it, but you're not going to support negative gross margins. You're going to pass through the token costs. So, I think it's a very dynamic moment right now for these companies.

And if you look at the ranking of these clusters and who has the most, right now people have forgotten about Colossus, which Elon has been building. He's expanding to 555,000 GPUs across 3 buildings—an $18 billion investment. If you then look at Prometheus, Meta's planned 2026 cluster, that's 150,000 GPUs.

Elon just announced a deal this morning with Cursor. Elon's renting a bunch of capacity, so he's now getting effectively into the data-center business. He's going to be a hyperscaler. He'll use as much as he can for xAI, and whatever is left over he'll give to—well, in this case, he's giving it to Cursor to train their model. He could give it to others.

You might as well overbuild capacity, because that way your own models will be in a privileged position and you can sell the rest to your competitors. But, Chamath, to your point about the thing I was saying on X—I think I was retweeting Mark Andreessen, who pointed out that one of the reasons Anthropic might have wanted to hold back Mythos is that they simply didn't have the compute to serve it.

The model was huge and very expensive to serve—something like maybe even 10 or 20 times the token cost of, say, Opus. They knew Opus 4.7 was coming out, right? So they held it back, knowing that they didn't have the compute to serve it anyway, and saved their compute for the next iteration of Opus. By holding it back, they created this impression of scarcity and altruism, and it turned into this gigantic marketing event for their product because everyone in the government was like, “Oh, wow, they're holding it back because it's so amazing.”

Now, look, I think it may have been genuinely altruistic as well, in the sense that Mythos does reveal coding vulnerabilities that people didn't know about before. It makes sense to give companies with large code bases time to patch these dormant bugs and vulnerabilities. But it's looking more and more like Anthropic couldn't have offered that model commercially anyway because it was just too big and expensive, and they needed to create space for Opus 4.7. So, it's an interesting theory about what actually happened there.

All right, guys, before we go to our next story, some breaking news here. Here's your Polymarket, gentlemen. I don't know if you're placing some insider bets here, Friedberg, but it looks like the All-In podcast has a 37% chance of being bought by Anthropic. This is live.

Real-time from Polymarket?

Yeah, this is by the end of the year.

By the end of the year?

Okay, good. By the end of the year. So, this isn't real, is it?

Yeah, absolutely. People have been trading on this. This is heavily traded.

It says $92 million of volume. That can't be.

I'm just calling it out. There's no way this is real, dude.

There's no way this is real. It's breaking news, guys. I don't control the news flow. It's J-Cal slop. Anybody can create a prediction market, I guess. Hey, what's the volume of that thing?

$92 million.

That's not real. There's $92 behind it. There's no way that's real, guys. Come on.

Guys, time to upgrade the planes. I don't need the PC-24. I'll take somebody's G650. Whoever's got a G650, go up to the 800. It's going to trickle down to J-Cal. Let's do it. I'm giving up my United Platinum status. Trickle-down economics.

Trickle-down avionics.

All right, listen. Story number 2.

Speaking of data centers, Allbirds just pivoted from ugly sneakers to AI, and the stock has ripped. We're talking about podcasts getting bought by frontier models and sneaker companies pivoting to data centers. Allbirds, as you know, has the ugliest sneakers on the planet. This became a massive dilution in our industry: that this company was worth billions of dollars.

They went public in 2021. This might be one of the peak ZIRP moments. They raised $350 million in their IPO. Sacks, were you an investor in this thing?

Allbirds? No, no, no. You're thinking about the scooter company.

Bird. That's right. Okay. Don't remind me of all the investments that didn't work.

Bird was crazy. This reminds me of the late ’90s, when all you had to do was change your name to whatever.com and you'd get a huge pop in your valuation. You could spin it out. BarnesandNoble.com became a separate company, and then Barnes & Noble's stock crashed and never looked back.

Collective delusion. People in Silicon Valley liked them and thought it was the next Nike. Tulips. It was tulips.

Well, I think that was an era in Silicon Valley when people rewarded rapid growth without really looking at gross margins or cost of goods sold. People didn't really make the distinction between software and everything else, right?

Right. With software, you never really had to worry about COGS or gross margin, because the incremental cost of serving a customer with software is almost zero. So, people in Silicon Valley weren't really trained to look at gross margin, and there was this rash of physical-world companies that suddenly started getting crazy valuations.

The success of Uber, quite frankly, might have ushered in this era of physical-world companies that started getting valued like software companies, even though obviously they didn't deserve it. David, thank you so much for that shout. I really appreciate it. You and Airbnb—

No, no, no, no. Hold on. The era of 2021 is super interesting, though, guys. I don't believe it was a physical-versus-digital thing. I believe it was a moment in time in the COVID ZIRP era, with massive amounts of money going in, before we had seen inflation. It was all of that happening, and the investor class was basically deciding, “We're going to look 2 and 3 years forward on your current growth.”

It wasn't 1 year forward; it was 2 or 3 years forward. That's where these crazy valuations got weird. So, if you went from $100 million to $300 million in the last year, they'd play that 2 or 3 years forward and go, “Oh, yeah, you're totally 20 times bigger. We'll pay you for that now.” That's where it got weird. It got weird.

Hey, Bird Scooters was valued at $2 billion or $3 billion. They were doing micromobility—

You're just trolling them, Jay.

I'm not. We're giving everybody their flowers for incredible investments and saying even the mighty Sacks could trip up. Did you sell the shares at the IPO?

No, no, we don't do that. It was a Series A we did. I mean, look, we got it right in the sense that it was a total phenomenon.

But then the cities just cracked down on it and killed it.

If the cities had leaned into it—if they had reacted differently, if they had created, let's say, a scooter or small-EV lane—it could have transformed cities. It would have been a lot easier to get around. But instead, they banned it, they limited it, and they didn't create designated areas for it.

The coup de grâce was when they would take a city where Bird already had, say, 80% dominant market share, and then say, “We're going to choose 4 operators and give them each 25% of, say, a 1,000-scooter allocation.” That just killed the economics for everybody.

Yeah, there's no market anymore.

There was no marketplace network effect, right? When they're basically just picking the winners and deciding the market—

You can't compete and split it evenly, so—

Yeah, there's no competition. You can't drive value.

Regulatory capture at its worst.

Guys, we should watch for this in the autonomous-car space, too. Cities may get cute and start doing things like that. What they did with scooters, they could do with cars.

That's a really good point. Instead of just letting the market play out, they say, “We're only going to have X number of thousands of autonomous cars.” New York is literally doing this right now, like the medallion system.

And Boston.

The prices go up, the innovation doesn't get realized, consumers don't benefit, and no one benefits. Basically, you've deleted the market.

Anyway, just to wrap up this story on the shoe company, they're now Newbird AI. They bought 8 H100s, I think, with a $50 million convertible note, and the stock has gone up.

Not 8. Stop. 8, really?

No, it's a joke. The stock's now at $14 a share. It's up 450% in the last week. Shout-out to WallStreetBets for staying solvent—

More than the shorts can stay solvent.

Can I say something serious?

Yes, please.

There are a handful of transactions that have happened in the last few days that, if you look from far away, are head-scratchers.

So, this is one. I don't know if you guys saw recently, but Jane Street did a $1 billion investment in, essentially, a new Neoscaler. And then it also did a $6 billion compute deal with them. That was a little interesting.

What would I like to say about this? I think the thing that the capital markets are getting right is that we are massively compute-constrained. Massively. And there are 2 problems. One is power.

If you look at companies like Bloom Energy, it has gone absolutely straight up—vertical, nuclear. The reason is that Bloom has a solution that allows you to use natural gas, do something on-site, and, critically, get your clean-air permits very quickly because it has very, very few emissions, and that has been proven.

Instead of waiting for years to get on the grid, if you wanted to build a data center, you can now use their services. The other part that's going absolutely nuclear is the actual land and the shell, because it's turning out to be impossible to get these approvals.

Now, why is that? I sent Nick an image. The reason is that underneath, at the core of it all, there's a tide shifting on AI. The American population is incrementally getting more and more negative on the whole subject matter writ large, and it's not clear exactly why they're doing that.

Maybe it's the doomerism that we talked about last week. Maybe it's the fear of job loss that Jason has been talking about. Maybe it's just this idea of yet another wave of innovation that's only going to benefit a few in an extreme way, minting trillionaires all over the place while everybody else stands still.

I don't exactly know what's causing it, but the sentiment is shifting. And as the sentiment shifts, the most scaled action that they can take is to vote down data centers. Here's an example that was insane: a town approves a $6 billion data center build, and half the board gets ousted—voted out overnight—so that they can put in new people to undo the decision.

If you look at this all around the country, the answer is not, “Oh, we're only going to build in Texas.” That doesn't work. There's not enough power, not enough grid capacity, and not enough natural gas to allow that to happen. Maine just passed a bill that bans all data center buildings.

So, I think the reason why All Birds went crazy is a very, very small canary in a very important coal mine: we are absolutely compute-constrained. I think if you play this out, the real problem again goes back to Anthropic and OpenAI. If I were them, it would be a five-alarm fire.

They, more than anybody else, need to get their hands on compute. They need to have land, power, and a shell, because otherwise that revenue could either slow down or hit a wall. And it will not be because of product quality and adoption. It will entirely be because of the Friendster effect: you just couldn't keep the site up.

And I think that that would be a huge, huge problem for everybody. Sacks, final thoughts on data centers and the build-out.

Well, I think Chamath is right that the data center has become very unpopular. Probably in 30 states, there's going to be a ban on them outright. And then it's very hard to get projects approved.

Look, I think there are a few reasons for this. One is that there are a lot of, let's call them, real-estate developers who are kind of wildcatters out there trying to get entitlements. They brought a lot of projects up for local permits where they didn't have a power solution.

There's no question that local communities do not want data centers drawing off the grid, thereby increasing residential prices if that data center is not bringing its own power generation. And the administration agrees with this. This is why the president did the Ratepayer Protection Pledge, where we got all the major users of the data centers—all the hyperscalers—to agree that they would not build new data centers without bringing their own power.

Again, it was designed to be power-neutral to the grid. In fact, it would increase the amount of energy available to the grid because these data centers would give back power when they're not at peak usage. So that's category number 1: there's this fear of electrical rates going up.

But there are a couple of other categories of groups. The second one was Future of Life Institute, and a lot of these doomer groups saw that data centers were a way to stop AI progress. There are interviews with some of these doomer folks who say things like, “We have to meet people where they are,” meaning that they've been unable to convince people that AI is going to lead to the Terminator, but they can convince them that AI data centers are going to use up their water, for example, which isn't true.

A lot of the NIMBYism has been astroturfed by a lot of the doomer groups, which have a lot of money thanks to contributions from a few tech billionaires, as we've talked about in the past. So that's category number 2.

Category number 3, ironically, is Anthropic itself. It has allied itself politically with a lot of the doomer groups and a lot of the NIMBY groups. It didn't seem to matter in the first couple of years because Anthropic had made the strategic decision not to build its own data centers.

They probably thought that they were just throwing sand in the gears of OpenAI or xAI, their competitors, and that they would just rely on hyperscalers to get their compute. I think that strategy has now backfired, in the sense that they apparently have reached the limits of the compute available to them by buying it from a third party, and they need to build their own data centers.

It's going to be very interesting to see how they adapt to that and how the message around data centers changes over the next year, as the effective altruists decide that all of a sudden data centers—or certain kinds of data centers—might be a good thing because they serve their mission. I think that's going to be a very interesting thing to watch.

I'll tell you the one thing I think you're missing, which is that most people in America really are starting to really hate rich people. There is no physical space that better represents the wealth in America—the wealth creation that's happened that a lot of people feel left behind by—than the data center.

What other physical space is there to go to? It is the temple of the wealthy. It is the mechanism, the tool, the machinery of the wealthy. It is the way that the rich, elite, tech, politically connected billionaires that we're obviously all attached to are taking from the poor, getting themselves ahead, shooting themselves into space, and leaving everyone else behind.

The data center, I think, is the representation of their progress. And it is a representation of the progress that others don't feel. So that's why I think it is physically the manifestation that people want to attack and destroy.

There's very little rationale about, “Oh, let's stop AI robots from killing us.” I think people just don't see the value in AI. The average person doesn't see the value in AI today. As we just talked about, so much of the value of AI is showing up in the enterprise and in the rebuilding of enterprises.

But for a consumer's life to actually be altered in a meaningfully positive way, most people don't feel that yet. The best thing they see is some medical advice they're getting on ChatGPT or something, and that's kind of the end of it for them.

So, I think there's a lot of this populism that's swollen and taken over not just the U.S., but probably a good chunk of the West. And the data center is the target. It is the pied de terre of this space, in a way. This is their attack vector.

To David's point, they'll probably ban data centers in 30 states. Data centers can output and input at the speed of light, so they can be anywhere. The reason you put data centers in different states is cheap power and low latency. Maybe you get a couple of milliseconds of latency.

The data centers can go anywhere. So, as soon as all the states start banning data centers, the data centers will just go to space, to Iceland, to Texas, or wherever. And then everyone will be forced to move on, right? They'll basically find the next target for populism, which will be something else.

The piñata, the data center—what's next? This is part of what's going on right now. Private-jet fuel.

I always bring it back to this because I do think this is at the root cause of populism: the fact that the government promised so much, it's so inefficient, and, as a result, it's destroyed the value of the dollar and not given anyone anything that they thought they were getting.

And we do have the Ratepayer Protection Pledge, Sacks, to address the energy issue. That was a big win, I think, and you worked on that, I believe.

Yeah. But hold on, can I just say something? I think that Ratepayer Protection Pledge is important. It doesn't change the business model of the utility.

Meaning, what the utility is allowed to do is, every year, build a budget. And the way that they decide how to charge you is that they are allowed to make investments, right? Then they're allowed to earn roughly 10% on the amount of investment that they make.

What do you think their incentive is? Their incentive is to find ways to keep investing and upgrading the infrastructure. So, I suspect what you'll see is, independent of what the data centers do on-site—which I think is smart and good, and I'm glad you guys did that—it doesn't slow down the actual fundamental business model of the utility because those are local monopoly licenses that are granted at the state and county level.

When you go inside and look at those utilities, what they're allowed to do is say, “Well, I'm going to bury the lines underground because there's a wildfire threat.”

That's going to cost $10 billion. They present that, the PUC has to say yes, and they're allowed to make 10% on $10 billion. So the business model of the utility has to be looked at because it's independent of all these other things. They have an incentive, like insurance companies, to just walk prices up over time.

It's true. That's why I think behind-the-meter is so important there, for sure. But look, this idea that data centers don't create jobs is just wrong. It's been a huge boon for blue-collar jobs and the construction industry while they're building the data center. That construction could be a wave that goes on for a decade or 2 decades.

The CapEx is not going down; it's increasing. Listen, for example, to what Jensen says about his projections. It's not like the CapEx is a 1-year thing. We're seeing tens of thousands of new construction jobs being created, with 25% to 30% higher wages for electricians, carpenters, the guys who hang drywall, pour concrete, create roads, and install equipment. These are blue-collar jobs, and it's creating not just new jobs but also wage increases. It's a good thing. It's not a bad thing.

But it's not the same as a fab, where the jobs are permanent and they're there. We can debate it all we want, but there are about 100 data centers right now that are being contested. The data is that for every 100 that are contested, about 40 get canceled. That number is increasing this year; it's already more than doubled from what it was last year. The total economic value of those 100 data centers right now is about $162 billion.

Jason, the thing is, there are permanent jobs. Where's the energy that's powering that? Where are the semiconductors that are powering that? We actually had Chase from Crusoe and Michael from CoreWeave on the All-In interview show, and they said they're bringing energy with them.

That's their big thing: BYOE—bring your own energy—to the space. They're bringing in natural gas, diesel fuel, and solar. BYOE is the model.

Well, can I also say, if you were of the opinion that you didn't want more data centers in America because they were using up scarce energy resources or something like that, but you weren't just anti-progress altogether, then what you would want to do is at least see more data center construction among U.S. allies—energy-rich U.S. allies, right?

A year ago, one of the things I worked on was allowing the Gulf states to build data centers with American technology for American companies. Remember the controversy that caused? Everyone accused us of somehow serving China. Well, those data centers are getting bombed, and I don't think they'd be getting bombed right now by Iran if they were serving China. These were—

They're very strategic, yeah. The IRGC has put these data centers on a list of assets that they've threatened to destroy because they're American assets—

Yeah, they're assets of our Gulf-state partners in partnership with America and American companies. This whole idea that somehow this was a threat to American national security was a total hoax, just like the idea that data centers use too much water.

What I'm saying is that a lot of the same forces were behind this hoax as are behind these data center hoaxes. Friedberg is right about the resentments, but those resentments get whipped up and marshaled by people who have an agenda, and that agenda is well-funded and strategic.

I can tell you that in the case of the GCC data centers, Anthropic was adamantly opposed to that. They were lobbying against it, and they were again salting the earth against those projects. I think the whole issue is kind of moot now anyway, because I think the data centers have been blown up, or they've been threatened to be. But it just shows how ridiculous—and what a psyop—some of the opposition to these ideas is.

And it's a cell phone in many ways. If you think about who the top 2 spokespersons for our industry are and what they're communicating to Americans, versus how the Chinese view AI, which is incredibly positive, we've got Dario, who says it's the end of days: everything's going to be hacked, all your files are going to be hacked, and all your accounts are going to be hacked. On the other hand, everybody's losing their job. That's our top spokesperson, Dario from Anthropic.

The other top spokesperson just had a 70,000-word piece written about him, where Ronan Farrow said, “I had a dozen people tell me, unprompted, Sam Altman's a sociopath.” Those are the top 2 spokespeople. Those people cannot be the spokespeople for this industry, and this industry has got to get focused on fixing the big 3.

Healthcare is going to be dramatically improved by AI. Housing could be dramatically improved. I'm not sure who's working on that. And obviously, education and the cost of education. We need somebody out there—a Michael Dell, Jensen, or Elon—just explaining how we're all positive this could be, because right now AI is less popular than ICE, the Democrats, and the government of Iran.

Joe Lonsdale at Alpha School.

On the education side, I think you just have to look at Alpha School. It's working.

So, that's your story, Chamath. That's the story of All Birds, the most overvalued startup. We're going to play a little game show for everybody. This is The Price Is Wrong—David Sacks is wrong. The price of the startup is wrong. This is the game show where we guess which overvalued disaster we're talking about.

First up, hailing from Austin, Texas, an enforcer of the PayPal Mafia, it's Mr. David Sacks. Welcome to the program, David Sacks. Are you ready to play The Price Is Wrong?

Let's do it. Let's play along.

All right, here we go. The Price Is Wrong. This startup was once valued at over $13 billion. Its main business model was selling JPEGs for fake internet money. People lost their minds and started spending millions of dollars on monkey images. Play the thinking music, please.

David Sacks, can you name that startup, once valued at $13 billion?

Is that the Bored Ape thing?

Okay, you're closing in. Remember, phrase the answer in the form of a question.

I know, I know. They're selling JPEGs for fake internet money.

Me, me, me. Pick me, pick me, pick me.

You'll get your chance, Chamath. I think I know what you're talking about. All right, let's move on to the next one. I can't remember—they're spending millions of dollars.

Can I say it? Can I say it? Can I say it?

Time, Chamath. It's your turn to steal.

OpenSea. OpenSea.

Correct. 100 points for Chamath Palihapitiya. Okay, an amazing steal. And next up that puts

Here we go. He puts the dick in dictator, hailing from Palo Alto, California. It's your favorite DoorDasher, Chamath Palihapitiya. How long have you been DoorDashing there, Chamath?

9 years.

9 years. You like that? What do you like about that? Is it the interaction with the people? Is it the tips? Is it stealing a couple of french fries? What do you like about being a DoorDasher?

I typically take 4 fries. I lick the burrito.

Okay, licking burritos. Did he just say “licking burritos”?

That's why we need robot delivery drivers. I mean, it's just—

It's coming.

All right, let's get the music going here. This startup was once valued at $4 billion. It allowed you to talk to other people on your mobile device during COVID, and you could listen to mid-level VCs. Can you name that overvalued startup?

Yes, yes, I got it. I got it.

It's called Generic Startup Device.

God, it's called Clubhouse.

Oh my God, he's on a heater. That's 200 points for Chamath Palihapitiya. That's a lot of DoorDashes. He's going to do very well here in the final round.

Memory. How soon we forget. I couldn't place any of these names.

Okay, finally, streaming in from a potato field at an undisclosed location in Idaho, it's Dr. David Friedberg. You're in the potato sciences, correct, Mr. Friedberg?

That's right. Potato science.

You like the potatoes? Do you have a favorite potato dish? Is it the scallops? What do you like to do with your potatoes?

What is your name, Mr.—what is the host's name? What is your name, host?

It is me, Jocular Calacanis. Let's go. Here we go. Jocular Calacanis here at the helm. You like a certain type of potato there? What do you like?

Jocular? Jocular? It's Jocular Calacanis.

And what do you like on the potatoes? What's your favorite? You like a creamy mash? A little cheesy? You're liking the cheese. Okay, very good.

Okay, David, this startup was once valued at $270 million. This is your favorite. I understand you're a vegan. They sold juice. These juice packages went to a juice machine. Can you name it?

I know it. I know it. I know it. The next iPhone, the iPhone of juice. I know it. Pick me. Juicero. Juicero.

Yes, that's right. That's 100 points on the board for you. Wait, are you searching? It looks like he's searching. Did he search? Oh my God, are you on your device? He looked like he was on his device. Internet-free here. He's on his device.

All right, folks, there you have it: a clear winner, Chamath Palihapitiya. Tell him what he's won. He's won a trip to Temptation 2 in Cabo. You're going directly to Temptation 2, an adult resort. That's for you and 2 of your friends. Enjoy Temptation 2, Chamath Palihapitiya.

Two of your friends. What a fun game this is. You can bring your throuple. I think this is going to need to become a regular feature of the pod.

Okay, landing. We find out everyone forgets the unicorns that don't work. They just disappear into the ether. All right, there you go. You won a pied-à-terre—you can finish with a tax bill in New York. You got a pied-à-terre available to you as the first-place winner, Chamath Palihapitiya.

All right, listen. There's been a lot of shenanigans going down in D.C., the most boring city in the world, where apparently everybody's getting a little frisky after hours. TMZ launched a news bureau. Eric Swalwell is out of the governor's race. There's a lot of dark stuff that's been released. He is innocent until proven guilty, but it's not looking good. He also resigned from Congress as well.

Friedberg, our investigative journalist, now working as a stringer for TMZ. Friedberg, what have you learned? What's in Friedberg's day?

Have you learned anything?

Okay, you have an analysis, then, maybe.

My anecdote on this is that back in December, when it was first rumored that Swalwell was going to run for governor, I started making some calls to various folks to be like, “Hey, what do we think of this guy? Is he going to be a good candidate?” Obviously, I and a lot of other people, before they evacuate the state, care a lot about the future of California, so I started checking around.

I spoke to several people who independently told me that there was knowledge about this guy sending pics to employees, and that this guy had a bunch of stuff that was going to come out about him. I heard all of this not from 1 person, but from several different sources. This was back in December, going into January, and I largely dismissed it because I was like, “If this is true, this would have all come out already.”

I thought, “There's no way this is true. If it was true, people would have talked about it. They would have made a thing about it.” If multiple people were telling me about this, then I had to assume that it was a rumor being used to block him from running for governor, rather than it being a real thing, because multiple people had this knowledge and this information.

This was December and January, when I had these conversations, and at that point nothing had come out. I was like, “Okay, it doesn't seem like this is real.” Then everything that I had been told started to come out in the last week.

So, the striking aspect of all of this for me was how much knowledge there was about these various incidents with the guy, how so many people had this knowledge, and how no one had actually brought the knowledge to bear. That begs the question: Why did they not do what was right by the victims? Or why did the victims sit on the sidelines, waiting for the right moment to all come out together?

This was broad knowledge within a community of people, and they made the choice not to bring it forward with that knowledge. That's what was so striking to me about this whole thing. I had honestly dismissed the whole thing as just being rumor-mongering to try and besmirch the guy.

It turned out that these were all being held back purposefully and deliberately for a very particular moment in time, when they were all brought forward to be used. Let me just be clear: These are all allegations. Nothing's been proven in court.

He does get his day in court. Absolutely. Again, we just want to make sure there's a bunch of allegations here.

All that I'm saying is that people had told me about these supposed claims 5 months ago, 4 months ago—multiple people—and had chosen not to bring them forward. The victims had not come forward publicly with these claims. Then there was this coordinated effort to bring everything forward at the same moment.

That's what was so striking to me: just how coordinated all of this was.

Who's controlling it? Who do you think? Is there a meeting? Is there a council?

Nancy Pelosi.

I don't know if that's true.

Well, I don't think it's just her. I'll tell you my sense of it. My sense of it is that there are certain insiders, and he's not an insider. Those insiders are people like Katie Porter. Katie Porter, I think, is who the Democratic establishment wants to be governor. She's an insider to national Democrats. She's an insider to California Democrats. I think she's the preferred candidate.

Tom Steyer. She's a spousal abuser.

I don't know that Katie Porter is this ultimate insider, but I think there are—look, there are clearly insiders. The Democratic Party is a machine that exists to siphon off as much money as possible from the public till to the interests that support the party. It's their gravy train, and they're not going to let anyone stop that gravy train.

The Democratic Party had a huge problem in this California governor's race, which is that the Democratic field was very fragmented. The 2 Republican candidates were actually polling the highest.

Just so the viewers have context, California has this weird jungle-primary system where the top 2 go to a runoff. They don't have a Democratic lane and a Republican lane. They just take the top 2 in the jungle primary, and then they go to the runoff.

Even today, Steve Hilton and Chad Bianco are both polling at around 14% or 15%. If the election were held today, you'd have 2 Republicans in the runoff. So the Democrats needed to winnow the field down and have fewer candidates.

In addition to that, they must have been concerned that all of this opposition research would come out once it was him versus, say, Steve Hilton. They didn't want it to come out later, when they could lose the election. So the powers that be made the decision to lance the boil.

Probably there was a conversation with him to tell him to get out of the race. He didn't listen. By the way, this feels a lot like what happened with Joe Biden when he had to drop out of the presidential race.

Surprisingly like that.

The Democratic establishment and all of the mainstream media were saying that Biden was sharp as a tack, okay? Then he had that disaster debate performance with Trump, and it became clear that—

Yeah, you could just see the text messages flying during the debate between the Democratic Party insiders. By the time that debate was over, they had congealed on a new position, which was that Biden had to step aside.

Then Nancy Pelosi was reported as having gone to the president and said, “We can do things the hard way or the easy way.” Imagine that—telling the president of the United States, “We can do things the hard way or the easy way.”

Biden disappeared for a week, and magically he stepped aside by tweet. Remember that? He published a statement that appeared to be done by an autopen. People were speculating whether he was even behind this or whether the staff pushed him to do it.

The whole thing was extremely weird, but he was clearly muscled out of it. Just a few days before, he had said, “I'm not leaving the race no matter what.” It was like straight out of that Wolf of Wall Street meme: “They're not getting me out of here.” Then, a few days later, he's resigning by tweet.

Again, Nancy Pelosi appears to be the figure at the center of both these things. She was—hold on. Hold on. Pelosi—

Hold on a second.

Pelosi is reported as having been Swalwell's mentor. I guess she found him roughly 20 years ago to run for Congress in the first place. When the Republicans wanted to kick Swalwell off of the House Intelligence Committee for allegedly being involved with that Chinese spy, Fang Fang, it was Pelosi who protected him.

She's been a central figure in his career. I'm not saying she approved of anything he did, but—

Listen, politics is dirty.

But there's no way that button gets pushed without going to Pelosi for the sign-off, right? I mean, she's the boss of this operation. I think the same thing that happened to Swalwell happened to Biden: They went to him and said, “We can do things the hard way or the easy way.” He was too dumb to listen, and they did things the hard way.

Yeah, shout-out to Nancy Pelosi, incredible day trader, and shout-out to friend of the pod Ro Khanna, who's now beaten Nancy Pelosi. I think his trades this year—he's actually even beaten Nancy Pelosi.

Ro Khanna has traded $600 million of stock. He trades more frequently than Citadel Securities.

What?

I mean, it's incredible. Where's his PR director? I want to get in on it. The congressman you've been supporting for years is a great stock trader. Apparently, we should have been talking to him not about taxes, but about trades. You supported him?

There were things I liked about Ro Khanna. He did support freedom of speech with the whole Twitter thing. You know what it was? The Overton window shifted so much.

When I supported Ro Khanna, freedom of speech was a big issue, and he was 1 of the only Democrats to support that. He was also, I think, the only member of the Progressive Caucus to support a diplomatic track for Ukraine, which I supported, and I gave him credit for that.

I knew that he was in favor of a wealth tax, but I thought that was just a very unserious proposal that wasn't in play in any way. The Overton window shifted so much that now the wealth tax really is a possibility. So, you know, things have changed.

Here it is. Shout-out to our boy Ro Khanna. Look at this. Right about now, Chamath's thinking about making a managing director offer here. I mean, Ro, if you get booted out of office, you could become a managing director at Alt Capital.

Statistically, even if you had insider information, making that much money is very hard.

Holy cow, Nancy Pelosi is going to shiv Ro Khanna next for not giving her the inside track on whatever he's betting on. What did he bet?

Well, look, any investor, as we all know, can have a good quarter or a year, but to put up the kind of returns that Nancy Pelosi has done over decades is nothing short of miraculous. It's generational. It's a generational run. Give her her flowers. She's substantially better than Warren Buffett. I mean, Warren Buffett, Stanley Druckenmiller, and Nancy Pelosi are 3 of the most accomplished investors of all time.

Whatever you did to the poor Buffett fans, Chamath, they are incredibly angry at you for desecrating his legacy.

I didn't desecrate his legacy. I pointed out 1 unavoidable fact, which is his returns were bimodally distributed pre- and post-Reg FD. When you have to follow the rules of disclosure, everybody's returns got knee-capped. When there were no disclosure rules, his returns were off the charts. That's just a mathematical truism.

Now, what's interesting to note is the reason why Nancy Pelosi's returns are so consistently good: Reg FD does not apply to people in Congress. That should be the takeaway.

To stop them? They can learn things in their committee meetings. In fact, there are situations where things are disclosed and then they are trading in real time. Shout-out to the SCIF. Yeah, get out of the SCIF quick and get that trade in.

So, I have enormous respect for Warren Buffett and what he's done. He's the GOAT of GOATs. But the returns post-Reg FD are just materially worse than they were pre-Reg FD, and that's just—

What's he going to do with that cash position, Chamath? Speaking of Buffett, they have—what? $300 billion sitting there?

The market is in a very complicated moment right now. If you look at historical indicators of value—if you look at Shiller as an indication of value—it's peaking. If you look at the Buffett indicator, it's peaking. So, there are things that, when you look at them, look like all-time highs.

The problem with that is you would say, "Oh, man, but there's this weird dispersion happening in the market." Dispersion means literally a few companies are hitting all-time highs—I think it's 8 or 9—and everybody else is not. So, it's a really complicated moment. It's hard to understand what's going on, but he's got a lot of cash.

If he's sticking to his knitting, he's looking at the Shiller index and he's looking at his own indicator, which shows all-time highs, and he's waiting for a correction. He's not really in charge anymore. Berkshire Hathaway is not Warren Buffett anymore. Even though he's a big owner, he's not really doing it.

But the fact that they're sitting on that massive pile of cash says something. The fact that they're not putting it to work in the market says they don't see an opportunity yet.

Sacks, Friedberg, I'm curious about your takes. Either one of you can go in whichever order you want. How is the market crushing it while we're in Week 7 of a war and we've put $100 billion or something into this military activity in Iran, and the market is pricing it in, shrugging it off, and we're hitting all-time highs?

Yeah. Look, I think it's pretty straightforward, which is that in the wake of the meeting in Islamabad, the market is feeling confident and pricing in that the war is going to get resolved. The president also recently said that it's very close to being wrapped up. The military objectives are close to being achieved, and he's made it sound like it's going to be resolved.

Yes, a deal was not signed in Islamabad. I always thought that was an unrealistic expectation—that these 2 countries, which are at war with each other and, in fact, have had hostile relations for almost 50 years, were going to resolve all their differences in 24 hours. It's not realistic.

But the impression that the market has, and clearly is trading on, is that the war is going to be what Donald Trump said: an excursion and something that is on its way to being resolved, and that they've made progress. And JCal, you're right. All week has been trading incredibly strongly. I think by Tuesday the market had recovered all of its losses since the start of the war. I think it made a new high yesterday, on Wednesday, and is making fresh highs today, on Thursday.

So, you just have to say that at the present time, the market thinks—and I would consider the stock market to be the ultimate prediction market—that this war is on its way to being resolved.

Just to qualify, I'm not speaking as a member of the administration. I'm not saying that I know something, okay?

You don't know anything. I got it. There's no clipping.

I'm not representing anyone, and I don't know anything different from what any of you know. I'm just saying that I think this is what the market is clearly pricing in, and I'm paying attention to the statements of the president and vice president.

And this, Friedberg, you and I were talking the other day while you were slaughtering me in backgammon when you won our 8-point match. You're almost caught up. We were talking about the TACO trade—"Trump Always Chickens Out." You have a theory about Trump always caring about optics. Unpack it for us, Friedberg.

Yeah, so while I was beating you in backgammon—

Ooh.

—I sort of had a thought. Because JCal asked me, the insight is—I think this maybe is just the more simple view of things—that Trump's weather vane is the stock market.

We see volatility up, we see the VIX is really high, but the S&P is trading in a band that's actually fairly tight, given the craziness that's going on. Everybody's nervous, but he moves in the policy space. He does not let the S&P go down too low.

People also get the sort of panicking thing now, which is that he gets people nervous, he makes moves, and then he comes back to reality and gets things done. It's practical and probably the better parts of what he does well, and people are pricing that in.

Yeah, it's like playing poker with Chamath and Alan Keating. It's just like, how is this guy solvent? He's literally playing every hand of poker, he's losing tons of money, and then all of a sudden, by the end of the night, he hits 2 big pots, and he's got the nuts both times, and he pulls out of the stall.

What's your take on the market today, Chamath?

Yeah, I mean, I'll just do it again, but I think the Shiller P/E—JCal, I sent it to you—shows near all-time highs. Then the second is the Buffett indicator, which is the sum of all U.S. equities divided by GDP. It's also at all-time highs.

So, this would generally mean that you need to be increasingly a little bit more risk-off. But then the opposite side of that—and I sent you a third one, and this is why it's so confounding—is you have signals showing everything, which typically doesn't happen.

This is the dispersion point: When you see this performance and it's up 5% in the first half of April, typically the market is up almost 32% on average for the rest of the year. And we were already up, as Sacks said, 7.5% already in the first 10—

Read this graph. This is crazy. I don't even know what's going on here.

Hey, this is just a dispersion. But the idea of all of this is, I think we're in a moment where you can find a piece of data to underwrite your bias. I think that's where there is a lot of danger.

For me personally, I'm generally more risk-off right now. More importantly, I'm waiting for these IPOs so that I can, to be very honest with you, delever and get some chips off the table.

I think that it is crucial that the SpaceX IPO get done ASAP. And then I think it's even more crucial that one of Anthropic and OpenAI front-run the other one and get out first. The first 2 charts are the reasons why.

It's fairly obvious to me what's going on here. You have multiple trends going on at the same time. The reason traders are valuing this—and I think there is some smart money in the market right now—is what you're seeing in terms of the earnings potential of these companies as they deploy AI, as they have unlimited intelligence, and as the top employees at the top companies become 10, 20, or 30 times more productive.

That's never been seen before. Microsoft Office may have made you 30% more efficient. The internet may have made you 50% more efficient. But none of these things made you 10 times more efficient.

And you had a really interesting point that you slipped in earlier, Chamath, which was that nobody knows how to harness these things yet, right? And it's producing slop. The truth is, 10% of people do know how to harness it—20%, maybe. I watch this in my own organizations, and I watch it in 600 portfolio companies. The ones that do deploy it correctly are running the table on the ones that are not.

So, the efficiency boom that we're going to see at these companies, whether it's Meta, Uber, or Airbnb—

Maybe I'm just an idiot. Whoever executes it properly is going to be phenomenal. The earnings will be insane.

Maybe I'm an idiot, but it has not translated into a tsunami of more revenue and more profit for me yet. Maybe I'm the only one, and maybe it's everybody else but me, but I haven't seen it. I see it in a lot of companies.

So, I think in private companies, we have 2 companies as examples. Micro1, which is doing data—dark pools of data—for these language models. They have built technology that allows them to build and collect data to help the large language model companies grow, and they are on a tear by using this to identify and find great people who can then contribute to these corpuses.

Then we have TaxGPT, which is making accountants—I think they have 6 or 7% of all accountants using their platform.

Do you agree with me? They're just ripping. I hope you can agree with me. I hope you can agree with me with the following statement: Small companies enabling at the edges is not where these guys will build a multi-trillion-dollar market cap. I think we're both in alignment.

Big companies, dumb companies, slow to implement the technology. Startups? No, they are dumb when it comes to implementing new technology. They're always the laggards when it comes to implementing new technology.

The reason that they are slow is not because they are dumb. The reason that they're slow is their business is much more sophisticated and much more complicated than many other businesses. What I'm saying is, if you can't prove that this works in the big-time, prime-time, big-league use cases, it's a toy.

All right, Sacks, where do you stand? Are you in the JCal position—the startups are showing the way, they're being massively efficient, they're growing revenues like we've never seen in the startup community before with fewer people—or are you in the Chamath camp: Big companies are not having it drop to the bottom line, and they're smart? Or are both things true? Where do you stand, Sacks? Be the adjudicator of this case.

I mean, honestly, I'm probably closer to you, J. Cal. I hate to say that. Listen, I think people are still figuring out how to drive business value out of AI. Change management is hard, and the bigger the company, the harder it is.

What's happening right now is there are a lot of transformation projects at large enterprises that are failing. There's a big McKinsey study on that. But if you look at activity from the bottom up, I think it's very interesting. It's becoming more interesting.

Over the last several months, obviously, with coding reaching a new level, we're starting to see very interesting things happening there. Obviously, the revenue that's now being generated from these coding models is exponential, like we've never seen before.

So, the ROI is finally there at the model layer. Before, people were saying that it's a bubble because you had all this massive CapEx at the data-center level, and there was no ROI coming at the model layer. Now we have the ROI at the model layer, and I guess we're questioning whether the ROI will be there at the application level.

In any event, I think things are progressing. Look, fundamentally, I'm bullish on this whole thing.

No, no, no, let me be clear. Hold on, hold on. Obviously, I'm bullish. I'm in this. I'm in the space. I'm doing it. You're in the arena.

We have seen in every single wave—in the mobile wave—we needed consumers to show up at scale. We needed consumer experiences, and it was very obvious that there were these consumer businesses that were going to be, if not already, incredibly, incredibly profitable. Google and Facebook were profitable within the first few years. They never looked back.

All I'm trying to point out to you guys is there is not 1 great example yet. If we believe enterprises are where all the money is, and if we believe that's what's going to underpin these trillion-dollar valuations, please, somebody show me a couple of good examples of scaled profits.

Okay, fair enough.

Travis, final word here on— But by the way, can I say one other thing? I think Chamath has a good point about being risk-off right now because if you just look at valuation metrics, they do seem to be quite high. I've seen other versions of those metrics as well.

Look, it's very hard to time the market. I don't try. But I think you could make an argument, based purely on valuation levels, not events, that you want to adopt a more conservative posture right now.

I think the bet you're making then, Travis, is: Are these valuations so high right now, by the standards that the efficiency AI could bring to these companies is not real enough to continue the growth from here? It's not a catalytic enough technology. And, you know, when you look at—

The opposite. All these traditional companies have horrible valuations. They've been crushed. So, my point is, if AI is real, the upside is also real.

All I'm saying is the details matter, guys. It's very hard to take a very complicated business and all of a sudden, quote-unquote, transform it. It's not as easy as it sounds. That word is easy to say.

For sure. Travis, go ahead. Final word. Final word from Travis, please.

First, when it comes to big companies, I think the big thing about, let's call it, the autonomous enterprise, is change management. That's the big boy.

Change management is actually about all the people that already work there: the middle managers, the technocrats, the bureaucrats, the whatever—the crats.

And getting the change management going there is a human thing, and it's very tricky with very complex processes, many of which are not even documented. In theory, it's just all going to happen real fast, but in practice, that's hard. So, that's part 1.

Part 2: What I'm seeing with true tech companies—real companies, public, hardcore public companies—I mean founder-led companies, folks that are really cranking public companies, and tech companies that are sort of up-and-comers—I talk to CEOs across the board, and they're fired up about the development, the productivity, and the deployment schedule. They're able to roll out new features much faster because they've pivoted their culture to be very pro-AI development.

I'm getting, at this point, almost consistent feedback from real founder CEOs that this stuff's real and it's not just hype.

Now, there are folks pushing, selling their book, that are like, "Oh, we're at AGI," and all this. Anybody who's worked with these agents and done the AI dev stuff knows there's a lot of good stuff, but they're not that smart yet. They're just not that smart.

I've got a side quest where I'm just investing—I have agents investing and betting on Kalshi and Polymarket and these other places—and it's silly how dumb the agents are, even their best agents, to be honest.

With agents, you have to be human in the loop. The agent has no taste. The agent can do repetitive tasks. The agent is not going to do something novel, and it can quickly get lost in the forest.

We had to spend a lot of time with our investing agents getting them on board with the idea that if you want to make money investing, you can't be on both sides of the same bet.

You know what I mean? There it is. That's where we're at. AGI is not here, and it's kind of silly for folks to suggest that it is.

Absolutely. Hey, guys, we've got to wrap. Travis, you didn't get to play The Price Is Wrong. I have 2 more. Do you guys want to do a bonus round of The Price Is Wrong?

Yeah, do it. Do it. Do it. See if Travis gets it. Okay, 2 more. But you guys can steal. You guys can steal here. Let me get my background music. Give me the music.

We got Travis Kalanick. He's here. He is a professional water-sports player from Tallahassee, Florida. How is it down there in Tallahassee, Florida? I understand you do a little handiwork to pay the bills, but you spend most of your time out there on the lakes and the oceans doing water sports.

Yeah, you like the water sports.

I love the water sports. I do water-ski lessons on the weekends. Anybody's interested.

All right, I'll be out there in Tallahassee, Florida. Okay, now here we go. It's your job to figure out The Price Is Wrong. The Price Is Wrong. Here we go. It's so funny.

This is your bonus round. All right, this startup raised $900 million at a $9 billion valuation before dissolving in 2018. Its famously deep-voiced founder and CEO is currently serving an 11-year sentence in a federal prison.

Can you name that mispriced startup? Can you name the mispriced startup? It rhymes with “Schmarinos.”

“Schmarinos.” Okay, you're closing in here. Be careful. You have to formulate a question. I'll answer.

Theranos.

Theranos. Okay, very good. 100 points. You're on the board.

And here is your next one. This is the overtime. All 3 players get to play this one. Whoever zooms in first and says it—this is your final. It's the mispriced startup here. The price is wrong on this startup here.

Chamath, you can do this. Come on, Chamath. You got this. Here we go. You can do this, Chamath. This short-form, mobile-first streaming platform raised $1.7 billion from top investors. Come on, Chamath.

Jeffrey Katzenberg. Snap—the... what is it called?

It raised $1.7 billion from top investors across all of Silicon Valley and shut down in just 6 months. Quibi.

Quibi. Wrong. No, David Sacks, wrong. No, no, I know. This was the...

Company. Jeffrey Katzenberg. Quibi! Quibi, quibi, quibi! I was so close.

Yes! There’s your champion, everybody. Let’s tell him what he won. Apparently, he’s won a pied-à-terre in downtown East Austin—a pied-à-terre in East Austin by the airport. God knows what shenanigans are going down there. I think if you need methamphetamine or a date, you’re going to have an easy time there, Chamath. Okay. Oh, hey Chamath, I'm getting hammered with people who want to come to the sold-out Liquidity. 500 was the cap you put on it. You've been around as dictator. Can we add 50 seats? I think we can add 50 or 100. I don't know. Alyssa, can we add 100? Get a simulcast room going. All right. We'll see you all. Oh, and uh the All-In Summit tickets are on sale. Go to allin.com. Uh and yeah, don't get shut out of the summit, folks, cuz this happens every time. You you guys email me 2 months out, the tickets have been sold out for 4 months. So, get your tickets now uh at allin.com. Another amazing episode. And we'll see you next time. Bye-bye. Bye-bye.