OpenAI需要救助吗？Mamdani胜选、社会主义抬头，废除阻挠议事规则的核选项

Jason Calacanis

Brad Gerstner's here, joining us hot after crashing the stock market and popping the AI bubble. Well done, Brad.

All of our portfolios, thank you, were all down 15% this week.

Chamath Palihapitiya

Can we ask OpenAI to just put a moratorium on any more public statements or appearances—

Jason Calacanis

Yeah, right.

Chamath Palihapitiya

—for another couple months?

Jason Calacanis

Good job, Brad. You decided you'd be a podcaster. You're like, “Hey, let me ask a couple—

Mm. Mm.

Jason Calacanis

—of hard questions here,” and you popped the AI bubble.

Yeah, something like that.

Jason Calacanis

Do as I say, not as I do.

David Friedberg

You let your winners ride.

And instead, we open-source it to the fans, and they've just gone crazy with it.

Jason Calacanis

Love you, S.I.G.

David Friedberg

Queen of quinoa.

David Sacks

Are we getting into it? Because I think it is interesting, actually.

No, it's super interesting.

David Sacks

Yeah.

Super interesting. Let's get into it.

1. OpenAI Spending Problem

Jason Calacanis

So Sam, of course, if you're not in the industry, Sam Altman appeared on the fabulous BG2 podcast last Friday, and it got a little frisky when our fifth bestie here asked what I thought was a completely—

Chamath Palihapitiya

Totally reasonable question.

Jason Calacanis

—legitimate—

Totally.

Jason Calacanis

—mundane question.

Chamath Palihapitiya

Mundane. It's actually a softball question, to be honest.

Jason Calacanis

It was an underhanded pitch.

Chamath Palihapitiya

The way that it was asked, I think you did a very reasonable job of asking a good question in a very fair way.

Jason Calacanis

So let's just show this clip here, and then I want to go behind the pod with you, Brad.

So I think the single biggest question I've heard all week, and hanging over the market, is how can a company with $13 billion in revenues make $1.4 trillion of spend commitments? And you've heard the criticism, Sam.

Guest

First of all, we're doing well more revenue than that. Second of all—

Yeah.

Sam Altman

—Brad, if you want to sell your shares, I'll find you a buyer. I just—

Ooh.

Sam Altman

—enough. People are—

Chamath Palihapitiya

Enough?

Guest

I think there are a lot of people who would love to buy OpenAI shares. I don't think you want to sell, Brad.

Including myself. Including myself.

Sam Altman

I think a lot of people who talk with a lot of breathless concern about our compute stuff or whatever would be thrilled to buy shares.

Ah.

Sam Altman

So I think we could sell your shares or anybody else's to some of the people who are making the most noise on Twitter, whatever, about this very quickly. We do plan for revenue to grow steeply. Revenue is growing steeply. We are taking a forward bet that it's going to continue to grow. There are not many times that I want to be a public company, but one of the rare times it's appealing is when those people are writing these ridiculous things—“OpenAI is about to go out of business,” and whatever. I would love to tell them they could just short the stock, and I would love to see them get burned on that.

Jason Calacanis

So, Brad, you asked what Chamath and I were just saying was a pretty mundane question.

Yeah.

Jason Calacanis

You said it very nicely. I guess we could give Sam a little bit of grace. I don't know if he was being a little cheeky, or maybe he's tired of answering the question.

For sure.

Jason Calacanis

But the internet took this and ran with it in a very viral way, that he was angry and hostile. How did you take it?

That's the interesting thing, right? Listen, we bust each other's chops all the time. We get feisty with one another. Sometimes it runs amok. We don't know if somebody's being serious or not serious, and Sam and I had a good laugh after.

I think Sam was feisty, but I think he also intended it as a joke. He knows that I don't want to sell my shares. He knows that I would like to buy more shares in the company, et cetera. But I think the reason that it went so viral is because it is a super-important question. People are really nervous. They're wondering, “Are we walking into an AI bubble?” How can these huge numbers—how can you be talking about $1.4 trillion in spending when you have GAAP revenue that's been reported at $13 billion this year?

So I was a little disappointed, and I tweeted about this afterwards, that the feistiness got in the way of the answer. But if you listen to his words during the rest of the segment, he basically said, “Listen, we think we're going to have $100 billion in revenues over the course of the next couple years.” And JCal sent the team a chart that basically just shows The Information's forecast for what OpenAI and Anthropic's revenues are going to be over the course of the next several years. The Information is reporting that their internal numbers are both over $100 billion.

Jason Calacanis

Is The Information reporting on leaked internal numbers, or is The Information taking a guess?

No, I think this is based on leaked internal numbers, according to The Information. And so I think Sam's in his head saying, “I believe”—and he says multiple times on the pod—“We're going to have revenues in excess of $100 billion.”

And the $1.4 trillion, it's super important to remember, this is over a period of 5 or 6 years. I estimate about half of that spending is going to be borne by the partners, so now we're talking $700 billion in spending. Spread that over 5 or 6 years. In the out-years, you're probably talking about $150 billion of CapEx to OpenAI.

So he's probably sitting there saying—and he said, “We're going to have over $100 billion of revenue”—so if we have $150 billion of revenue and $150 billion in CapEx, now it begins to pencil out a little bit more. But importantly, he said, “And if we don't have those revenues, we've got to match our revenues to our expenses,” right? I think they will just extend, recut the deals in order to make those expenses doable for the company.

Jason Calacanis

This is an important point because we don't know. We haven't seen these actual deals, or whether they have conditions or outs, or whether they can push them out or cancel them. That will come out, I guess, in the public filings. But putting all that aside, the market was not happy about this.

Mm.

Jason Calacanis

Microsoft, NVIDIA, Oracle, Broadcom, and CoreWeave—all the partners we're talking about—were all down 6% to 20%. When you see these charts of all the deals Sam has done, and Sam's a great dealmaker, obviously, this has in fact been a significant correction in terms of the AI boom.

So before we get into their CFO's comments, Chamath, I'd love to hear your general response here.

Chamath Palihapitiya

I think it's fun to give these guys shit, but they're totally and completely uncorrelated. Every now and then you have a bad day. I've done thousands of hours on TV. I've had a couple of really bad days. You guys have been there.

I suspect that if Sam had to do it over, he wouldn't have said what he said in the way he said it. Even if he was joking, he would've practiced it a little bit more and just landed it.

So what's actually going on? I think right now we are in a period of getting a little risk-off and rebalancing. Why? There are 2 sets of things that are happening.

The first set of things is the market is learning to digest all of the CapEx that has happened, and they're all breathlessly trying to build models that try to predict what the ROI is of that spend. The second part of that is they're trying to figure out how this new spend will actually impact future earnings.

This is less to do with OpenAI, but it has much more to do with the big stalwarts of the Mag 7. Google's earnings were phenomenal. Their AI numbers were blazing hot. Facebook's was terrible. Apple is now in this really interesting place where it seems like they're going to cede their AI business to Google and pay them billions of dollars a year, like they get paid billions of dollars a year for search from them. I think that's what's happening.

The other part is, as you go into year-end, there's just a little bit of fall in the market, and people are like, “Let me just consolidate. Let me book some wins. Let me get ready for the new year. Let me tax-loss harvest.” Let me do all the things that people used to wait until mid-December to do. Now they are smart enough to know that the price action starts in mid-November.

Jason Calacanis

Hmm.

Chamath Palihapitiya

So I wouldn't pin this on Brad and Sam. I just think this is natural market machinations. But to be clear, we are very much getting into a phase of risk-off.

Jason Calacanis

All right.

2. The Bailout Backstop Debate

Yeah, and this got exacerbated, Sacks, because on Wednesday, OpenAI decided to be in the news again when their CFO, Sarah Friar, told The Wall Street Journal she hopes the U.S. government—that's you, Sacks—will backstop the financing of its $1.4 trillion in data centers.

Here's a direct quote: “The backstop, the guarantee that allows the financing to happen.” And she said that the federal guarantees would, quote, “really drop the cost of financing.” Of course it would. And this would allow OpenAI to borrow more money at lower rates from a much larger pool of lenders.

That went viral, and everybody said, “Oh my God.” It started feeding, I think, the narrative that maybe OpenAI is insolvent, in fact, and there’s no way for them to pay their bills, which is obviously a little bit ridiculous. People are trying to correlate this to the dot-com bust and then the great financial crisis.

But on Wednesday night, Friar walked back her comments: “I want to clarify my comments. OpenAI is not seeking a government backstop for our infrastructure commitments. I used the word ‘backstop,’ and it muddied the point. I was making the point that American strength in technology will come from building real industrial capacity, which requires the private sector and government playing their part.”

She also said that OpenAI was on pace to generate $13 billion. I’ll get to you on this one, Brad, because he took offense to the $13 billion revenue number and disputed that, so we’ll see if there’s some clarity there.

But, Sacks, you came over the top and tweeted that there will be no federal bailout. There are plenty of people available to pick up the mantle if OpenAI needed a bailout. You’ve got 5 major frontier model companies: Grok, Claude, Gemini—plenty of them.

So, Sacks, you came in, “Daddy’s home,” and you settled it. Daddy came home. Everybody has to sit down in the kitchen and explain what’s going on. Take us through how you think about this as our tsar of AI.

David Sacks

Well, this morning my entire feed was full of comments by people—analysts, consumers, businesspeople, and politicians—saying that we can’t allow OpenAI to have a federal bailout. I think they were connecting Sam’s original comments, or prickliness, to what Brad talked about: “Hey, do they have the money? Can they justify this?” They connected that with Friar’s comments that they need a backstop to say, “Hey, this company’s not solvent. It’s going out of business, and they’re demanding a federal bailout.” I think that’s how the pieces got put together.

What I said is, “Look, there’s not going to be a federal bailout for AI. It’s not going to happen.” We have 5 major frontier model companies right now, and there are new companies being formed all the time. If one of them fails, it’s going to go out of business, and the other ones are going to replace it. So nobody is talking about a bailout.

In fact, I would say that the AI sector is maybe one of the healthiest—meaning most competitive—sectors of the entire American economy right now, to the extent that you love ruthless competition driving innovation. That’s what we have right now. If one of these companies gets over its skis and ends up going bankrupt, the chips are going to fall where they may, and I’ve never heard anyone serious disputing that fact.

I also made the point, which I think is important, that to give OpenAI the benefit of the doubt, I don’t think anyone at OpenAI was asking for a bailout. If you watch the video with Sarah Friar, she’s clearly searching for the right word to describe what she’s trying to say, and then she settles on a word that she now regrets, which is “backstop.” Definitely not the right word.

So I don’t think they are asking for a bailout. I don’t know what she meant by backstop. It doesn’t make sense. I think this is a little bit of a tempest in a teapot.

What I think is important, and I think maybe where she was going, is that I do think we want to make it easier to build infrastructure in this country. That means making permitting easier and making power generation easier. That’s all about regulatory reform.

I think the goal here is to enable a rapid infrastructure build-out without increasing residential rates for electricity, which nobody wants. That’s in the process of potentially creating a little bit of a NIMBY backlash, when local communities fear that their electricity rates are going to go up because someone wants to build a data center. That’s the thing we have to combat.

But the way that you solve that problem is by making it easier for these AI companies to stand up their own power generation behind the meter, and that requires regulatory reform. That’s what the president has called for: allowing the AI companies to do behind-the-meter generation.

So no one’s talking about a bailout, nobody’s talking about a backstop. We are talking about—

Jason Calacanis

What about—

David Sacks

—making permitting easier and making it easier to do build-out. So “build out, not bailout” should be our motto here.

Jason Calacanis

Build out, not bailout. Now, there was a little bit of reading the tea leaves. There have been times when loans were given to incentivize an industry, so I just want to be clear with you: there are no discussions of Solyndra-type loans, Fisker—

David Sacks

Nobody’s discussed any of that with me.

Jason Calacanis

Okay.

David Sacks

So it’s not even on the radar, I would say, from the government standpoint.

Jason Calacanis

And why would it be if there are so many people trying to pour capital into this, Brad, and there are so many people trying to buy your shares in a company making $13 billion that’s currently valued at $500 billion, which I think is a 30-to-40 price-to-sales ratio? This company is fully valued, and people are still trying to buy the shares. So, Brad, wrap us up here.

I think it was brilliantly said by David. Listen, it’s a national imperative that we accelerate the build-out of AI infrastructure across the country. I’ve said before, the $4 trillion that Jensen Huang has estimated will be built out over the next 5 years is 10 times the size of the Manhattan Project, which was totally federally funded.

This is all being privately funded, but it wouldn’t be possible without the government—Secretary Wright, others, Burgum, et cetera—clearing the regulatory hurdles out of the way. You heard on the same pod that power is really the gating issue here, so it’s been amazing to see what the federal government is doing.

I think that’s what Sarah was trying to get to: they need to have a public-private partnership. They’re going to do their job and raise their money. “Backstop” was not the right choice of words.

I talked to Sarah this morning about this, and I know they are deeply grateful for what the federal government is doing to accelerate the build-out of power and infrastructure. The federal government could do that without taking risk on its own balance sheet.

In fact, we’ve seen some of the investments they’ve made as a result of the Japanese deal they got on tariffs. They can reinvest those dollars to accelerate some of the nuclear build-out, et cetera.

Chamath Palihapitiya

By the way, Sam just posted something about 15 minutes ago.

Jason Calacanis

Oh.

Chamath Palihapitiya

He was pretty authoritative in addressing the 3 critical questions. The first thing he says is that we—meaning OpenAI—will end the year on a $20 billion forward run rate, which means December revenue will be at least $1.666 billion. So we kind of know where the revenue’s going, from $1.2 billion to $1.6 billion over the next few months, which is a pretty staggering growth rate if they were at $13 billion and they’re going to end at $20 billion.

Then he addresses the whole “too big to fail” question and whether they want government sponsorship. He’s pretty unequivocal here.

Yeah.

Chamath Palihapitiya

I think this is a tempest in a teapot. People are on pins and needles. They’re agitated. Some people have had no gains; other people have had incredible gains. Everybody’s agitated.

I think we are getting into the risk-off phase for at least 2 or 3 months. We will be back firmly in risk-on mode in February, is my suspicion. But these next few months, I think people will overblow every random little thing.

Jason Calacanis

Well, in fairness, $1.4 trillion is a very large number. This is a number we’ve never seen one company say they’re going to do a build-out at.

Chamath Palihapitiya

Okay, well, can I take the other side of this?

Jason Calacanis

Please.

Chamath Palihapitiya

If it is—

Jason Calacanis

Yeah.

Chamath Palihapitiya

If I were the U.S. government, to the extent that we are doing public-private partnerships, if there were a way for U.S. taxpayers to own a piece of OpenAI, I would say, “Great.”

Jason Calacanis

I mean, so, Brad, will you sell—

Chamath Palihapitiya

I’m sorry—

Jason Calacanis

—your shares to the Trump sovereign wealth fund?

Chamath Palihapitiya

So hold on. Before everybody breathlessly complains, whether you see it or not, there is an enormous sovereign wealth fund being built by President Trump, and it is for the benefit of all American taxpayers.

To the extent that the people in government could underwrite an investment structure like they have done in things like MP Materials, which is way up, and things like Intel, which is way up, these guys are really good, smart people. Mike Grimes, Dave Shapiro, Stephen Feinberg, and his team at the DoD—they’re cutting hard deals, tough deals.

I don’t know. If they did do a deal with OpenAI, I think they’d probably get the best of it, and the American taxpayer would win. I wouldn’t be angry at that.

Jason Calacanis

Hmm.

I think, as you see from Sam’s tweet, they’re not looking for the government to invest. They’re not looking for a government bailout. What they are doing, and I think David said it well, is pushing us very hard as a matter of national security and economic growth to go faster, to accelerate, to build out infrastructure.

To give a little credit where credit is due, all of this build-out—all of the Stargate stuff that people were laughing about 18 months ago—thank God, as an American citizen, that we are running this fast. China has 100 nuclear fission plants under construction, and we were sitting on our hands.

So if anything, they’ve helped jump-start that conversation and get us moving faster, and I think that’s good for all of us.

3. The China AI Race

Jason Calacanis

Great segue, Brad. Thank you. Jensen told the FT straight up, quote, “China is going to win the AI race.” His argument is that U.S. state-by-state regulations and power constraints are making it harder for U.S. AI companies, as we’ve discussed here countless times, whereas the CCP is obviously just making it super affordable to run all those GPUs. NVIDIA put out the following statement from him: “As I have long said, China is nanoseconds behind America in AI. It’s vital that America wins by racing ahead and winning developers worldwide.” Obviously, he has—

Chamath Palihapitiya

He’s 100% right. I don’t know if you guys saw, but Cursor 2.0 launched this week. My team at 80/90 uses it. It’s an incredible product. Guess what they did? They swapped out Anthropic for an open-source Chinese model.

Jason Calacanis

Yep.

David Sacks

Do you know what they’re using? Is it like Kimi, or what is it?

Chamath Palihapitiya

I think it’s Qwen.

David Sacks

Qwen. Okay.

Chamath Palihapitiya

I think they’re using a spin of Qwen.

To be clear, they’re cloning these Chinese open-source models, but they are—

Jason Calacanis

They’re forking them.

Yes.

Jason Calacanis

And then running them, setting them up themselves, obviously.

Chamath Palihapitiya

My point is we’re right now running with one hand tied behind our back. We’re going to have to deal with 50 different sets of legislation from state legislators who think they know what AI is. They don’t. Sacks knows.

So there should be a federal framework, and that should be it. Meanwhile, the Chinese open-source models get better and better and better and better. We’re making technology decisions that tie our wagons to that steel thread. Jensen is right. We need to clean this up quickly.

David Sacks

I thought it was reasonable for certain politicians, especially Republican ones, to say, “Look, there’s not going to be a federal bailout for AI.” Great. We all agree. No one’s asking for it. But I was a little disappointed to see that some of them were associating a bailout with a single federal framework, as opposed to the patchwork of 50 state regulatory regimes.

If Republican governors think that they’re the ones who are going to be writing the rules, they’re sorely mistaken about this. 25% of the bills going through state legislatures are in 4 states: California, New York, Colorado, and Illinois. In other words, the biggest blue states. Those also happen to be the states where these big AI companies are all headquartered, with the one exception of xAI, which is in Texas.

These companies are in California. The blue states have the most market power, and if they end up creating the regulations, I just think it’s naive to think that the AI companies won’t write their models to those regulations of the blue states. I don’t think the red states are somehow going to find themselves exempted from the blue-state regulations that are being imposed.

I’ve talked on a previous podcast about how what the blue states are going for here is to reinsert DEI into AI models to achieve ideological capture. The way they do this is they don’t say, “We’re requiring DEI.” They say that they’re prohibiting algorithmic discrimination, which means that the model says something bad about a protected group. You end up with the same end result, which is, again, ideological capture.

I think all Republicans should be opposed to this. There’s only one way to stop it, which is with federal preemption. Otherwise, the states will do what they want, and the blue states will basically dominate.

Now, I think that part of what’s going on here is that Republicans have muscle memory around what happened during the Biden years. What happened during the Biden years is that the Biden administration pushed for censorship, shadow banning, and deplatforming, all that kind of stuff. They were working very closely with the big tech companies to push the censorship agenda.

The only pushback that Republicans were able to achieve was at the state level. So you had cases like Biden v. Missouri, where Senator Eric Schmitt was on the pod talking about that when he was attorney general. He was able to make a states’ rights argument to push back on the Biden censorship.

I think Republicans remember that, and they think, “Well, states’ rights is our solution.” But now we have a completely different situation. The board’s been completely reversed, where Republicans are in power in Washington, and the states are making a bunch of bad decisions with respect to AI.

I think, to be honest, we need to just realize that. The arguments that make the most sense right now are the Commerce Clause arguments. The Constitution wants to create a single national market for interstate commerce. AI clearly qualifies.

Let’s give President Trump, not Gavin Newsom or J.B. Pritzker or Kathy Hochul or Jared Polis, the ability to write the rules. Let’s have a single federal framework that will prevent ideological capture of AI and keep it unbiased, which every conservative should want.

Chamath Palihapitiya

I’ll just reiterate again: if you want to see the impact of having a state set of regulations that basically mugs a market up, just go and use your favorite AI tool and ask what happened when California passed CARB, which are the emission standards that forced the entire U.S. auto industry to have 2 sets of cars, one for California and one for the rest of the market.

What did it do? It completely flipped demand upside down on its head, and it has made it very difficult for the auto industry to be sustainable. If you apply that same idea across 4 states instead of just 1, across the most important technological revolution we’ve had, I just don’t think it’s going to be a good outcome.

Jason Calacanis

The steelman’s—

David Sacks

Right.

Jason Calacanis

—version of that, Chamath, would be that we got rid of smog in California and that it did an amazing job in terms of getting rid of pollution, which also matters. I’m trying to think of the steelman—

Chamath Palihapitiya

Are you sure that’s the steelman that you’ve come up with, or do you think that the tax credits did that?

They have the toughest—

Chamath Palihapitiya

Do you think making 2 types of everything was the way that solved smog, or do you think the $7,500 federal tax credit solved smog?

Well, no, but the smog regulations predated the EV ones. Those have been going on for decades.

Chamath Palihapitiya

Okay.

So it did get rid of smog. Using cleaner gas and having better exhausts on those cars would be the steelman of it. I’m not saying I’m for or against state regulations, but Sacks, aside from the federalism argument and states’ rights, have you heard any great defense of states having some say in how AI is deployed in their communities?

David Sacks

Well, they can have some say. There are definitely areas where you don’t preempt. You have to decide how wide the preemption is. But when it comes to things like notifications about model safety incidents, things like that, it doesn’t make sense to have model companies needing to report to 50 different states, 50 different agencies within those states, each with a different definition of what needs to be reported, each with different reporting deadlines.

Why would you have that? It doesn’t make any sense. Why would you allow the big blue states to essentially insert DEI into the models, which will affect the red states, too? You’re not going to be able to keep that out. If California pushes algorithmic discrimination, Florida, Texas, and Arkansas are going to be affected as well.

I think we need to use the opportunity we have right now, given that we have a majority in Washington, to set a sensible federal standard that preempts the excesses of the blue states. The constitutional arguments can be made either way. I personally believe in the Commerce Clause. But when it comes to the merits of the policy argument, we should let Donald Trump write these rules.

Let me just say one other thing. Part of what’s going on here on the right is that there’s so much anger toward the big tech companies for what happened during the Biden years with censorship and deplatforming that I think there’s just this knee-jerk reaction where we don’t want to do anything to help the tech companies; we just want to hurt them.

I think we have to have a more nuanced approach than that, because the question is: what will the result be? No one was more critical of the big tech companies engaging in censorship in Silicon Valley than me when Donald Trump—

Yeah.

David Sacks

—was kicked off every big tech platform. I think I was literally the only person in Silicon Valley who was publicly objecting to that on this podcast. I’m perfectly willing to criticize the big tech companies when they make a mistake, but when they’re engaging in healthy competition and innovation and we want to prevent ideological capture—and that’s what we’re talking about—let’s make sure it ends up in the right place, not just engaging in this knee-jerk anti-tech reaction, which will play into the hands of the blue-state governors.

We’re in this race to win AI globally, and one of the major concerns I have is that AI is becoming deeply unpopular in America, right? Silicon Valley is losing the battle around AI. Doomers are now scaring people about jobs. They think all these job cuts that are going on in America are the result of AI.

Number 2, they’re seeing their electric bills go up, and they think that’s also the result of AI. I’ve talked with a lot of Republican senators and House members who say they’re afraid to mention the words “AI” because their popularity ratings go down.

We need to get on the other side of that because that is a losing proposition for America. If what takes hold here is that it's politically popular to push back against AI, then, David, I think that's what you're seeing at these state levels—

David Sacks

Right.

…with Republican governors as well. I think both of those are false narratives, but we need to get on the other side. In China, they're not going to slow down. So if we do an own goal here and slow down because we think somehow that this is the path to greater economic growth, it's going to be a real problem for both national security and economic security 3 to 5 years down the line.

David Sacks

Can I build on that, actually? In terms of the public discourse, it's true that the doomer narratives have had this tremendous effect that you can see in the poll numbers, and then the politicians feel like they can basically play into that in one way or another. But where do these narratives come from? Three big tech billionaires who are on the left contributed over $1 billion to these doomer think tanks. Basically, it was Dustin Moskovitz, Jaan Tallinn, and Vitalik Buterin.

From Open Philanthropy and some of these other entities, they have spun up hundreds of these astroturfed organizations that are spending literally hundreds of millions of dollars to spread these doomer narratives. A writer named Nirat Waisblatt on Substack has basically broken down how this all works. It's actually a great article. I'll put it on the screen, so people need to understand that these narratives are coming from somewhere. They're astroturfed. I don't think they're true, and the narratives are contradictory.

So let me give you an example. Right now, the 2 biggest narratives that we're seeing on social media and in mainstream media are, number 1, the idea that there's a huge AI bubble right now. In other words, it's all totally fake. The other biggest narrative is that AI is on the verge of superintelligence, and we're all going to get replaced.

Chamath Palihapitiya

Also fake.

David Sacks

Right? In other words, AI is completely real and super powerful. Well, these 2 things are contradictory. If it's a bubble, it's not going to be on the verge of superintelligence. And if AI is really that powerful, then obviously, on some level, this economic activity is justified.

So these narratives are completely contradictory. I think it's possible to believe in neither one of them, which is where I'm at, but it makes no sense for people to believe in both. And you literally have the same people on social media and in mainstream media pushing both of these doomer narratives. So I think we need to increase somehow our immune defenses or our—

Chamath Palihapitiya

Antibodies.

David Sacks

…we need to improve our antibodies to, I think, these memes that are being pushed out by groups that have these weird doomer ideologies like effective altruism. They literally just want progress to stop on AI, and if we do that, as Brad's saying, China will end up winning this AI race. Progress in AI is not going to stop. It'll just all be in China.

4. OpenAI Revenue Challenge

Jason Calacanis

Brad, let's talk brass tacks before we move on to our next topic about ChatGPT and its revenue. If they're at a $20 billion revenue run rate, I think it's pretty well known that the majority of their revenue is consumer. 75% is the number I heard. You can tell me if that aligns with your estimates as well. It costs $20 a month, $240 a year. That's about 60 million paid subscriptions a year.

And on the other side, you've got Anthropic, which has kind of got the opposite, right? They're mostly APIs. Do those roughly correlate with what you know, Brad?

Well, I would just say they correlate with what The Information has reported as leaked data from both companies. And that makes them both the fastest-growing companies in the history of Silicon Valley. Let's just be clear about that.

Jason Calacanis

But specifically, the 75% coming from consumer at OpenAI.

Yeah, more—

Jason Calacanis

So—

I think it's well known more comes from consumer at OpenAI, and more comes from enterprise at Anthropic.

Jason Calacanis

Yeah.

So, 2 challenging questions for you since you have a big bet here. Google and Apple make these products free. They have pretty robust ad networks, so that's a massive headwind. And then, on top of that, in the startup community, we just talked about Cursor, people are not trusting OpenAI and its API anymore because they know OpenAI is creating competing services. So there is a big movement in the startup community not to use OpenAI's API products.

How confident are you that you can overcome those 2 headwinds—free for consumers and better products? Gemini is a great product. Grok's a great product. Claude's a great product. There are a lot of great products out there, and I don't think consumers can tell the difference. Why would they pay $240 if they can get it for free from Google?

And second, startups are realizing, “Hey, Sam has to make a lot of money. Therefore, he's going to do what Microsoft did.” There were companies called Lotus 1-2-3 and WordPerfect that were on the Microsoft platform, Windows, and then Microsoft killed them. People right now are experiencing that from OpenAI. How concerned are you about the revenue growth?

Yeah, listen. I'm betting on the supercycle. This is the biggest supercycle of all of our lives. I'm an investor in OpenAI, Anthropic, Google, Microsoft, Nvidia, et cetera. And so I don't think you have to make the call right now on one of these companies winning. The fact of the matter, as Sacks said, is that we have one of the most vibrant and competitive ecosystems of AI in the world.

I love the fact that Sundar's coming off the mat swinging. I think Gemini 3 is going to be great at Google. They may, in fact, make it free. I think Apple—I love seeing Apple pay Google to make Siri better. I think that's going to be a great consumer experience. And the only way that OpenAI wins is that they've got to build a product that we all love.

You've seen those cohort curves. The reality is they are the verb at the moment. It's theirs to lose in consumer. The cohort curves are things of dreams, right? This is the retention rate and the engagement rate as people stay with the service longer. And, by the way, Anthropic's numbers, despite Cursor doing some of its own thing, are off the charts.

I think that this market is as big or bigger than current estimates are out there, but it's not going to be a straight line up and to the right. We're going to have these moments, as Chamath said, of risk-off panic, just as we did with the internet, just as we did with mobile, just as we did with cloud. The key here, as an investor, is conviction. There is a massive conviction tax to be paid, right? If you lack conviction and sell when these things are down, I'm going to bet on the supercycle, but I'm betting over a much longer period of time than most people.

Jason Calacanis

And among the startups out there, they basically believe Claude is not trying to take their business, and Claude has been very careful to say, “Hey, we're not going to encroach into the application layer.” So, all right, yeah. If I was starting—

Chamath Palihapitiya

By the way, the most impressive revenue chart, when you showed that leak from OpenAI, was Anthropic's. If that revenue chart is real, what's really impressive is Anthropic is not really in a J-curve at all, and they get to very similar points of free cash flow but will not have burned through near as much capital to get there. That's my singularly interesting observation about this chart.

Can I say again, those numbers are—

Chamath Palihapitiya

If it's real.

Yeah.

Chamath Palihapitiya

If it's real.

As reported by The Information. But what I would say about that is this: I've been forecasting companies for 25 years. I know the numbers today. Sam just told us a $20 billion run rate by the end of the year for OpenAI. Everybody who is forecasting 3 years out on these companies is totally guessing, right? That's why I said time is on your side if you're betting over a 5- to 10-year horizon.

But if you think with precision that the company itself can forecast what's going to be happening in 3 years, I think you're misleading yourself, right? So we'll see. I think it's going to be a lot bigger. It could be even bigger than those numbers. But those forecasts are highly uncertain because the rate of growth has never been seen before.

Chamath Palihapitiya

How do you think about expenses, then? Because you're going to have to have some sense of whether the spend is accurate or reasonable.

I think you have to build in an expense structure that has the flexibility so, if the numbers don't show up, you have the ability to extend your runway. So that's why I was saying I don't think this $1.4 trillion—I think it's kind of the red herring out there. I think it's kind of a fake, made-up number. It's all of the obligations of all of the deals that have been announced.

And the truth of the matter is, only a portion of that is borne by OpenAI, and I'm certain there's flexibility in all of those deals to match the expenses with the revenue side. But listen. Let's steelman the opposite case, Chamath. Let's just say that all of these revenues start flatlining for Anthropic, for xAI, for Google. People aren't willing to pay for these products. Then our CapEx build-out for AI is going to be a lot slower.

Because at the end of the day, there's got to be somebody either in the enterprise or a consumer willing to pay real money to pay back all of this infrastructure build-out, or it doesn't happen.

Jason Calacanis

All right, listen, we’ve got to talk about markets. We’ve got Brad here, our fifth bestie. The stock market’s pulling back, as Chamath just said. We’re going to be risk-off, and we’ve been getting a lot of data. There’s a lot of hand-wringing going on, Brad.

GDP growth—maybe that’s mostly due to AI. The unemployment rate’s ticking up. Inflation’s ticking up. Lots of concerning signs. Help us make sense of this, Brad.

I think Chamath teed this up perfectly. Let’s rewind the clock a little bit to April of this year. I think on an intraday basis, the Nasdaq was down 20% year to date in April. We’re now up 20%, so a 40% move higher in a few months. Same with the S&P: I think it was down over 10%, and now it’s up 14% on the year. So we’ve had some pretty massive moves, and I think you have to reflect on where we are.

We used to do this on market checks when I was on, and you’d say, “Where is Altimeter?” Early in the year, when we were worried about tariffs, Altimeter was positioned small. By May, we thought they would land the plane with the Besant consensus on trade. We’d get the Big Beautiful Bill passed, and we went to extra-large positioning. We’ve been there most of the year.

So now we’re back to—Chamath just nailed it—we’re back to medium, medium-small positioning in the market. I’ll walk through a couple of slides as to the reasoning for that. Maybe we can kick it around a little bit.

The first one is that there are growing signs the consumer is pulling back. You heard it out of Chipotle, you heard it out of Cava, you heard it out of NCL, and you heard it out of JetBlue. The nature of it is that we have this two-tier economy: the low-end consumer is faltering, while the higher-end consumer is hanging in there. But the consumer situation is making people nervous.

Compounding that, U.S. credit card delinquencies are now back to 2009 levels, right? And so you have the consumer cracking a little bit, delinquencies, and now we’re seeing regional banks roll over, et cetera. We’re seeing the credit markets beginning to crack a little bit, with credit spreads blowing out a little bit.

And then, if you go to this next slide, this goes to what Sacks has talked a lot about on this pod: we’re still in highly restrictive territory when you look at the 10-year TIPS. The Fed still has the market tight because they’re seeing the market at all-time highs. They’re seeing AI stocks rip. But under the surface, I think there’s a lot of concern and questions about what’s going on.

The good news about this is that we still have firepower. On this next slide, earnings have come in really strong. We have 70% of companies beating, and they’re beating by wide margins in earnings. But if you look at the forecast, earnings have come in about 11% higher than last year, and the stocks are up about 13% or 14%. The real question is what happens as we roll into next year.

Scott Besson, on the one hand, is saying, “Listen, in Q1 we’re going to have some big tax refunds because of no tax on tips, no tax on overtime,” and he thinks that’ll give it a boost. But clearly, you hear them pounding the table that we need to get rates down. I tend to agree. The low-end consumer who pays a lot on their credit cards, on their car loan, et cetera, is clearly hurting. We see these layoffs, and these layoffs are hurting some of these people as well.

When you look at the multiples of the market, how do we look at it from an Altimeter perspective? We still believe that owning all of compute, as we have for 3 years, is an AI trade that has room to run. But obviously, multiples have come up a lot. I talked about a 40% move from the bottom this year, so you can just take your position size and make it a little bit smaller.

In the world of less or more, I think Chamath’s exactly right. There’s a pause as we head in over the course of the next couple of months, reflected by the stuff that we’re doing. But I do think the economy is set up as we head into next year.

Chamath Palihapitiya

There’s been a total decoupling.

Yeah.

Chamath Palihapitiya

It used to be the case that as goes Main Street, so goes Wall Street, or vice versa. The unfortunate reality is that there’s a handful of companies that have bids that are, as you said, totally speculative and well into the future, but they are so well bid and so highly valued that they drag the entire indices forward, even though underneath the waterline, you’re leaving 493 companies behind.

The reality is that those companies that more accurately reflect what’s happening to middle-income America are not doing so well. And so we need to find some visible wins in the domestic policy arena. I think that’s what we need now, because if you think of each year of a presidency as an act in 3 parts, I think act 1 was tariffs. I think act 2 has been foreign policy, and now it’s an opportunity for act 3 to refocus on domestic earnings and the domestic wealth effect of middle-income Americans. I think that’s what has to happen.

There are a lot of clever ways, by the way, that we can do this. I’ve mentioned this before, but I think the really interesting opportunity for us is to use the money that’s been committed by all of these countries as part of their trade negotiations. If you add up the money that was committed from Japan to South Korea to the Middle Eastern countries, you’re talking trillions of dollars. I think it’s like $3.2 trillion is the number.

So it’s an enormous amount of capital that can improve the state of earnings of the middle class, and I think we need to figure out a way to more aggressively put that money to work now. I think that’s what needs to happen, I think. If you look at some of these other markets, it’s really incredible. Bitcoin is about to break through $100,000 to the downside, which I think is a psychological barrier that probably has another 5% or 10% more to run. There’s a dispersion happening in the Magnificent 7 as of this past quarter’s earnings. So I think all of it has to get sorted out.

You nailed it. 29 days ago, I talked to you guys a little bit about the net approval rating here and how the country’s feeling, and you can see Trump’s net approval rating: negative 9.4. We’re going to get into the election as well, because it dovetails really nicely with this discussion.

It’s gotten worse. Here are the numbers from November. It’s gotten 30% worse now. Trump’s down 13%. I attribute it, I think, to the fact that we’ve had all this layoff news. Inflation ticked up to 3%.

Everybody knows it was down at 2.3%. We had made great progress on it, but now it has gone back up. There is a lot of survey data coming out right now, and I think if you look at this next chart, it is super interesting. What gets all the press is the foreign policy, right? We are talking about wars ending and a great job in the Middle East for Trump, Jared, and the team.

Border security and immigration were, of course, a big win in terms of shutting the border. It is a bit controversial on ICE, but that is my personal stuff, so I will leave it out here and just talk about the brass tacks. If you look at where the country feels Trump has fallen short, it is where he is strongest, or where he was elected for being so strong. Look at these last 3 here, Chamath: the economy, where 63% believe Trump has fallen short; looking out for the middle class, where 65% of the country believes he has fallen short; and inflation and the cost of living, at 66%.

5. The AI Jobs Fight

If you look at this next chart, this one has been trending online. I have been talking about unemployment, and I have been attributing it to AI. I know some people think I am crying wolf here, but the statistics are starting to trend toward my position, I believe. If you look at people who are 20 to 24, they are at 9.2% unemployment, and it is spiking. Why is it spiking?

I can tell you from the front lines: I hire young people, I am at startups with young people in them, and I talk to them about their contemporaries. As companies cut and use AI to solve problems, they are saying that hiring young people and training them is not as rewarding or efficient as training an AI to do the same task.

Chamath Palihapitiya

Why do you keep thinking it is AI? Even with the Amazon cuts, Andy Jassy said on the earnings call, “This is not AI. This is us digesting all this hiring through ZIRP and DEI.”

I can answer that. He did say that. He also said in his memo before that he thought there would be a lot of changes because of AI, and they had the leaked memo saying that they would not be hiring those people.

But I am talking specifically about young people. I will take Jassy at his word: those 30,000 people are just redundancies. But what I am seeing on the ground, because I work with startups and with young people who are in them, is that their contemporaries are not finding jobs. The companies we see, and the companies selling the technology solutions, are getting rid of what I will call entry-level white-collar work.

Jason Calacanis

In addition to this, student-loan delinquency is also going up massively. They had a big hiatus, as you know.

But does this not indicate that there is no ROI to these expensive degrees, in a lot of cases, that these kids are being sold? All this debt is being foisted on them.

Jason Calacanis

Yes.

This has been going on since before November 2022, when ChatGPT launched. I do not think AI has had enough time to have a huge impact on this yet.

Jason Calacanis

Yeah, but that is not—

In this case, I think you are wrong, respectfully, even though you are the czar of AI. I am seeing it with young people, and I am seeing it with the software that the companies I am investing in are deploying at other companies. But we will see who is right in the coming months.

Inflation is back up to 3%, and that is going to make it particularly hard to cut rates. When we go back to slide 3, the reason people feel Trump has fallen so short on the economy and taking care of the middle class is inflation. Trump has said over and over again, and some people in the administration have been very publicly gaslighting the country, in my opinion, saying, “Hey, inflation is not happening.”

Inflation is happening, and Trump sold people on the idea that prices would go down. We are back up to 3% inflation, folks. Not only is it not going down, it is going up. When you put together what people are seeing—the focus on a ballroom being added, the focus on tariffs, things Americans are not seeing reflected in their wages, and their kids not being able to get jobs with 9% unemployment—this is what is concerning Americans.

Trump has failed the middle class. That is the bottom line here.

David Sacks

I do not think that is true.

Jason Calacanis

The country does: 66% of them.

David Sacks

I think everybody can find a piece of data to hang their bias on. What this most recent spate of elections shows is that most people are looking for some form of a price break. If you are a New York mayoral candidate and you offer giveaways, you get a lot of attention. If you are running for governor and you offer giveaways, you get a lot of attention.

The question is: Why is that happening? I think it is fair to say that the reason is that we have spent a lot of time setting the table for a wave of domestic-policy initiatives, and now is the opportunity to follow through on those. We need to figure out where the pressure is, and I think Trump is doing this. For example, today he announced that Eli Lilly and Novo Nordisk would be selling GLP-1s for $149 a month. I think that is good. Is it as expansive as it could be? Probably not, but there are all of these other programs that need to be addressed.

The student-loan issue is a key one. It is time for legislation that says we are going to stop federally underwriting these loans. Force them into the public market, force them into the open, allow transparent pricing of loans and risk, and you can stop that curve dead in its tracks. We need to clean up the AI stuff.

There are a bunch of things now where, when the government shutdown ends, we just have to get to a legislative agenda and start to fix domestic policy. But you cannot do that when folks will not even come together and will not even end the shutdown. Part of this is a frustration that is going to keep brewing. I think it is less a reflection on him and more a reflection of the point we are at: people want these structural problems fixed.

That can only happen through legislation. But that cannot happen if the House and the Senate are not even convening because there is still a shutdown.

Jason Calacanis

Yeah. People are seeing equity holders like us and people in crypto get rich. We are all up 20% or 30% this year, and the working man and woman in this country are not. That is the truth of it.

I think that is why 65% of the country blame the Republicans for the shutdown. People are not getting food stamps, people are not getting paid, and people see their health-care prices about to spike. We are doing all these tax cuts. It is great that the economy is ripping for equity holders, but there is a reason why two-thirds of the country feels this administration has failed the working man.

It is time to get back to work. It is time for Trump and his group to focus on the people who put him there, not just people like us who are already extremely affluent. What do you think, Brad?

If you are right and we have inflation over 3% next November, higher youth unemployment, an economy growing at only 1.5%, and interest rates staying high, then I think it will be a rough midterm for the Republicans.

I do not actually think that is the flight path for the country. I think you are going to get 3 or 4 rate cuts. I think you are going to see a reacceleration of GDP. I do not think these job cuts are the result of AI, but that is what makes a market.

You heard the president say after the elections this week that there are a lot of things around affordability that have to be delivered and are not being delivered. The number-one affordability issue starts with lower interest rates, and he has been pounding the table on that all year long.

A big mistake by Republicans in this last election cycle is that they were not talking about affordability, while the people they lost to were talking about affordability. On that score, you are correct. But I happen to think that inflation is going to continue rolling over. I think you are seeing some one-time effects in here, so I will take the under, Jason—

Jason Calacanis

Cool. Yeah.

—I will take the under next year.

Jason Calacanis

I am just sharing the data.

Can I respond?

Jason Calacanis

Of course. That is why we are here.

You are really good at finding these cherry-picked charts, so let me show you a couple of them. With respect to—

Jason Calacanis

How are they cherry-picked?

David Sacks

Well, as Chamath said, you find charts or polls that support the point you are trying to make, which is fine. But let me show you a couple of other ones to provide some balance.

U.S. white-collar jobs as a percentage of total employment in the U.S.—a very steady line. You see a blip there around COVID because so many blue-collar jobs were lost. That caused the percentage of white-collar jobs to bounce up. But post-COVID, it has been on a very stable trajectory.

If it were true that we were seeing massive AI job loss, you would see the percentage of white-collar jobs dropping in the economy. That has not happened, okay? Every time there is some sort of job-loss story, you glom onto it being AI-related, like you did with Amazon, and then Andy Jassy comes out and refutes that. Then you are like, “Oh, no, no, I am talking about something in the future.”

Chamath Palihapitiya

Just one note about your chart here: that chart ends in the 1st quarter of 2025, so it is missing the last 2 quarters.

David Sacks

So now you're saying that all of a sudden this has just happened in the last 6 months?

David Friedberg

No, it's just flat. You can see it's flat there for the 4 years—the percentage on your own chart—and it only goes to Q1. But sure, keep going.

David Sacks

I don't see a big change here in the percentage of white-collar jobs.

David Friedberg

I'm not even sure this is relevant without the last 6 months of data.

David Sacks

Of course it is. If there was some sort of job shock in the economy, it would be white-collar jobs that were affected, right?

You're all over the place with this narrative. Whenever we point out that—

No, I was very clear. I think it's happening with young folks, and the number of layoffs right now is the largest number of layoffs we've had in a quarter since 2003. So this has been a lot of layoffs. I know it's hard to hear, Sax, but this has been the largest number of layoffs in 20 years.

David Friedberg

It's not hard to hear. Look, you keep bringing up these anecdotes, and the plural of anecdotes is not data. In any event, when I look at this chart, I see that there is no job-loss shock in the economy, and that's what you've been claiming. Every time we specifically refute your anecdote, you're like, “Oh, no, no, no, that's going to happen in the future.”

So you don't know what the cause of that is. The cause of that could be the fact that all these kids graduating from college are woke and majored in degrees that don't have economic value. You don't know what the problem is. You're just glomming onto this AI story because you just like it. You like this narrative, and you're repeating this doomer narrative that's been fed to the narrative by the media.

No, it's not doomer.

It is a doomer narrative.

David Sacks

No, I told you what it is. I'm just saying, the anecdotal stuff I see on the front lines is what I base it on. It's not just me glomming onto random data.

David Friedberg

Okay, the second chart I want to bring up: What you see here is that median real wages and earnings are going up. Now, if it's true that inflation is high and wages don't increase, then people's purchasing power will go down. But what we're seeing here is that purchasing power is still going up, and we have seen that during the first year of the Trump administration.

It is true that Republicans need to focus on affordability. The price of gas and eggs, for example, has gone down, and we need to keep making progress in that area. I don't know if you saw the tweet by JD Vance, but he basically said that Republicans need to focus on the domestic picture, and they will.

Now, if you're talking about the election, one thing I want to point out is that all the major races were in blue territory, and we got blue results. So this wasn't unexpected, and furthermore, Trump wasn't on the ballot.

I concede to you that Republicans have a problem: When Trump is not on the ballot, our voter turnout is lower. In this election, the only thing that was a little disappointing, I think, was the gap in excitement between the Democrats and the Republicans. Democrats were able to turn out their voters; Republicans were not.

Again, that's a problem that Republicans are going to have to solve for the midterms and in 2028: How do we turn out our base when Trump is not on the ballot? But you can't really lay that problem at Trump's feet. When he's on the ballot, he turns out the base and he wins, and that is the problem that we have to solve.

Okay.

6. The Filibuster Nuclear Option

David Sacks

I agree with the part of it that said Republicans now need to focus on the domestic picture. But one of the biggest problems we have right now is that the government is shut down. Who does the public blame for that? Unfortunately, they're going to blame the party that's in power, not realizing that the reason the Democrats are able to shut down the government is because of the filibuster.

Unfortunately, I think the public perceives that they gave power to Republicans, but the government is shut down. The reason for that is that Democrats can shut down the government with just 41 votes, basically, and President Trump has called for getting rid of the filibuster. If the Democrats won't reopen the government, I think we should do that.

By the way, just to go off on this tangent for a second, most people don't understand the filibuster, how it works, or why that's even a part of democracy. Basically, what it says is that for votes where the filibuster is in play, you need 60 votes. You need a supermajority in the Senate as opposed to 50.

If you ask most people in the country, “What is democracy?” they would say it's 50 plus 1 in the House, 50 plus 1 in the Senate, and the presidency. That's all you need. But that's not true.

The question is, if Democrats won't reopen the government, then I think Republicans are within their rights to get rid of the filibuster, and we know the Democrats will do it. The next time the Democrats have the trifecta, they will absolutely get rid of the filibuster.

David Friedberg

We know that's true because when we interviewed Joe Manchin, he was very clear. He was asked by Schumer and Biden, “Let's vote out the filibuster,” and he drew a bright line in the sand and said, “No.” But you're right, they would have.

David Sacks

We're getting to a place where, if these folks in the Senate and the House don't get back to work soon, there will be no domestic-policy agenda that gets passed. There's critical crypto legislation, AI legislation, domestic health care, and other legislation that has to have a chance to see the light of day. If you're not going to show up, then it doesn't leave many other options except to get rid of the filibuster.

That has all these downstream effects for when you're not in charge, right?

Jason Calacanis

Because then it'll be a simple majority for the other team as well.

David Sacks

Right.

There are 2 Democrats who are opposed to getting rid of the filibuster. It was Manchin and Sinema. They're now both gone. So the moderate Democrats are largely out of the caucus now, and if and when—because at some point the Democrats will return to power—you know they're going to get rid of the filibuster.

Here's the crazy thing: You can get rid of the filibuster with 50 votes in the Senate. I don't think people realize that. Basically, you've got this custom, because that's all it is—a custom, convention, or tradition in the Senate—of having this filibuster vote. But the majority can just get rid of it, and the Democrats have already said they're going to.

So why would the Republicans foolishly abide by this sort of gentleman's agreement that doesn't exist anymore and hold themselves to a 60-vote majority? The country gave President Trump a mandate. They gave him a sweeping reelection: 7 out of 7 swing states, the House, and the Senate. They want results, and those results are being thwarted by a government shutdown and, more generally, by this crazy filibuster rule.

Why wouldn't we get rid of it now and actually pass the reform that the country wants? Then let's be judged on that in the midterms and in 2028.

Jason Calacanis

That's the calculus.

David Sacks

And if it ends up not being good, then we'll pay the price.

Jason Calacanis

That's the calculus. All right. So we should get to the election results. And just as—this is the chart I was mentioning, Sacks—here are the layoffs for any October, highest since 2003.

We're all trying to figure things out in real time here. But it's not just the All-In Pod talking about what's going on with jobs. It's a big discussion on CNBC and everywhere in between, on social media, et cetera.

David Sacks

I know, and I remember that article that got pulled up when we discussed Amazon. It was basically about automating its warehouses, and you were saying that this was all AI-related job loss. The article itself said that they bought a robotics company a decade ago, and this was all in line with what they've been planning for a decade. The humanoid robots aren't even here.

Then you're like, “Oh, yeah, yeah, yeah, I'm talking about the Tesla Optimus robot that'll be here in a few years.” You keep pivoting between what's going to happen in the future and what's happening right now.

Jason Calacanis

I know you're trying to win the debate club, but you're misconstruing my point. You're trying to spin my point of view, and you're the greatest debate club captain. I'm not trying to debate. I'm just putting out facts here for the audience.

David Sacks

You are debating me.

Jason Calacanis

No, I'm just putting out facts here as the world's greatest moderator.

David Sacks

You're putting out spin. It's okay. Spin.

Jason Calacanis

It's not spin.

David Sacks

Go ahead, spin. Spin away.

Jason Calacanis

Okay.

David Sacks

Don't pretend you're not debating when you're debating. Just debate. It's okay. Don't be a—say it. It's okay to debate. Just disagree with me. Don't pretend you're not.

Jason Calacanis

No, no. Here's the thing. The 600,000 jobs were what they were saying in a leaked document at Amazon that they were not going to fill in the automation group.

Chamath Palihapitiya

You want to be a defender of the weak and the blighted, and the problem is—

Jason Calacanis

No, once again, you're trying to diminish me by saying this is bullshit.

Chamath Palihapitiya

I'm not diminishing you.

Jason Calacanis

What I'm trying to do is balance out the fact that you two guys are captured and are no longer objective on this podcast. I'm the objective voice here of reality.

Chamath Palihapitiya

No.

Jason Calacanis

You said something about a company where the CEO said the exact opposite, and you won't take him at his word.

Chamath Palihapitiya

He said that a week later. I think—

Jason Calacanis

Okay, so you get a week to make up your own story. That doesn't make it right. He was very clear.

No, no, here's what I'll tell you. They literally were saying in the leaked documents that they had a PR crisis on their hands, and they were going to try to spin the AI job losses.

Chamath Palihapitiya

Jason.

Jason Calacanis

And part of what he said they were going to do was—

Chamath Palihapitiya

The CEO of a multitrillion-dollar public company said on the record and filed with the SEC—

Jason Calacanis

Sure.

Chamath Palihapitiya

— that the reason he did these layoffs was because of ZIRP and DEI.

Jason Calacanis

There's 2 sets of layoffs. One of them is the 30,000 white-collar people. Sure, we'll take him at his word. The other one is that they're canceling jobs explicitly because of the AI that they're planning to deploy and that they're making PR plans to cover themselves for how bad that looks to them. Let's move on to Mamdani.

David Friedberg

I want to read you something from Morgan Stanley this week on this question of whether the job losses are caused by AI or whether they're caused by something else. The title of this section is “Flatter Is Faster.”

It says, “You know, the cynical take has always been that this just wouldn't last, that companies wouldn't be able to maintain this level of discipline, and that they were doomed to repeat the mistakes of the past. It's never different this time, as they say, but the cynical take seems wrong again, and the number of examples of companies that have adopted this mindset, culture, behavior of efficiency and getting fit appears to be growing quarterly. Guess what? It's cool to have fat margins, no debt, maximum flexibility, and be efficient and move faster.”

Credit to Ashton Curtis. They're saying they're out there talking to the companies. I'm not saying that there's no AI effect, but I just think the idea that you're going to hang all these layoffs on AI—

Jason Calacanis

I would never hang them all on AI.

David Friedberg

Yeah.

Jason Calacanis

I think there are—

Hold on. I'm going to be very clear here.

David Friedberg

Yeah.

Jason Calacanis

I am not hanging them all on AI.

David Friedberg

Okay, good.

Jason Calacanis

I'm saying for the young folks, that's what I suspect is happening there. For the existing folks, I do agree that they're all getting rid of excess fat, and they want to be leaner, and they want to juice their earnings. Those are 2 different things.

7. Mamdani Wins New York

Now let's go on to Mamdani because it's getting boring. We're starting to circle here. Zohran has won New York City. The Dems won across the board. Zohran won every borough except Staten Island. He finished with 50.4% of the vote, a 9-point lead over Cuomo. Curtis Sliwa was a distant third at 7%.

No surprise if you're on Poly Market, because they have been saying this for months. Pull up the chart, Producer Nick. Thank you. He's been a big favorite since he crushed Cuomo in the primary in June, and he's held the lead the whole way.

He's a self-proclaimed democratic socialist. Some people think he's a communist. He's the first Muslim mayor and the youngest in over 100 years. Big promises seem to have resonated with young folks, especially women in New York: affordability, rent freezes, free public transit, higher taxes on the rich. You're going to pay an extra 2%. I think it'll be 54%, maybe the highest tax rate in New York, the highest tax rate in the nation.

Hilariously, he added supervillain Lina Khan to his transition team, and she's going to go break up the bodegas. Shout-out to our friend who's not here today, the sultan of science, David Friedberg, who predicted the rise of socialism on this very show about 6 months before anybody knew who Zohran was. Here's the clip. Give him his flowers since he's not here.

David Friedberg

The party line is that socialism was defeated in this election cycle, and that there was a resounding vote from the American populace against socialism. I actually think my contrarian belief is that we'll see a dramatic rise in socialist movements in 2025 in the United States.

We're going to see an unleashing of economic growth because of deregulation and AI. There are going to be some parts of the economy that are going to be big winners and some parts of the economy that are going to be big losers. When you have this sort of change this quickly, there are often large contingents of people that are left behind, and when that happens, I do think that the socialist policies and the socialist movements gain steam.

Growth does not mean that it benefits everyone equally, and I think that some folks will see people go from being billionaires to $100 billionaires to the world's first trillionaire, and it will also start to fuel this rise. So I think that we will see—

Jason Calacanis

Hmm.

David Friedberg

—an increase in the breadth and depth of socialist movements in the United States.

Jason Calacanis

Chamath, what can we learn from the Mamdani effect?

8. Fixing The Generational Compact

Chamath Palihapitiya

Peter Thiel predicted this in 2020. There were a bunch of leaked memos between Peter, Zuck, and Andreessen, but he put his finger on it in January 2020. I retweeted it. Basically, what Thiel said was, “From the perspective of a broken generational compact, there seems to be a pretty straightforward answer to me, namely that when one has too much student debt or if housing is too unaffordable, then one will have negative capital for a long time and/or find it very hard to start accumulating capital in the form of real estate. And if one has no stake in the capitalist system, then one may well turn against it.”

This has been floating around for years, and I think Thiel, yet again, has seen the forest for the trees many years ago. So what's happening? I think we've put our finger on it. We need to come back and now focus on domestic policy and fix some of these core pernicious issues.

One is clearly that we now need to address how student loans are underwritten. We cannot allow generation after generation of people to graduate with degrees that they don't quite understand and hundreds and hundreds of thousands of dollars of debt that they have zero chance of paying off. That is a horrible way to start your life, and we have not done right by these folks.

We need the free market to be able to tell somebody, as hard as it may sound to hear, that an art history PhD will cost you $800,000, so that the people who take it have the money or are willing to bear that cost. Meanwhile, if you became an electrician, it would only cost you $40,000 or $50,000, and you could have an incredible life. Or if you went and got a degree in AI or stats or something. My point is that we need to differentially price degrees based on the value and the earnings power that they create for people.

That is a policy-level initiative that needs to cascade through America. That cannot happen if the government is not working, so we need to fix that. The problem with housing is much more state and local, and I'm not sure I have a great diagnosis for how to fix that, except that certain states just need to get their act together.

I mean, in California, we have absolutely abysmal building regulations that prevent anything and anybody from doing anything. So that's what we need to do. We need to fix these things legislatively. We need to do it right away, and we need to fix this broken generational compact. This was the first moment in years where I have now become sympathetic to this idea of student-loan forgiveness.

Jason Calacanis

Hmm.

Chamath Palihapitiya

I was never sympathetic to this idea. I am sympathetic to it now.

Jason Calacanis

Why are you sympathetic to it, Chamath?

Chamath Palihapitiya

Because I think that we should have fixed this problem a long time ago. We should not have allowed these loans to be underwritten the way they were for so long, and we should not have allowed all of these effective subsidies to pervert the free market's ability to tell people that some of these degrees were not worth their time.

Jason Calacanis

Right.

Chamath Palihapitiya

We have Palantir today saying that they are not going to hire from college anymore.

David Sacks

Let me build on that. Just on the narrow loan-forgiveness point, I actually agree that maybe loans should be forgiven if you get a total overhaul of the system.

Jason Calacanis

Yes.

David Sacks

What you don't want to do is acknowledge that all the loans are bad and they need to be written off, and then continue making them. That was the problem with, I think, the Biden loan-forgiveness program: he's going to write off trillions of dollars of loans while keeping funding the system without any reform.

So let's talk about a big reform package where we completely re-underwrite how we do this, and maybe some loan forgiveness can be part of that, so that we don't have all these kids graduating who are basically socialists because they're so deeply in debt they're never going to own capital in the system.

By the way, J-Cal—

Jason Calacanis

Hmm.

David Sacks

—maybe the fact that all these millennials are socialists might have something to do with the fact that they're unemployable. Who the hell wants to hire a young Lenin 2.0, communist revolutionary, to be in their company? If they don't believe in capitalism, how are they going to work their way up through a capitalist system? Just saying.

Jason Calacanis

Fair enough.

That's your best point so far. Yeah, it's your best point.

David Sacks

Just saying.

Jason Calacanis

Best point so far.

David Sacks

Maybe it has something to do with the fact that all these kids are socialists and junior communist revolutionaries, rather than blaming our favorite scapegoat, AI. Okay, so anyway, that's that point.

Now, in terms of Mamdani getting elected, you have to remember, New York City is 80/20 Democrat, and he won very narrowly. It was 50.4% of the vote, so this was not some overwhelming victory. It was a narrow victory. He squeaked by.

But he did win, and it's because the base of the Democratic Party is energized by this socialist ideology. So what we saw in this election, I think, was blue cities and states getting bluer, meaning moving to the left. And that is a problem that's a little bit scary because historically in this country, politics was played between the 40-yard lines. You didn't have one of the parties being fundamentally a revolutionary party.

It does seem like the Democratic Party is gradually becoming a party of Mamdanis, Katie Porters, and genuine socialist revolutionaries. If we ultimately lose, then the country's going to be in for a big shock. But I don't think that's going to happen. This is why I think we should take very seriously the idea that, if the shutdown continues, we should end the filibuster, because while the country has empowered Republicans with all the different parts of the federal government, we have to deliver genuine results now. Otherwise, these socialists are going to take over in 3 years.

Jason Calacanis

Brad, what was your take on this? I know you were involved in a debate, a “Ban the Billionaires” debate that you did very bravely. I don't know if you can talk about it or not. I know it was at a certain university. I'll leave it at that. But if you were willing to talk about it a little bit, I think it relates and dovetails quite nicely with what happened in New York.

By the way, I think the most important statistics—and since it's my hometown, I think a lot of New Yorkers don't fall for this kind of spin that he's selling. If you look at the chart, if you could pull up the chart, Nick, of people who were born in New York as opposed to people who've been there: I was born in New York City. Thirty-eight percent of those people went for Mamdani, and the other 60% went for the other 2 candidates.

But if you were there for less than 5 years or less than 10 years, there was a 78% and 85% chance you voted for him. The people who were not born there fell for it. This guy's a charlatan. Nothing he says he's going to do is going to happen, and it's going to be total, complete, utter chaos.

I think—

Jason Calacanis

—what we're seeing, like, bigger picture, not just what's happening in New York, but what you heard when you debated kids about “Ban the Billionaires” and socialists.

Yes. Yeah, so the debate was at Stanford, sponsored by the economics department, on whether or not billionaires should be allowed to exist in America. I can't talk about what we debated in the room. It's Chatham House rules. But I will tell you this: a preponderance of people on their way in thought that they should ban billionaires in the United States.

I think it's a fundamental fight for the soul of America going on right now. This goes to the very basic premise of the American dream: Is there economic mobility in America? But I think Republicans have to get real about this. I think they have. The president ran on a Main Street agenda. He passed the Invest America Act. I think he absolutely gets what's at stake.

Seventy percent of people feel left out and left behind. They feel like the system is rigged against them. And when you have somebody like Mamdani who comes along and promises everything under the sun for free, that he can solve all of these problems, I think it's very enticing for young people who are frustrated.

David Friedberg

They also don't do any diligence because he can't deliver on any of these promises.

Right—

David Friedberg

He's not allowed to do that as mayor. And it takes the State Assembly.

Chamath and I were with Vivek Ramaswamy last night, and Vivek tweeted about this yesterday. Republicans have to talk about the issue of affordability, right? And I think they really have a good game plan around the Main Street agenda for affordability.

Remember no tax on tips. That doesn't help rich people. It helps people who are feeling left out. No tax on overtime helps people who are feeling left out, but that's getting drowned out, Jason, as you pointed out in the moment, by people who feel like their grocery bills are going up, the cost of rent is going up, et cetera. So that is going to be the struggle over the course of the next 12 months. That is the battling narrative.

And in the Democratic Party, clearly, this was a fight within the Democratic Party between Andrew Cuomo and Mamdani, and the centrist Democrats are losing to the more socialist wing of the party. On a national level, folks in Indiana, Ohio, and Wisconsin don't fall for that. They believe in the American dream. They believe in economic mobility. They may be frustrated by affordability, but they're not ready to burn down capitalism in the way that Mamdani is suggesting.

David Sacks

Those voters may not want to burn down capitalism, but that's what they're going to get, because we have a 2-party system in America. And if the Democrats go socialist, then get the trifecta, and get rid of the filibuster, that's what they're going to impose.

Jason Calacanis

I think the person who wins 2028 is the person who puts as much energy into, say, building data centers or ballrooms as they put into building affordable housing. That will be a great thing for somebody running in 2028 to champion.

Chamath Palihapitiya

Why do you guys think young women overwhelmingly supported Mamdani?

Jason Calacanis

The vibes?

David Sacks

No. Look, I think there's a lot of polling showing that between young people and old people, young people are much further left, and between women and men, women are much further left. So the most left-wing group of the electorate is young female professionals. Add in professionals versus blue-collar workers, so basically, professional class versus working class is another very important overlay.

Jason Calacanis

It might also be that there's the overlay of Cuomo being #MeToo'd as well, so maybe some women didn't feel comfortable voting for him.

David Friedberg

He's a particularly odious candidate, by the way.

Jason Calacanis

The worst candidate ever, yeah.

David Sacks

I agree that he is a weak candidate. He represents this washed-up establishment vibe. It's just not a fresh candidate at all. That's a problem.

But look, the Democratic establishment did a terrible job here. Not only was Cuomo the candidate they put forward, but they also lawfared Eric Adams. They basically wrecked his mayorship through this really weak lawfare based on airplane upgrades or whatever. And they did that because they thought they were going to be able to get a more compliant establishment figure in the role, and it completely backfired, and they ended up with Mamdani, who hates them as much as they hate Trump.

So the establishment Democratic wing of that party has completely failed. This is what we're seeing across the country: the so-called centrist Democrats, the Manchins, the Sinemas, are being driven out of the party, and all the energy is with this base that's turned socialist.

And so, Brad, this is where I get very nervous. It's true: the American heartland does not want a communist revolution, but if these guys get power, then that's what we're in for.

Jason Calacanis

Build affordable homes and this will end.

How ironic is it that San Francisco ran this experiment? We had feces in the street, we had people dying in the street, et cetera. Daniel Lurie, a centrist Democrat, gets elected in San Francisco and is now putting the city back on the right trajectory, while New York is heading in the exact opposite direction.

That's going to be a really interesting A/B test for America, because I think San Francisco is trending in the right direction. I think that, to the extent that Mamdani can do any of these things he's talking about, he's going to put New York City on a very, very bad track.

9. The Jail Debate

David Sacks

In addition to all the economic redistribution, the idea that the government's going to run grocery stores and things like that, he's also said he's going to abolish all the gifted programs in New York City schools, and he wants to close Rikers, and he believes in cashless bail.

Jason Calacanis

I think that's the plot of the second Dark Knight movie by Christopher Nolan—

It is.

Jason Calacanis

—when they open Arkham Asylum.

Yeah, yeah.

Jason Calacanis

—and the Joker and the Sandman come out.

David Sacks

Okay. So, speaking of which, at least I think New York City's got multiple jails.

Yeah.

David Sacks

In Los Angeles right now, there's only 1 county jail. I think it's called Men's Central Jail or something like that. And the Board of Supervisors for Los Angeles County has been talking about shutting that down because it has fallen into disrepair and needs to be upgraded.

They have a $50 billion budget. They haven't allocated any money toward building a new jail, and now they're talking about demolishing that jail without a replacement and just having all the inmates basically be sent to diversion or social services, which means they get an ankle bracelet and they're turned out on the street.

There are 7,000—

Chamath Palihapitiya

What?

David Sacks

—of the most hardened criminals in that jail, and half of them are severely mentally disturbed cases, like the guy who killed Brianna Kupfer. So they're talking about doing this right now. There was supposed to be a vote a couple of weeks ago.

Jason Calacanis

But the idea is that they're going to make new jails. They're going to shut down this—

David Sacks

No. No. They have not allocated any money toward building a new jail. They should have done it years ago, and they're actively discussing whether they're going to shut down this jail, whether they're going to demolish it, and send all the inmates to diversion or social services. This is a serious possibility in LA County right now.

Jason Calacanis

That's insane.

David Sacks

So they've gone—

Chamath Palihapitiya

Yeah.

David Sacks

—from defund the police to demolish the jails.

Jason Calacanis

Yeah. Mamdani's been trying to clean that up and saying he's not going to shut it down without the new one being built, but, yeah, Arkham Asylum—not a great idea.

David Sacks

It is Arkham.

Jason Calacanis

It is.

David Sacks

It's like releasing—

Jason Calacanis

It is.

David Sacks

—all the crazy people.

Jason Calacanis

Well, a lot of these people who are in jail are actually suffering from mental illness, and we should definitely have addressed that as well at the same time.

10. Building The Founder Funnel

Chamath Palihapitiya

JCal, what are you doing with Founder University? Are you trying to raise money? What are you doing over there?

Jason Calacanis

We did a deal with Senable to bring Founder University to Riyadh, and we were going to have 25 companies, mostly Saudi nationals, and teach them how to build companies and the best practices from America. We had so many applications that it went to 50. Then all these people we know were trying to get people into the program, so we wound up with 60 companies.

Chamath Palihapitiya

Oh, wow.

Jason Calacanis

I spent 3 days working with them. There were really great fintech, construction technology, and real estate companies. It was really inspiring. I saw all of our friends from Saudi there, and then—

Chamath Palihapitiya

I saw that. Did you have Tareq from HUMAIN actually talk to the founders?

Jason Calacanis

Yeah, Tareq came. He made an offer to them. He matched Google's offer for cloud computing for all of them, giving them a bunch of credits. It was really nice of him to come by. He was going to his board meeting, and he came to speak to 60 founders.

There were probably 60 teams, but maybe 90 founders in the room. It's a tiny little thing, and a lot of people came out to support it because they're trying to get more domestic startups going.

Chamath Palihapitiya

Totally. Totally.

Jason Calacanis

It was really good. Then, since that went so well, JETRO, which is the Japan External Trade Organization, called me and said, “We want to do it in Japan.” So I'm leaving here to go to Japan and announce the launch of it, and then we'll have it in 3 countries—the United States, Saudi Arabia, and Japan—which will officially start in January. I'm just going there to do a press tour about it.

Chamath Palihapitiya

Are those to feed the funnel of angel investments for you, or are you making money on this?

Jason Calacanis

Yeah. We got a fee to run the program, but it's not going to match venture investing, obviously. It just underwrites the cost of it, and then that creates the funnel. About 5% to 10% of the companies that come to Founder University wind up going to the accelerator, or we make a direct investment in them.

It's a pre-accelerator, but it lets us get to the companies before Techstars or Y Combinator or other folks even know they exist, because half of them aren't incorporated yet, right? That's the exciting part about it: We meet them when they're just 2 or 3 people working on a prototype.

Chamath Palihapitiya

It is. I've had 3 founders come in here over the course of the last 2 years. I made angel investments in 2 of them. The third I did not, unfortunately. The valuations of those 3 companies are remarkable: One is Cognition, now at $15 billion; one is Decagon, now at $4 billion; and the other one is Distyl, which just raised at $2 billion.

I mean, the rate at which these companies are scaling, for the best companies, really—

Jason Calacanis

Angel money—

Chamath Palihapitiya

—is now real money.

Jason Calacanis

It's going to be interesting to see how resilient and robust this revenue is, with churn and people sampling stuff, but it's feeling pretty good right now.

Totally.

Jason Calacanis

The cohort data for people who have had products out for 6, 7, or 8 months is starting to turn into the smile, where the cohorts go down, down, down, but then they turn back up.

Yeah.

Jason Calacanis

That means people who signed up for it forgot they were using it, and then they came back and started using it again. It's a really good sign for some of these companies.