Trump行情，还是Bessent托底？Elon Musk重返Tesla、Google的Gemini难题与中国的钍发现

Let your winners ride. Rainman said, "We open sourced it to the fans and they've just gone crazy with it." Love you.

JCAL is not here this week. He is off in Detroit at the Knicks game. Congrats to the Knicks. Here's a photo of our boy JCAL hanging out with—who's that?

Ben Stiller.

Man, what happened to Ben Stiller? That looks like Chalamet right next to him.

Timothée Chalamet.

Is that Chalamet? You know, Ben Stiller was once referred to as the Jewish Tom Cruise, but he hasn't held up like Tom Cruise. I have to say that he should have joined theology. Keep it up. You have to represent for us a little better.

Joining us this week is our boy Andrew Ross Sorkin, journalist, author extraordinaire. Andrew, welcome to the show.

Long-time listener, first-time caller. Thank you for having me.

Andrew, I'm going to give it over to the pro to do his work and take us through the docket today.

The world's greatest moderator is here today.

That's fighting words.

Shots fired.

We'll see. Hold your tongue. You can tell me when it's over.

Before we kick it off, I just want to give a quick plug. The All-In Summit, if you haven't seen the video, check it out on YouTube and Twitter going into our fourth year, the 2024 recap. We'll kind of give you a high set of highlights from last year. It's September 7th through 9th in LA. Trying to have the world's most important conversations. Andrew, I know it probably competes with one or two of your conferences, but let me just tell you, this one's way better. We're going to be blowing it out this year with parties. Allin.com/summit to apply to join us at the summit this year. Awesome. Andrew, thanks for being here. We're excited to have you.

Okay, boys, here we go. I'm curious—I want to get your take on all this stuff because I'm fascinated by what's going on in these markets this week. Let's start with the markets, because we got a rally. The question behind the rally, I think, is: Is this a Bessent put? Is this a J-Pow put? Does he have a put on this thing? I don't know what you think is going on.

We're up 3% to 7% for the week. There's been a she-loves-me, she-loves-me-not situation with what Trump has been saying about China and whether there's a deal or not a deal. China is saying nobody's talking; he says everybody's talking. What do you guys think is really happening? And then we have to get into Alphabet. We have Tesla, and we have a lot of earnings here that's also moving this around.

I think the most important piece of financial logic that we have to break is this idea that there is always a put. The put, for the folks who don't understand it, is this weird thing: When the market goes down, somebody can bellyache and cry, and somebody in the government will say, “Okay, we'll buy your shitty securities at your bad prices.” So that's what the put is.

Typically, what's happened is you can cry to the White House, and they'll call the Federal Reserve and it will buy it. That's the Fed put. In some very rare cases, when things get really calamitous, like in 2008, Treasury will set up a program and start to buy things. That's what was called TARP back in the GFC.

So the question is, is there a put here? My honest answer is no. The simple reason is that, if you look at where we are, the stock market is back to where we were in May or August of last year. If you had said that we would have upended 50 years of economic policy and the markets would only be off 500 or 600 basis points, I would have been shocked. So there's no reason for a put.

Maybe the more interesting question is, who's asking for the put? The ones that have suffered from the volatility, and a lot of those are the active market traders. If you trace the feedback, who's been the most negative? Ackman's been negative. Ken Griffin recently turned negative. That probably comes from the fact that they're not monetizing the vol the way that they used to monetize the vol, more than the market has spiraled down to such a degree that the world is coming to an end, because it isn't.

Okay, can I build on that? There was a tweet two weeks ago, when all this started, by Nick Carter that I retweeted because I thought it was really interesting. It was: “If you think a market sell-off delegitimizes Trump's presidency, you're going to give him unconditional credit when it rallies, right? Because those are the rules now.”

And that's exactly what's happened. The market has rallied. It's up 7% this week, but the media doesn't want to give Trump any credit for it. So it's created this narrative of a Bessent put. In other words, Bessent is entirely responsible for it.

Look, Scott is a very smart man. He understands markets and the bond markets, and he is a reassuring voice to those markets. I think he's doing a great job. But I think part of what's going on with this Bessent-put narrative is that no one ever wants to give Trump or his administration as a whole credit.

So what they do on any particular week is valorize a particular member of the administration that they can then say, “Well, this is the person who's really responsible for this, not Trump,” because they don't want to give Trump that credit. Next week, they'll be tearing Scott down, and there will be some other member of the administration they'll be trying to valorize to basically prevent the administration as a whole from getting credit for doing something good.

Again, I think Scott's doing a great job making these deals, and he's saying a lot of correct and reassuring things to the markets. But if the media is going to tear Trump down every time the market goes down, you have to give him credit when the market goes back up. I think there's just a general reluctance to do that.

Where do you land, Dave?

I think the “is he crazy enough to let it all fall apart?” signal got wiped out this week. That's really what it was. It was the fact that everyone realized that Bessent and Trump are not going to let trade come to a standstill. They're not going to let the whole global economy grind to a halt. They're going to have to do something.

I think the read is that they're indicating they're going to get deals done, which means that their intention is to make sure that the market doesn't tank. Roughly three weeks ago, when we started talking about this, I think I made the point—and Chamath made the point—that part of the art of the deal here, so to speak, is an opening salvo, which is bold, and some people would say maybe it's too extreme, but it's basically an opening bid to shift the conversation completely. That would create tremendous leverage to get deals done. Isn't that exactly what's happened? Isn't that the way it's played out?

Yeah. I remember there was a guy from Harvard who used to teach a class in negotiation. He put out all these videos, and what he always said was, “Set your anchor point as far away as you possibly can if you want to get maximum leverage in a negotiation.”

I don't know what's farther away than a 145% tariff. So it's clearly a point for, as Sacks said, negotiating to a deal that seems like one that would have been impossible or implausible if you started from a 0% tariff beginning point in negotiating.

There was some percentage of the market, or some probability, that people were assigning to the fact that this administration could be crazy enough to tank the global economy by leaving tariffs really high and letting this kind of run. I think that's being taken out of the market, right?

But now that it's out of the market, that means there's less leverage. There's less leverage in the conversation. And if the goal is to—

No, that's not true. You're saying a critical thing, and that is not true.

Well, hold on. If the goal is to effectively pressure China, which I think is the ultimate goal, and the bond market has spoken, and the Trump and Bessent administration has said, “Okay, we understand where the floor is here and what we've got to do,” I assume that if you're Mark Carney, if you're the EU, or if you're India or Japan, you say, “I'll make a deal with you.”

But if you thought I was giving you the world before, no, no, no, no. You need me as much as I need you, and you seemingly need me even more now.

So the part where I agree with you is that last part. I think too many times you look at the stock market and think that that's reflective of what's actually happening, and it's not. The stock market, I think the best way to think about it is, it's an estimation of what the future would look like at a rate of return. You're superimposing a view on reality.

What is actual reality right now? If the United States was running a negative $2 billion-a-day trade deficit, right? If our current-account balance was minus $2 billion on March 31, the day before Liberation Day, the real question is, what is that number mathematically after April 1? Because you're ratcheting up tariffs and implementing a scheme that should take that negative $2 billion that's going out the door of America's shores and flip it positive. How positive? We don't know. But I'm sure that the government knows, and I'm sure the White House knows what that number is. That's important fact number one.

Separately, the country that matters the most on the other side of that, which was very positive—let's say China was plus $3 billion or $4 billion a day—where did they go? I suspect that it's probably breakeven or possibly even negative.

So if you think about what the new reality is today, April 25, as we tape the show, we've gone from minus $2 billion to maybe plus $1 billion.

And China's gone from plus $4 billion a day to maybe zero or minus $1 billion. I think that's what instills the leverage that is required to get the deal done. Then we can all debate whether the forward P/E of the S&P should be 16 times or 22 times. I don't think that matters, but the actual cash flows have materially changed.

By the way, I think long term America's got leverage here. Everybody who wants to invest is ultimately going to want to invest in America, including China. There's no question that we have the advantage on a long-term basis. The question is, do we have the patience?

Well, if you go from minus 2 to plus, given our democracy, given our bond market, given everything that goes on, we're such an ADD society.

Andrew, I think that's a very good point. What I would say to that is, then, we need to communicate better. It needs to be a single point that's repeated by everybody in the Cabinet. What they should probably focus on is, "Hey, guys, we just flipped cash flow. We went from negative cash flow to positive cash flow. Here are the actual numbers." Don't deviate from that, because then I think Americans would say, "Hey, this thing is working. Let's let it play out."

To your point, hey, Sacks, here's my question for you. We hear the president say he's talking to President Xi. We hear Scott Bessent say that he's in talks with China, or that talks are going to begin. And then you hear China say, "No, no, no. We're not talking. There's no talks happening. We don't know what you're talking about. Stop saying these things." What are we all supposed to believe?

Well, look, first of all, let me just say that when I'm on this show and I'm not talking about AI or crypto, I'm just a civilian, right? I'm not a spokesman for the administration on anything but the 2 issues that I work on. So I don't want to come across as someone who has special knowledge, because I'm just not part of the trade conversation. I can't answer your question directly.

But what I would say is that, with respect to the China relationship, I think what's happened over the past few weeks has been a very important stress test. It's basically flagged some serious weaknesses or dependencies that have evolved in this relationship over the past few decades. I flagged this in a previous episode that we did with Larry Summers, where I said that, 25 years ago, it was a huge mistake to walk China into the WTO.

I think, just to build on that thought, one of the problems with walking them into the WTO is that they got what's called developing-nation status. You can either be a developing nation or a developed nation in the framework of the WTO. Maybe there was something legitimate to that in 1978, when Deng Xiaoping began his reforms. The average Chinese person was making $2 a day. By the year 2000, it was much more questionable, and it certainly makes no sense in the year 2025. Still to this day, China gets developing-nation status.

Now, what is the benefit to China of that? They have a whole different set of rules. They're allowed to have tariffs. They're allowed to subsidize their industries. They have all sorts of different time frames for doing things. They're allowed to do things that the US simply can't do.

How have they leveraged that? They have identified strategically certain industries that are choke points in the global supply chain, and they have taken them over. The best example of this is rare earths. There were some good stories in The New York Times about this over the past couple of weeks, which I think were largely accurate.

China identified this as a critical industry. The rare-earth ore itself is distributed all over the world. China has some advantages there, but they're not huge. It's the processing of the rare earths that's very expensive and very complicated, and they decided to dominate that industry. They're responsible today for over 90% of the processing of rare earths.

The next step in the supply chain is that those rare earths get cast into rare-earth magnets, which are a critical component in pretty much every electric motor. They're a critical component of the automotive industry, but not just cars—lots of different products. I think China makes something like over 90% of the cast rare-earth magnets.

By the way, this issue still has not been resolved. China has now cut off the United States. As part of this trade negotiation, we're going to have to resolve that issue, and I trust that it will be. What you've heard from the president and the Treasury secretary over the past week is that they've indicated a desire to engage in bilateral negotiations with China and to essentially de-escalate this trade war.

I think it's been very useful, again, as a stress test to reveal the critical dependencies on China that we've exposed. We never should have let this happen. From a national-security standpoint, we worshiped at the altar of this free-trade god to the point where we became dependent on China for these critical components in our supply chain.

I think that was a catastrophic mistake, and I think it's exposed these dependencies we've created on a nation that is not our ally and that we can't count on. It's a country of concern. I ultimately think that we're going to need to learn from this experience and very rapidly make some major corrections here.

Andrew, let me just highlight what I think is one of the not-talked-about points about the discussions that are underway. There's a lot of conversation about tariffs and trade deficits, but very little about regulatory parity. I think this is really critical for these trade negotiations to actually resolve to a positive outcome for American businesses and American enterprises, because there is not parity in how American businesses can do their work overseas relative to how foreign companies can do work in the US.

If you're an Indian company and you want to sell something in the US, you set up an LLC, you set up a bank account, you set up a store, you take a lease, and you sell your product. There aren't a lot of hoops and challenges to that business operating in the United States.

I'll tell you a story. In 2005 or 2006, Monsanto, which was a seed company that doesn't exist anymore, launched cottonseed in India for cotton farmers. It basically had a trait in it that would prevent worms from eating the cotton. Indian farmers were paying 450 rupees an acre for cottonseed before Monsanto put this cottonseed out. The yields went through the roof, the farmers made 5,000 rupees per acre of incremental profit, and Monsanto charged 1,500 or 1,400 rupees per acre for this seed.

It was more expensive seed, but it had a 5,000-rupee profit improvement, and it took off. It took the whole market. Ninety percent of farmers bought the seed, not because they had to, but because it was better.

What ended up happening was the Indian government stepped in and said, "You've got to lower prices. You've got to go back to the price of 450, because that's what's fair for farmers." The farmers were like, "We're doing great. We don't need to have this kind of price control." But the government put price controls in and ultimately sued the company, went to court, stole their IP, and they left the country. This was a company that had invested well over $1 billion in developing and launching this product.

The same is true, by the way, in pharmaceutical companies, software companies, and hardware companies. If you want to go do business in China today, you have to set up a 51% joint venture where some local company owns 51% of your equity and your IP.

The stories around the world start to paint a picture of why American businesses find it so hard to develop international markets and sell into those countries, when in the US we make it so easy for companies based in foreign countries to come and sell in America. That's a big part of where the trade imbalances arise from.

It's not just because Vietnamese people can't afford expensive American goods. It's because it's so much more difficult, from a regulatory standpoint, to do work in these countries. They can come in and take your IP. I think this is a really important part of the trade discussions and negotiations that a lot of people miss, and that American businesses will see massive revenue growth and massive market adoption if we can get regulatory parity in some of these key trade markets.

Okay, let me ask you a different question. Chamath raised the issue of Ken Griffin, who made some comments this week about the brand that is America and effectively said there was sort of a "sell America" situation going on.

My question about the brand that is America is whether you think, after these tariffs—whatever deals get done—you can put the toothpaste back in the tube. Can you get the trust back? Is there still going to be some kind of toothpaste on the rim of the tube that you can't stick back in, and that gets all nasty and crusty? To me, that's a fundamental question.

I want to read to you what Ken had to say. I think Ken was probably a bit of a reluctant Trumper in the second term, but nonetheless, he's been outspoken on a couple of things. He says, "If you think of your behavior as a consumer, how many times do you buy a product with a brand on it because you trust that brand? In the financial markets, no brand compares to the brand of US Treasuries, the strength of the US dollar, the strength and creditworthiness of US Treasuries. No brand comes close. We put that brand at risk." What say you?

Well, look, I think there's a fundamental question here about whether you think there's a problem with the status quo or not. Ken Griffin is one of the biggest winners in our economy under the current status quo, and I don't think he sees a big problem that needs to get fixed.

But I think, like we just talked about, there are some big problems. Friedberg just laid out the way in which trade with China is nonreciprocal: Our companies cannot participate in their markets the same way that they can participate in ours. But it's worse than that. As I talked about, these WTO rules gave China the opportunity to strategically annihilate our core industries that are critical in the supply chain.

There's one other piece of this that I think is really important, which is the race to the bottom. If you're an American company and you are still producing in America, and your competitor is able to go to China and undercut you, obviously you have to do the same thing because those are just the rules of the game. We've had this race to the bottom where, if you're an American company operating under these rules, you have no choice but to export your manufacturing and your supply chains.

We've had these rules that, under the status quo, I would say are not free trade. It's unfair trade. It has all these perverse consequences. It's bad for middle America and this manufacturing belt of the country, and it's created massive strategic dependencies on an American adversary.

I'll say 3 things. Let me just say 3 things quickly. Number 1, I think Ken Griffin is massively long Pax Americana. He's built a $42 billion empire. He owns $1 billion of American real estate, so I don't see him investing in Ecuador anytime soon relative to the opportunity set in the United States.

I think that's what speaks to the second point, which is that investing is always a relative exercise. On Wednesday, I had a meeting with a Brazilian multibillionaire, a very, very, very successful guy, and we were talking about the tariffs and the investing posture. He's like, "There's still no other country other than America that I can really put my money to work in that makes any sense."

I pushed him: What about China? What about India? In all of these markets, there are, as Friedberg mentioned, vagaries in China. There are different problems in India. There are different problems in these other markets. Europe is roughly uninvestable. So, the more generalized answer is that, on a relative basis, the United States is still the shining beacon on a hill. There is no investable alternative for scaled capital except the U.S.

The last thing I want to say is just to build on what Sacks said, Andrew. The thing that we keep forgetting is: When economic leverage spills over to political leverage, what happens to the United States? Specifically, what happened this week was, on the rare earth side, not only did China constrain rare earths in the supply chain, what they then did was tell South Korea what they could do with the rare earths that they gave them.

Now, all of a sudden, you have this geopolitical spillover where it's not just a country constraining supply. It's a country now dictating the political and trade practices of another country. So let me ask you a question: How do you think America is in the best position to make its own decisions?

For example, let's say that there's a China-Taiwan situation and we need to decide one way or the other. I'm not saying which way is right or wrong, but for the purposes of this thought exercise, China and Taiwan get into some spat. We have to take a side. China says, "We're going to constrain all the pharmaceutical APIs in the supply chain. We're going to constrain all the battery cathodes in the supply chain. And we're going to constrain all of the rare earths to anyone who doesn't take our side."

How is America in a position to actually decide for itself what is in our best interests if that's looming over us? I think that this is the big problem that this has exposed. This rush to globalism, the race to the bottom, has actually created a level of dependence that doesn't allow you to do what's in your best interest.

Real quick, I agree with you in so many ways. The question that I have is not what the ultimate goal is; it's how you get there. Is there a way to do this with a velvet glove, if you will? Maybe there's no way to do it in a better way.

No, but what's to say this isn't the velvet glove? The stock market is down 6%. It's not down 60%. What I think is so interesting, Andrew, is the way that Trump has already shifted the conversation.

The truth of the matter is that, before Liberation Day on April 2nd, which was 3 weeks ago, no one was talking about the unfair trade practices. No one was talking about the dependencies on rare earths. No one was talking about the race to the bottom. Trump has shifted the conversation 100%.

When Jared Kushner was on the show, he talked about how the Trumpian approach is "controversy elevates message." We said at the beginning that, by having this Liberation Day, by planting this flag in the ground, Trump was creating leverage to then have these negotiations, and he completely shifted the conversation.

Now, where I will agree with you is that the administration has to stick the landing. Bessent—and it's not just Bessent, but also Lutnick—these are all smart, talented people. They do have to then negotiate these deals, and we have to basically stick the landing. But I think the fact that you're saying that you don't disagree with where Trump is trying to get to, but it's mostly just tactical, is a huge shift in the conversation.

It may raise the issue the way you're describing, but at some level it's also undermining a semblance of trust. I just want to relay 2 stories. One is, I've been talking to a whole number of multinational CEOs who do business in China. I'm talking about the McDonald's of the world, the Starbucks of the world, the Nikes of the world.

So many of them now talk about not just how consumers aren't necessarily going into their stores the way they used to, but also about the American dream and the American brand. If you were a young kid and you got a job at Starbucks, you used to go home to your family in China and say—and your family would go, "Wow, that's amazing." They're not doing that right now.

So many of these companies are even trying to reestablish how they market and advertise their companies, almost as if they're a local business, not an American brand. I don't know what that ultimately means.

I was also talking to a Chinese CEO who made what I thought was just a fascinating comment to me, which was, "We're cool with the U.S. being the leader. We were always cool with that. We're okay with that. We don't love it, but we're okay with it. It gets a lot more complicated when you have to be the winner."

I think that's an interesting sort of construct, this idea that we're first and we have to win. I know Trump is telling citizens in the U.S. that we're going to win, but he's also telling the world that. I think that Chinese CEO should have written that memo to Xi Jinping and seen if he agreed, because I think he would say, "No, we are the winner."

Look, you've got to remember: In 2003, 3 years after China was admitted to the WTO, Hu Jintao became president. He had this incredible speech, and in it what he said is, "There are these 10 boxes, and I've written inside these 10 boxes all the critical industries that, in 25 to 30 years from now, will dictate supremacy in the world. We are going to create national champions," he said, "in those 10 areas."

What's interesting about Xi is that, despite the ideological differences from the person he followed, he followed through with it. This is a group of people. We can say what the CEOs are going to say, but at the end of the day, the CEO of China, Inc. is one man, and he and his coterie have been pretty resolute.

Even when he inherited a political ideology that he didn't completely agree with, he followed through with it. These guys in these 10 areas now own the critical inputs that make the world go around. I think it's pretty reasonable to do a risk assessment and say, "How do you stand up to that and actually have your own point of view if you don't have your own method of having access to those things?"

I think that is the most important question. Everything else about brand and all of this other stuff comes after this question. This is the critical question, because if they can tell you what to think, that is not winning.

I would just add to that: You hear this criticism of Trump a lot, which is, "Okay, he's right on the issue, or his instincts are correct, but there should be a nicer way of doing it." I would give that story more credence if the people making that criticism had actually been advocating for a change in the status quo prior to Trump. They weren't.

We had this bipartisan globalist consensus that pretty much all trade practices with China, no matter how unfair they got, were basically good for the United States. No one was really doing anything about it until President Trump made it this issue and made the unfair trade practices conspicuous.

If somebody else had been willing to take that on, then I maybe would give credence to this idea that there's some non-Trumpian way of doing this. But I just don't think that's the way our political system works. I think that the way our political system works is that you have to completely shift the conversation first to create a new consensus, and then you work out the deals. That process can be a little bit disruptive.

Okay, one tiny follow-up. I don't know if you guys saw the news this morning, but there was a report out that Apple is going to move the manufacturing of iPhones that are exported to or imported into the U.S. They're going to manufacture all that stuff in India, not in China anymore. So they're going to move all the manufacturing to India.

See you in 2035. They say they can do it, supposedly, in 18 months, by the end of 2026. Bet. You say no way. Bet.

Well, I think if this is geopolitically smart, yeah, they should do it. But I'm just saying that's—it's very smart. It's not going to get done in 18 months. I don't know how long it's going to take.

But look, I think one of the big lessons here over the past 25 years is that you have security first. Security has to be worked out first, then you have trade. If you build your whole supply chain and all your trading relationships with countries that are fundamentally adversarial to your interests—and we can get into that part of the U.S.-China relationship—but I think most people would say that we're in some version of a new Cold War with China.

Obviously, you're going to have to revise your trade relationships because we can't be dependent on an adversary for core components that then go into our military, for example. Security always has to come first, then you work out trade. I think that India is fundamentally extremely aligned with us on security because they view their biggest potential threat and adversary as being China, and that is basically the way that the U.S. sees the world as well.

I do think that the U.S.-India relationship is very aligned, and I think, therefore, the investments that get made and the trade relationships that get forged will be very stable over the next couple of decades. Whereas, if you make a big investment in a country that could be your adversary, you should expect that that relationship could get disrupted.

I don't think the Apple deal is from a standing start. My understanding is that Foxconn and others have been working with Apple to actually stand up manufacturing capacity in India for some time now, and I do believe that they probably have 1 or 2 model runs already active in the country.

I don't think that this is necessarily, “From a standing start, let's go recruit the people, figure out the processes, find the builders, and stand up the facilities.” There's probably a model system already running that's now about replication and scale-up, which is why they have confidence in the stated goal.

It's not often that Tim Cook will stand up and say, “I'm going to do something in 18 months,” and then be delayed by years. He's not Elon. He's going to come out and be very clear. He's always been very clear. My understanding is that this has been thought about for some time.

And you, I think you've been to a lot of conferences where, for 3 years now, you've been hearing people banging the table saying, “Get everything out of China as quickly as you can.”

Yeah. Tim Cook's heard that message. He's been thoughtful about it. He's been planning for this day. I don't think he wanted to be declarative about this until the day came. Well, the day is here. How do we feel that it's gone to India? It's not coming to America. I think we knew it wasn't coming to America.

Well, India has 1 massive advantage over China, which is that it has 1/5 the labor cost of China, which has 1/5 the labor cost of America. India is a natural place for a lot of this next-generation labor to end up. Plus, they have a very educated workforce. Plus, they have a young workforce.

In many ways, you could say that it's China circa 2003. You have to remember, China from 2006, I think, to 2012, its GDP grew at about 10% a year. Can you imagine what that must have been like when a country that big is growing that fast? India is going to have that moment over the next 20 or 30 years.

The only reason I scoffed at what you said—“You invest in India, would you?”—is that I have a large rare-earth mine in India. It's one of the largest supplies of rare earths outside the United States. We negotiated that deal with the PMO, and they gave us a great deal to compete with the Chinese.

Specialty-chemicals processing is on the ground there. The idea is then to import all of those specialty chemicals into the United States and do all the—

I like the macro. I just—there's something about China where there's a lot of accounting fraud still, and you can't really trust a lot of the numbers. But if that all gets cleaned up, the macro is all right. It's exactly where it should be for—

You mean India?

India.

Yeah. I just think the structural realities make it a good place to invest. I think that India and the U.S. are going to be very aligned for a long time.

Agreed. Again, their main security threat is China, and our main security threat is China. That's just not going to change for probably the entire 21st century.

A couple of weeks ago, Sacks and I sat down with the Indian ambassador, and it's very clear they're fundamentally aligned with the United States. I think exactly what Sacks said is true. Everybody knows who needs to cooperate in order to have a reset of global trade policy and political leverage.

I don't know. I'm pretty pro-India. It's a very difficult place to invest.

I agree with you, Sorcin. I've never made money there. I tried my hand at tech investing over a decade ago. I may have returned 50 cents on the dollar.

Why? So then what went wrong? What's the macro reason?

The market was far too immature. My mental model was about what American success looked like, and when I tried to impose that on the Indian market, it all just blew up.

A lot of the Indian entrepreneurs raised money not from Indian VCs, but from American VCs. You can't tell an American VC an idea that makes no sense to them because their frame of reference is the United States.

What turned out to work were a couple of very large bets that were copies of American businesses. The Indian Amazon worked. The Indian PayPal kind of worked. The Indian DoorDash kind of worked. But everything else didn't work.

So why? Is it that the consumer market's not there, or—

No, because markets aren't there. We have these mental biases that will never allow us to believe that the way they would construct the business would work.

For example, they're like, “Here's a business where we're going to buy food from the local farmer, then we're going to put it in a warehouse, then we're going to drive it around on rickshaws and deliver it to people.” You think, “This business can never work.” Well, actually, that business turned out to be a ginormous success. Nobody in the United States would have ever funded that.

What happened 4 years ago, though, was that after a bunch of meetings, I pivoted to these other areas because when you talk to the Ambanis and when you talk to the Adanis, they're like, “Dude, stop beating your head against this wall. Build infrastructure. It just works.” And it was true. They've been right.

Okay, 1 geopolitical question: How do you feel about India buying as much oil as it does from Russia? I know we'll get into the whole Ukraine thing later.

Smart. Smart.

Look, India is a country that has 2 great powers in its neighborhood: China and Russia. It has a contested border with China. They've had border skirmishes. When you have a country that is much more powerful, that's actually a security threat to you, you then seek good relations, even possibly an alliance, with the other great power in the neighborhood.

That has been India's philosophy. They've had good relations with Russia for a long time, even going back to the Cold War. That is what makes sense from their security standpoint.

By the way, the exact same thing is true of the United States. We just haven't realized the strategic reality. China is the peer competitor. John Mearsheimer made this point at our All-In Summit. China is the peer competitor. It's the only country in the world that is a peer of the United States. It's the only country that's really capable of threatening our security.

What you want to do if you're in a head-to-head competition where there are 2 superpowers is make an alliance with that number 3. What we've done is push Russia into the arms of China.

What we ought to have been doing is a reverse Kissinger. We did this during the Cold War, by the way. We had the Soviet Union and we had China. In that case, the Soviet Union was the big threat to American security. So what did we do? We had Nixon and Kissinger go to China and make a deal with Mao.

China was just as communist as the Soviet Union. Mao had blood on his hands to a degree that was much greater than someone like Putin has. And yet we were willing to shake hands with him and make an alliance because that was realpolitik.

I think, in a similar way, this is what our strategy should have been with Russia for the last 2 decades. Instead, we've been foolishly pushing them into the arms of China.

Well, what's your read? I mean, what's your opinion on this? I think India is a tough look. India's a tough place to invest. Everybody I know who's been talking about India for the last 15 or 20 years has been talking it up as the next great utopia. It may very well be. It probably will be. It's just a question of when.

I think you have to be specific about the end market. There are people who have made a ton of money there, but it's not the way classical Western economics has played out by any means.

Andrew, what do you think about this sort of balance-of-powers logic 101?

I get the idea that your enemy's enemy is your friend, or whatever it is. I get where you're going with this power logic. I just think that, given all that Russia has done, it's hard to look at Russia and China. Maybe Russia, on the margins, on a relative basis, is better than China in certain ways, but in other ways it's not.

It's very complicated for me to look at it and go, “Okay, who's more evil?” Is China more evil, or is Russia more evil? I think there would be a big debate about that in the world.

Well, I think it's unnecessarily judgmental. I think it's dangerous to look at foreign policy from a position of extreme moralism because ultimately, the purpose of our foreign policy is to ensure American security, and we can't rectify every injustice in the world.

Over the past 25 years, we've become hyperinterventionist in an attempt to do that, right? We keep saying that we've gotten involved in all these places because we want to promote our values and spread democracy. And, by the way, we did the opposite. I mean, you look at the forever wars in the Middle East. We had interventions and occupations in Iraq, Afghanistan, Libya, and Syria. How did all those things work out? They did not promote our values or spread democracy.

So, my point is just that I don't think we should look at foreign policy from a position of extreme moralism. I prefer to think of neither of those countries as “evil.” I would prefer to think of them as countries that have their own interests.

I think that, with respect to China, the real issue is that it is a revisionist great power. It has made clear that it wants to basically annex Taiwan. It wants to turn the South China Sea into effectively a Chinese lake. It does not respect international waters, and it has territorial disputes with Japan. If we don't play a role in containing China, then China will revise the balance of power in East Asia in a major way, and I think that would have profound consequences for the United States.

China will just throw you off a roof one day, and you won't even know it. I mean, that's the difference. And I don't know if that's a moral argument. I'm not worried about the United States being thrown off a roof by Russia. That's the level at which we should think about these things.

If you look at Russia's behavior, Russia actually is not a revisionist great power. What it wants is just security on its borders. The revisionist power in Europe over the past 25 years was actually the United States, because we pushed NATO right up to Russia's borders despite the promises that were made to Gorbachev in the '90s.

We're the ones who've revised the balance of power in a way that was profoundly threatening to the Russians. And I think it would be far better from the American standpoint to just work out a security architecture for Europe with Russia that gives the Russians the security they crave on the western border and gives Europe the security it craves. I think if we had had that mentality, as opposed to this highly moralistic mentality where we basically want to spread American-style democracy everywhere, we would be in a much better position.

You don't think that American-style democracy, though, has helped the world in some way? I'm not saying we did it right everywhere. I think we've lost a lot of wars that we should not have attempted.

But if it could be done, you're right. But the problem is, the failure rate is too high. Look at what happened in Afghanistan, Syria, and Iraq. As we've gone in to make regime changes, we actually put those people in a worse condition than they were before. They don't end up with American democracy.

After $1 trillion in Afghanistan, the Taliban is back in charge, doing awful, brutal things, right?

$2 trillion.

And I would be a proponent of—by the way, this was also part of the promise of the open internet. From a fundamental point of view, the regulatory regime we had in the United States with the internet was driven by this notion that freedom of information would actually allow democratic proliferation.

We were talking about the Arab Spring, as you'll recall. That was going to be this big moment, but at the end of the day, it turned out that didn't work either. There's something much more complex in the history of a people and a geography—the deep ties that those people have. This idea that turning on a screen or showing up with a tank is going to flip the switch to democracy turns out not to be true.

Look, just to be clear, I thank my lucky stars that I'm an American, and we do have the type of government we have here. My parents immigrated to the United States in 1977 for political reasons because we have the type of government we have here. So, look, I'm a strong believer in American-style democracy. The problem is, our efforts to promote it or impose that style of government through regime-change operations simply have not worked. I mean, this should be one of the big learnings of American foreign policy over the past 30 years.

It's better to model it right.

Yeah. So, how do you model it?

I think you model it by being the shining city on the hill, by setting a good example, and by focusing on our own country and making America the best it can be so that people all over the world look to us and say, “Oh, we want to do that.”

I actually think that what I'm describing is American soft power. If you go back to the 1980s, we had that soft power. I mean, people in places like Russia loved buying American blue jeans and listening to American music. We had that soft power. We were inspiring to the world.

But it all changed, starting in the '90s, and then really in the 2000s, when we started going into these countries with our militaries and occupying all these countries. We soured them on America. We made them hate us. This is the problem with hyperinterventionism. What we should be doing is setting an example, not trying to occupy these countries.

Let me read something to you. This is from this week. This is Trump condemning the strikes going on in Kyiv. This is what he said: “I'm not happy with the Russian strikes in Kyiv. Not necessary. Very bad timing. Vladimir, stop. 5,000 soldiers a week are dying. Let's get the peace deal done.”

How is this deal going to get done? I should also say Secretary of State Marco Rubio reportedly skipped a meeting on ceasefire talks on Wednesday after Ukraine refused to recognize Russia's annexation of Crimea.

Well, the reason why Rubio canceled that meeting is because the Americans put forward a peace proposal and Zelenskyy dismissed it out of hand. Before the Russians could do anything, Zelenskyy dismissed it. Recognizing Russian ownership of Crimea, if you will, should be the easiest point for Zelenskyy to concede, and he dismissed it out of hand.

Now, why do I say that? The Russians annexed Crimea in 2014, so this was not something that happened in 2022. It happened over 10 years ago. And if you look at what the people of Crimea want—and there's been a lot of Western polling on this, as well as man-on-the-street interviews by NBC—the people of Crimea, by something like over 80%, say they want to be part of Russia. They're ethnic Russians.

I'd say the final point on this is that Zelenskyy and Ukraine tried to get Crimea back in the summer of 2023 with a summer counteroffensive. Remember, the whole point of that summer counteroffensive was to sever the land bridge to Crimea, destroy the Kerch Bridge, isolate Crimea, and essentially lay siege to it. This was the whole point of that, and they got nowhere with it. They didn't even break through the first Surovikin line.

So, my point is that it's militarily impossible to retake Crimea. The people of Crimea don't want Ukraine to retake it. They want to be part of Russia. And yet Zelenskyy cannot give this up.

Okay, so how do we get to yes? This was supposed to be done on day one, and here we are.

The basic problem is that Zelenskyy doesn't want to make a deal. I mean, he has rebuffed the Americans on Crimea, which should be the easiest point to give on. He's been completely unrealistic about their prospects on the battlefield.

He was insulting to the White House when he visited the White House in his T-shirt or whatever. He was dressed inappropriately. And that meeting was one of the worst tests for America, I think.

He was insulting to the president and especially to the vice president. He murmured a curse word under his breath to the vice president. You remember that? I mean, it was really quite something.

And remember, he's biting the hand that feeds him. He has an obligation, I think, to be respectful to his patrons in a way that we may not have to be to him. And he has demonstrated that he's unrealistic in every possible way. He's unrealistic about Ukrainian prospects on the battlefield, about retaking Crimea, and about which hand feeds him.

So, my view on this is quite simple: If Zelenskyy won't see reality, let him go find new patrons. Let Starmer and Macron support him. They say they will. They say they have the ability to do this, but they can't. My point is, Zelenskyy has made his bed. Let him sleep in it. He refuses to listen to the Americans, and he refuses to recognize that this administration even has leverage over him.

And you don't think that ultimately, if Putin takes Crimea plus whatever, he's going after Poland next? I mean, that's the domino theory that becomes the complicated part.

No, I don't think so. I don't think there's any evidence that he wants Poland. I don't even think he wants the part of Ukraine that's chock-full of Ukrainian ultra-nationalists, because you would have a massive insurgency there. He doesn't want a Gaza-type situation.

And, by the way, just another point on Crimea: Where is the insurgency? If Crimea really wanted to go with Ukraine as opposed to Russia, why has there been no uprising there over the past 10 years?

In any event, my point is there's no evidence that Putin wants Poland. Poland is part of NATO. That would start World War II. I think Putin has shown that he's a calm, rational decision-maker, and I don't think he wants World War III.

Now, the Russians have said what their terms are. They said it last year, in this root-cause speech they gave, I think, in June of last year. Basically, what it comes down to is they want Istanbul-plus.

There was a deal to end this war in the first month: the Istanbul deal. It required Ukraine to basically sever its security relationship with the West and recognize that it would not become part of NATO, and it had to give up Crimea. Basically, that was the deal back then.

What Russia has said is that that's still the deal, plus realities on the ground. In other words, the Russians have lost—I don't know—100,000-plus, hundreds of thousands of soldiers taking this territory in the east, and obviously they're not going to give it up.

So I think that the Ukrainians could have had a better deal in the first month of the war. The Biden administration rejected that. Now there's a different deal. It's Istanbul plus, and I think that the American proposal is pretty close to that. So I think this administration recognizes the wisdom and just reality of that deal, but the Ukrainians aren't.

Look, my view is simple. If the Ukrainians are unwilling to take the deal that's on the table and they want to find new backers with the Europeans, let them. But why should we continue to support this?

Okay, let's pivot. Let's talk markets and specifically talk about a couple of stocks. David, you disagree, because I want to hear your disagreement.

No, no, I could, but I think we're going to run out of time.

Okay. I want to talk Alphabet. So everybody thought the blue-links economy was dead. OpenAI, Perplexity—all of it was going to take the search market. They come out with earnings, and Sundar does way better than everybody expected. Revenue was $90.2 billion, up 12%. Net income was up almost 50% year over year. You've got the cloud still ripping, you've got a big buyback program, and he starts talking about Waymo on the call for the first time in a long time, if ever. The stock's up about 5% on the news. What do we think?

What an incredible business. My gosh, there's a lot of resiliency in Alphabet that I don't think people have given the company credit for. First of all, I'll just highlight that with the dividends they're paying plus the share buyback, you're basically getting a 4% to 5% yield on the stock. The dividends are about $10 billion a year, and they're announcing $70 billion in buybacks on $2 trillion in market cap. So you're making 4% a year just by owning the stock through the dividends and buybacks.

Then, if you look at the revenue—$90 billion, $77 billion in services, $12 billion in cloud—about half of that $90 billion was ads on search. A lot of people have been talking about the decline in search due to AI. That's the end of Alphabet. Here, you can see the breakdown.

What this shows is that there's such significant growth: $9 billion in YouTube, $7.5 billion in non-Google ads, $10 billion in subscriptions with 270 million subscribers across subscription businesses, and $12 billion in cloud revenue, which is growing 30% year over year. There's a lot of resiliency, even if you take out search.

And so the question is, what's the downside and what's the upside? Well, the downside is, what's the worst that could happen to search? What if they lose half of the market to ChatGPT? If they lost half the market to ChatGPT and the rest of the business keeps growing the way it is, you get back to parity pretty quickly.

What's the upside? Well, the upside is they're able to use AI to actually reinvent search, retain the audience, and create an entirely new search experience that's chat-based or has some chat integration into search. So it seems to me, trading at 18 times free cash flow, which is where this thing's at right now, plus the 4% yield on dividends and buybacks, plus the seemingly pretty significant gap between the downside and the upside here, it's just such a powerful business.

It just seems like it highlighted that this is not a business that's in decline and is about to die, as everyone's been proclaiming. There's enough growth engines here, there's enough resilience, and the price is so good. I feel quite good about it. I think there's $100 billion of hidden value called Waymo in this business—250,000 rides a week now, growing like crazy, and they're only in about 4 cities. It's crazy. I think the market's giving them zero for that right now. Zero.

I think the most interesting stat that I saw—and they own% of SpaceX by the way which is crazy—the most interesting stat in the earnings was that they said they had 270 million total paid subscriptions across YouTube and Google One.

My initial thought was, when are you just going to stick Gemini as the most obvious interface in front of it? I think these companies now need to confront that this OpenAI behemoth is growing at a rate that was probably a hindrance before and now is a real strategic risk. So if you have 270 million paid subscribers, come on, guys, put Gemini right in front of them. That was my first takeaway.

Again, they're doing it from a position of strength, but they need to get going because they're probably hemming and hawing, and there are probably all kinds of people—all kinds of cooks—in that kitchen. But I've heard a lot of stories about this question. I think Sundar, Larry, and Sergey just need to take the equity and the political capital they have as a CEO and the 2 largest shareholders and just stuff the decision down people's throats. So that's number 1.

The other thing that I think about actually relates to what I spoke about last week about Nvidia. Just to give you guys a little bit of a cleanup, after we spoke about Nvidia, the Nvidia team reached out to me and said, "Hey, Chamath, here's a bunch of stuff that will help you unpack what you spoke about in a little bit more detail."

So let me tell you what Nvidia told me, and then I think it relates to GCP. What they told me is that when you look at the SEC filings for Nvidia and add up all the revenue, it says that about 47% of the revenue goes to China, Singapore, Taiwan, I think it was. What these guys educated me on is that they've been trying to refine how this is described, but it's not what it seems.

They report revenue on what's called the billing or invoicing address. So, meaning, you're an entity, you'll buy a huge amount of chips or compute power, and you issue the invoice out of Taiwan, Singapore, or China. That's what represents 47% of Nvidia's reported revenue. It turns out that that's actually not where the silicon gets shipped. There's a Singaporean office that handles the invoicing.

Then I said, well, who are these companies that are using these places? It turns out that it's a bunch of major U.S. companies, including U.S. OEMs and clouds, that use Singapore for invoicing their suppliers. What industry sources have told us is that all of these products are almost always shipped elsewhere, including the U.S., Mexico, Taiwan, and all over the place.

So why is this interesting? There was a study by Artificial Analysis. Nick, I will send you the link. Basically, what it showed was that the clouds were not doing any real form of KYC on who is using their stuff. So now I think we get to this next place, which is that I think the Google business is profoundly amazing. I think they need to do something with the subscriptions that they have and put Gemini right in front, and really do a head-on attack against ChatGPT.

Those are the pluses. The minuses, or the risks that I think will now get uncovered, are: Are the clouds doing enough to make sure that China and other folks aren't getting access to compute that they should not get access to? And that could have an impact on some of their revenue. So that's a GCP problem. I think it's an Azure problem. It's an AWS problem, and that's going to need to get sorted out in the next couple of quarters.

If Western governments basically say to the clouds, "Hey, you need to KYC your customers so that it's not somebody distilling the best of our models into a place we don't agree with," I think there could be some blowback.

All these numbers are obviously backward-looking. Do we think—and I would be talking too much about tariffs—but do we think that the tariff story is going to impact Google next quarter? Do we think it's going to impact Meta next quarter? And to the extent that Google is saying, "Look, we're going to keep our big infrastructure spend—$75 billion—full steam ahead," which is great for the ecosystem, great for Nvidia, great for everybody in that world, do we think that holds up if we think there's a hiccup between now and whenever all this tariff stuff gets sorted out?

If it does, I think it's going to hold up. I mean, look, I just think that this investment in capex and the data center build-outs is so strategic right now. I think that the hyperscalers—it's only good for their businesses—but I think they see it as so strategic to their survival that they just have to do it no matter what. In other words, I think they're totally inelastic on this.

But let me just go back to Google. Nick, can you throw up this chart? I think the problem that Google has with respect to ChatGPT is that Gemini is just not getting the usage, and ChatGPT is just growing like crazy. If you look at how these models perform according to the benchmarks, Gemini is actually really good. Gemini has made substantial progress catching up to ChatGPT, but they have not caught up on the usage side.

Now, to Chamath's point, you could just make Gemini the default interface in Google Search, but that is true innovator's dilemma, because if you do that, you could be giving up more than half the search revenue. So I do think Google is in a really tough spot.

The longer they wait to make Gemini front and center, the worse this usage problem gets. What's basically happening right now is users are learning a new habit. They're learning to go to ChatGPT to get their questions answered, or just their searches answered, in a completely different way.

And let's be honest: They make using Gemini impossible. They can do just a much better job. At a minimum, give the best leading-edge Gemini model to the 270 million people who are already paying you for a subscription. Just do that. Do something so that you can start to blunt the growth of OpenAI. Otherwise, you're going to look back in 4 years and regret hemming and hawing today.

So, Chamath, you're the CEO of Google. You've got a $200 billion run-rate ad search business.

What's the right kind of integration of Gemini such that you don't massively disrupt the search ad business overnight? Or do you not care and you're just going to do it? I think that's the conundrum they're dealing with.

That's not a conundrum. If you don't understand how to value a dollar today versus a dollar tomorrow, and a discount rate to disambiguate the emotionalness of that decision, it's not a hard decision to make.

I think the more difficult question is: What does the integration look like? If you are already paying for a YouTube subscription or a Google One subscription, what should your experience be? That's a key question.

The second question is, today, they're already inserting Gemini in all kinds of uncomfortable ways. For example, if you use Gmail or Google Workspace, what happens today is all these random Gemini pop-ups come up all over the place. That is an implementation that happened at way too junior a level by people that have no product taste, so if you just stop that, it's not hard.

But right now, you have a bunch of people that don't have taste and are a little too junior creating a very cluttered experience. If you use the products every day, it would be hard for you to disagree with me. They have infinite capital and infinite access to talent. Capital doesn't buy taste, right?

So, what do you think's wrong?

I agree that the Google interface trying to incorporate Gemini into the 20 blue links or whatever is very clunky, but I don't know how you fix that without just getting rid of the blue links. I guess you can do a hybrid experience, but it doesn't compare to ChatGPT, which is all in on the AI experience. So, I think it's a real innovator's dilemma for them.

And look, the truth is the market was killing them already because people thought that that first little AI summary thing was actually going to kill the blue-link economy, right? That was their worry. So, it's a terrible conundrum if you're Sundar. How do you do both things at the same time?

Jason, I think you're right. There's a design way to do it, but they're not doing it. This comes down to taste. I think you can't have groupthink drive creativity and taste. You either have it or you don't.

You need to find the 1 or 2 people, and then you need to use the strength of your leverage as a CEO and the ownership that Larry and Sergey have to impose 1 person's taste on the decision.

Well, let me ask you a question. On the Google homepage, where they just have the search bar and you can basically submit your search, or “I'm Feeling Lucky” or whatever, would you replace that search bar with an AI chatbot?

No. Here's what I would do. I would first go to the critical other points that are around that today and not cannibalize the blue links. The obvious insight should be: If you look at the traffic patterns almost as a Sankey diagram—right? We just had one up. Show a Sankey.

What you should be looking at here is where are the entry points into Google that then result in a clickable link. What you would have is a very different Sankey, and what it would show you is that there are certain places that are highly deoptimized today for revenue-generating events. They happen as a byproduct, but they don't happen as the use case.

In that example, you would put Gmail as a critical place, the YouTube Premium subscription, the Google One subscription, and 5 or 6 other places. That's where I would put Gemini as the front door and start to habituate 300 to 500 million people a week in using that.

I think then you can figure out over time how much money you can make from all of that, or how it directs derivative revenue, and figure out what to do with Google.com last. My point is, the experience in Gmail should be done today. The experience in YouTube should be done today. The experience in Google One should be done today.

Okay, Nick, can you put up this? Here's their homepage. Right. This is the famous shot. I just checked; it hasn't changed in 25 years.

I would leave it alone because I think it's too disruptive and it's too politically fraught.

What if they replaced “I'm Feeling Lucky,” which is kind of antiquated now, with AI—I mean, Gemini, basically?

I would need to look at the Sankey to understand how many people are actually generating monetizable events or behavior from “I'm Feeling Lucky” in 2025. I'm not sure what the novelty was in. You know, it is the novelty. I think it's there as a whimsical thing.

I don't know if it drives revenue for them or not. If it takes the user off-site, I don't see how it drives revenue for them.

Here's a great Facebook story just to explain this and why I get so animated about this. I was telling this to my 80/20 team yesterday because we're dealing with this one product implementation issue, and engineers always end up fighting about whether things should be opt-in or opt-out.

It's almost like this religion where people feel like, “You cannot force people to do something. You should always make things optional.” I remember at Facebook, we bought this little company for a few million dollars that made contact importers, which essentially meant that if you were using any other email service to sign up for Facebook, other than Hotmail back in the early 2000s, we could import your address book.

Long story short, I remember getting into a huge fight about whether that thing should be opt-in or opt-out. I remember telling the guy—and I'm not going to say who it is because you guys all know him—“Honestly, just shut up and enjoy the success that will come from keeping this as an opt-out.”

It reminds me that at some point, somebody needs to impose their will to make very difficult product decisions. If you leave it to a team, you'll end up with—by the way, go to Gmail today with 95 Gemini pop-ups. It's not what will allow them to win with what they have, which is a better product, which, I will agree, is a meaningfully better model on multiple dimensions than anything else.

Okay, I know Sacks has to go. Before you go, David, I just want to hit Tesla for half a second. Tesla is out with earnings this week. The stock is up 8% on the back of that news, 23% up in the last 5 days, in large part because Elon basically said, “I'm getting out of DOGE and I'm going back into Tesla.”

So, the question is: What can he actually do to fix Tesla at this point? What does he have to do? And is he really out of DOGE?

No, I don't think he's out of DOGE. He didn't say he was out of DOGE. It was just a matter of how much time he could allocate to each thing.

I saw this before when I was part of the Twitter transition. For the first 3 months or so, he was basically full-time at Twitter headquarters, learning the business down to the database level. Every nook and cranny of that business he learned about.

Once he felt like he had a mental model and had the people in place that he trusted, he could move to more of a maintenance mode. I think the only way he can manage 5 companies is that he has these intense bursts where he focuses on something, gets the right people and structure in place, feels like he understands it, and then he can delegate more.

I think that he has reached that point with DOGE. But he was also clear that he's going to keep doing it because, if he doesn't, there's going to be a huge backsliding where all the corrupt interests will basically put back all this corrupt spending.

He's going to stay involved, but as an SGE, he's limited to 130 days a year anyway. It makes sense for him to now ration his days a little more closely. But he's got the people in place.

Remember, DOGE wasn't just him. It was also the U.S. Digital Service, which is basically the IT branch of the executive branch, and it got put under DOGE. My sense is that DOGE is going to continue. It's just that Elon is shifting to a mode where he can manage it 1 day a week or 2 days a week, as opposed to being there 5 days a week.

Steve Bannon says he wants the receipts. He says there are no receipts. We don't have the receipts. He says we need itemized receipts to see what has actually happened. What do you think? Why wouldn't we want to cut as much of this corruption as possible?

Well, I think we have to. We have to. The question is, should the public get to see exactly what's been cut and what hasn't? Why wouldn't they? Of course.

What we really need is for Congress to now embrace all the corruption that Elon has found and eliminate it from the budget. At the end of the day, in order to capture the savings here, we do need those appropriations eliminated from the budget.

My biggest concern is not something that DOGE is going to do or not do. It's not up to DOGE to do that. It's up to, frankly, these old bulls in Congress who control the appropriations process.

Are they going to backslide and just put the spending back in because it's easier to do that than to engage in this logrolling, or do we take advantage of this incredible sacrifice that Elon has made?

This has cost him enormously. One of the reasons why Tesla is down is because you've had crazy leftists engaging in terrorism, firebombing Tesla dealerships. In any event, he's made this enormous sacrifice in order to expose the corruption.

Look at what we've learned. We've basically learned that this whole NGO thing is a giant scam where the people in government give enormous amounts of money to their friends, probably with the expectation that when they leave government, they're going to be next in line at the trough. I think Elon's done an enormous service exposing this.

But it’s not entirely up to him. In order for us to realize the benefit, we need Congress now to act on that, and I’m just afraid that’s not going to happen. What kind of total number do you think we get?

I always say a dollar saved is a dollar made, and we should say amen to that. The question is, what do you think the total number is ultimately going to be?

Well, he says they’re at $160 billion right now. I think, by the way, that’s an annual number, as far as I understand it. What I always hear with spending is that they multiply everything by 10 because they assume it’s going to be 10 years. So if it’s $160 billion a year, that would be $1.6 trillion over 10 years, probably more because the spending always grows.

And look, that’s $160 billion that we weren’t planning on saving before Elon got involved in the government. So if that’s all it is, that would be great. I think there could be more. It really comes down to whether DOGE continues to be supported by the political process.

Elon can’t force that, right? Ultimately, it’s up to legislators to take advantage of the work that he did. Ultimately, it’s up to the people to put pressure on those legislators to embrace what he did. He’s done an enormous amount, but there’s an old saying that you can bring a horse to water, but you can’t make it drink.

If the entrenched political interests at the end of the day aren’t going to embrace the savings that Elon has found, I don’t know what we can do about that. There’s the administrative and discretionary aspect, but I’ve said for a while that it’s going to be necessary to have statutory resolve to actually fix the deficit problem in the United States.

That means getting Congress to act, and Congress has not shown a willingness to act. Every time I go to D.C., I come back more disappointed. I share the stories on this show and talk about the fact that there are very few members of Congress standing up and saying, “At the end of the day, the first thing we need to solve is: How do we reduce the deficit?”

Then we can address all of the policy issues that we want to talk about. But if we don’t solve the deficit problem, there is no policy. There is nothing we’re going to be able to do because Treasury rates are going to spike. There’s going to be no capital. There’s going to be no funding. We’re going to go through a debt death spiral.

At the end of the day, we really do need Congress to stand up and lead here. It is not on one man’s shoulders. There’s only so much Elon can do without statutory authority from Congress. I think that’s really important in this budget reconciliation process, and it’s really important for the president to also help lead Congress to where we need to go on this.

Members of Congress are elected because they get stuff for their constituents. The more they get, the more likely they are to get elected, and next cycle they’ve got to get their constituents more stuff. So it creates an escalating spiral staircase.

It’s a problem in how democracy operates. It creates a tragedy of the commons. There’s a collective-action problem because, like you said, each individual congressman or senator primarily wants to get stuff for their district or their state, for their special interests. There’s no one really looking out for the public interest, the overall common good.

These appropriators are basically engaged in logrolling, where they’re willing to give their colleagues their pork if they get their pork. This is why we have a $2 trillion deficit. It’s a really hard thing to fix. People realize that they can vote themselves all the money, and that’s what they do.

I remember Bush 41, in his State of the Union, called for a line-item veto. This was 30 years ago, and we still need that. I do think that if the president had more authority over this process, it would be less of a tragedy of the commons.

In my naive youth, I was very against that. I felt like it was giving too much power to the president, and I never really understood the wisdom of it until we found ourselves at the end of the cycle, which is where we are here.

Yeah. Unfortunately, Congress will never, I think, change its ways until it’s forced to by some sort of crisis. Eventually, we’ll be in a debt crisis, and then Congress will finally see the wisdom and appreciate what Elon did. There are all these entrenched political interests complaining about Elon. Again, he undertook an enormous sacrifice to try and fix our fiscal situation, which is unsustainable, before there’s a crisis.

One day there’ll be a crisis, and everyone will see that he’s right. But to answer your question, Andrew, Nick, can you pull up this Polymarket?

So this is the Polymarket on how much spending will be cut in 2025.

Polymarket is showing an 84% chance of less than $50 billion being cut in 2025, and only a 10% chance.

Isn’t the number on DOGE.gov right now something like $160 billion?

That’s what they’re saying they’ve cut in contracts. That means the annualized run rate of savings from cutting contracts.

Oh, because these are multiyear contracts.

Multiyear contracts.

But then there’s also reporting that, like, $92 billion of it, I think, isn’t actually itemized, so it’s hard to understand what’s actually on the list. They haven’t done anything yet in defense, right? So that’s a big honeypot, I guess, if you want to use that word—potentially, that system. Do you think he’s going to go there?

Well, I mean, there are people at DoD and at HHS, but I don’t think we’ve had a readout from any of those people yet, right? HHS is huge. If you look at their list on the DOGE tracker on the website, HHS is the number-one department for savings right now.

They’ve got the contracts. Defense is on. So they are finding—remember what they’re doing is they’re just cutting wasteful contracts. They’re cutting stupid stuff that isn’t used or needed, all the discretionary stuff.

No, no, but my point is like all of these things are still these are not line item appropriations, right? Because you can't cut those.

Well, here you can see the Department of the Interior had a $3 billion contract for refugee resettlement with a company called Family Endeavors, Inc. And so they put that as the number one contract they cut on the website. Saves $2.9 billion total. Treasury, they cut a deal with a company called Centennial Technologies, which is an enterprise software company. Saves $1.9 billion.

What is the Department of Defense technologies? Mike, I mean, that's the thing. Go into this stuff. You can in that are all like small. You can click on it. It shows you the contract. Here it is. Well, no. I'm going to go and try to find the Centennial Technologies. Who are these guys? You're going to invest?

Should we talk a little Tesla, or do you think we should move to the science corner?

What I would say quickly on Tesla is that I would encourage all of you guys to try FSD. This is the reason I bought my Tesla. I’ve been using it, and the quality of FSD is just so good. There was a revision of it probably in the last month or 2 that took it from, like, 90.8 to 99.5 or something. There are just very few disengagements.

Okay, so are you a believer that Tesla is doing real robo-taxis in 2 years?

Yeah. I mean, right now my car is effectively a robo-taxi. The shittiest part about my Tesla experience is that I have to hold the steering wheel and look ahead because the camera is always looking at me.

My wife will come home angry because she got suspended for looking away and looking at her phone. They turn it off. Tesla punishes you if you look away too often. They put you in timeout for, I don’t know, how many days or whatever. That’s the worst part about the experience, because it’s so good. You’re just like, “Can I take my hands off the wheel and do something else?”

Well, there are a lot of videos getting posted on Twitter of FSD being on, and then the car drives into the sunlight. At a certain angle, when the sun hits the cameras, the car goes into an emergency disconnect on FSD.

I’ve had that happen, so I know it’s true. I know it’s real. It is a bit of a risk with the camera-only mode on how Tesla operates FSD, if they can actually run a full, never-disengage robo-taxi service like Waymo does.

Still pushing lidar on them? Is that what you’re doing?

I don’t know. That’s the word on Twitter: Watch out, because without lidar you’re going to have a lot of disengagements.

Elon’s never doing lidar.

Here’s an interesting factoid. I’ve been working with somebody here on trying to figure out what next-generation AI data centers look like. Do you build inference POPs all over the country?

I talked to some folks at Waymo, and I talked to Teedra [?]. What’s interesting is that the Waymo approach is that they don’t need POPs anywhere because all of the models are on board. These things go fully autonomous out into the wild. They can make every decision themselves.

They have the lidar, and then they come back to the fleet and can resync the models. Whereas everything else—particularly if you have human intervention—all of the food delivery folks that have central NOCs of people who can engage and intervene need all this real-time ability, so that things are multi-milly maximum.

Awesome. Should we go to the Science Corner, my friend?

Yeah, if you're ready for it.

I'm ready. Today's Science Corner is on a discovery of a massive thorium reserve in China and a disclosed molten-salt reactor in China that's been running for some time. So, as a primer, before I do the tee-up for this, Friedberg, tell us about these thorium reactors.

I think the most important thing that everybody should know before I tee this up for Friedberg is that what he's going to talk about is a material called thorium, which doesn't split. It doesn't release energy unless it absorbs a neutron, and then it transmutes to uranium-233. By this, what I mean is that this is a transmutation that everybody should be able to get behind. Over to you.

Hey, he got it. He got it. You get it, Andrew?

I got it.

It's a beautiful thing. I don't think we brought it up in Science Corner before, but apparently there was a giant thorium deposit discovered in Inner Mongolia that has enough thorium to effectively power the entirety of China for 60,000 years. It was found at the Bayan Obo mining complex. It's a million-ton reserve in Inner Mongolia. Thorium is much more abundant in the Earth's crust than uranium, which is what we've historically used for fission reactors and what we use today for all the new-generation nuclear fission reactors.

For a long time, folks have talked about building a thorium reactor. There are a lot of challenges with building a thorium reactor. It's a new set of systems. But if you build a molten-salt thorium reactor, you can ultimately produce energy in a way that doesn't actually have the risk of a meltdown. It can passively shut itself down. It's not high-pressure, so it can't explode. So it's technically a much safer, much more reliable way of producing power with a much more abundant fuel source.

People in the US have been talking about building these thorium molten-salt reactors for decades. In fact, in 2009, the US Geological Survey went out and did a study to identify how much thorium existed just in the continental North American region. It turns out the US has about 64,000 tons, and Canada has 172,000 tons of thorium reserves. China just discovered 1 million tons in this one region. The amount of thorium that we have in the United States is enough to power our country for centuries.

So in China, it turns out they built a molten-salt reactor. This was another secret. They've been operating it now for some time. Last June, while the reactor was running, they replaced the fuel. So it basically showed continuous use of the reactor without needing to do a shutdown and refueling cycle. It's a 2-megawatt experimental unit, and they're actually building and planning to bring online a 10-megawatt unit by 2030.

This was all shared in a private meeting. It got out. The discovery in Inner Mongolia was also kept confidential for nearly 2 years, and then it got reported. After it got reported, it was ultimately confirmed in a government meeting. That's how we know about it now.

I think this is just another critical point about this. Once China gets this 2- and 10-megawatt system running, this allows for very small, very modular nuclear reactors using a very abundant fuel source that can be quickly built, stood up, and safely operated in a distributed way. You don't need to have 1 gigawatt, which is 1,000 megawatts, in a station running centrally, and then have to invest in a big construction project to build and operate the thing.

You can have many small reactors split around neighborhoods, split around cities, and so on. This is not built into China's energy projection. They actually have an electricity forecast that, over the next 15 years, gets them from 3 terawatts of electricity production to 8. That's through a combination of solar, hydroelectric, and the Gen 3 and Gen 4 traditional uranium nuclear reactors that they're building out.

So if you add this in, it provides even more, or perhaps lower-cost or more distributed, energy-production capacity than is even provided in the forecasts today. Again, I think this is one of the base drivers that is part of the compounding effect of China's economic advantage for this century: energy production, the cost of building new energy-production systems, and the cost to operate those energy-production systems.

Can I say something about the China conversation—about picking national champions and getting focused? This was an area that China chose, and now the US finds itself years behind on a technology that we invented.

This was research that happened at Oak Ridge National Laboratory. We pioneered this decades ago, and then we shelved it. The person in China who led it gave credit to the United States and basically said, “Americans let the research wait for the right successor. We were that successor.”

So here we are. We invent this cutting-edge technology. It's a huge leap in innovation. But we put a regulatory fence around it and blocked it from getting organized, and they're like, “You know what? We can just do this, and we'll figure this out.”

Yeah, it's really a self-goal here.

They have an effectively infinite thorium reserve. They don't need to refine it. One of the key advantages of thorium is that you don't need to go through a refining process. With uranium, you have a very expensive, very difficult refinement process where you end up, I think, getting less than 1% of the uranium that you pull out that can actually be used as fuel, and you have to enrich the uranium to find the fuel material.

With thorium, 100% of it can be used as fuel. So it's a very low-cost, very easy kind of fuel source that can be quickly put into production if we actually build out the supply chain and build out the energy systems to utilize it. It's a real reinvention of nuclear-energy technology, and this will take off. I mean, once this experimental reactor is working, reactors are going to go boom.

Imagine how many fundamental leaps in science we have funded that are just gathering dust or falling through the cracks, that other countries are using right now and probably snickering a little bit behind our backs about the fact that we did all of this work and then they were able to take advantage of it.

Well, here, I'll show you. Here's another thing. Let me put this in the group chat real quick. How much of this stuff do we—what's our ability to actually do this?

We invented it. We invented it.

Well, the big problem has been the regulatory fencing we put around it. Andrew, there's a bunch of administration. There are forms. There are pronouns. There's all kinds of stuff.

Just speaking to this point, we've talked in the past about nuclear fusion. Andrew, I don't know how familiar you are, but fusion is different from fission. Fission is where you take a heavy element, like uranium or thorium, and it breaks, and you release energy. Fusion is where you take a light element, like hydrogen, you accelerate it, you make it really dense, and you jam the protons together.

When you jam those protons together, they fuse. Hydrogen turns into helium. This is what goes on in the sun, and energy is released in a different form through accelerated neutrons. We capture that energy and make electricity.

Fusion is this next-generation technology where we could theoretically just use water from the ocean to create all the electricity we need on planet Earth. A 10-meter-by-10-meter-by-10-meter cube of water makes all the power the entire planet uses every year. That's what you could ultimately do with fusion technology.

So it turns out this is a satellite photo that revealed a very large-scale fusion research center in Mianyang, China, which we did not know about and which was just kind of stumbled upon. It turns out it's now the largest experimental fusion facility in the world—50% larger than the National Ignition Facility run in the United States.

Not only is China getting this new energy infrastructure built with all the stuff that was discovered last century, but it's now getting ahead of us on the new discoveries to be made in the next generation of energy systems, which is fusion. At some point this century, fusion will work, and we'll end up having the ability to unlock effectively unlimited free energy.

These are compounding value creators. If China can make new energy systems cheaper than the United States and scale them faster, it will ultimately have much cheaper, much higher volumes of electricity, which gives it the advantage in manufacturing, production, transportation, and everything. That's where the economic advantage will ultimately accrue to them.

So this is key to the strategic puzzle of what's going to happen in the great race with China this century: the build-out of electricity and the cost of that build-out. The regulatory constraints need to be addressed first and foremost. That's, in my opinion, the number-one mission-critical strategic imperative for the United States right now.

Awesome. Anyway, that's just a red flag on China Science Corner for the day. Thorium, man. So, how was it, Andrew? How do you like hanging out with the best?

I loved it. I'm still getting used to it. I'm still getting used to the whole flow, trying to figure out how you guys all roll. But it's good. It was a lot of fun.

Do the outro. You’ve got to do your outro. You’ve got to do the outro.

Oh, yeah. I’ve got to do the outro. You guys do the outro. Come on, it’s your show. Do the outro.

Well, okay. I’ll do it.

I’ve never done an outro on you.

Do the outro.

On behalf of our crypto czar, David Sacks; our prince of panic attacks and Sultan of Science, David Friedberg; and our new bestie guestie, Andrew Ross Sorkin, thank you. I am your chairman dictator, Chamath Palihapitiya. Thank you for listening to the All-In Podcast. We’ll see you next time. Love you.

Thanks, guys. That was a lot of fun.

Awesome. Let your winners ride, man.

And as I said, we open-sourced it to the fans, and they’ve just gone crazy with it.

We should all just get a room and have one big, huge orgy, because they’re all just useless. It’s like this sexual tension that we just need to release somehow.