Trump：向旧金山派国民警卫队、中国稀土贸易战与AI公关危机

All right. You guys were talking about the fun shenanigans you were having in San Francisco and at Mark Benioff’s Dreamforce. Did you see Bryan Johnson was at Benioff’s CEO dinner?

Oh, I didn’t see him.

Yeah. Matthew McConaughey was at our table at dinner.

All right. All right. All right.

McConaughey is fabulous. He is funny.

Did you see where I was sitting at that dinner?

Yeah. Table 1.

Right next to the king himself. The king himself. Exactly.

The king who? The king and I?

Benioff is a king.

Well, he’s the king of San Francisco at his party.

He’s the king of San Francisco.

Yeah. Obviously, he has to bend the knee to the king of kings, which is Trump.

I was sitting next to Benioff there. It got weird.

It got a little weird.

No. Well, what happened was—did you see The San Francisco Standard?

Yeah.

I don’t even know how they can write a headline like this. Benioff interviewed me at Dreamforce, and somehow this was headline-worthy for The San Francisco Standard. They said, “Mark Benioff dodges political questions, then fawns over David Sacks on Dreamforce day one.” That was just referring to the interview we did. So we trolled them by taking that photo.

That’s a little weird. I thought the fireside chat you guys did was great.

It was great. It was just an interview.

It was really good, actually. He’s a very good interviewer.

Yeah, totally. But apparently, to The San Francisco Standard, it’s some sort of transgression that I was even interviewed.

They didn’t get the memo that cancel culture is over.

Yeah.

They can’t.

I know. That’s what it feels like, right? They’re trying to gin up some sort of brouhaha over the fact that I was speaking.

Yeah. “Oh, you can’t talk to David Sacks. He’s in the Trump administration. You’re not allowed to talk to him.”

About 100 people a week say that.

Literally, the AI guy and the AI CEO shouldn’t talk to him. It’s like, you know, the AI company should not talk to the government’s AI person?

Right?

By the way, did you see that there was one other very important conference in town at the same time as Dreamforce?

Okay.

I discovered this on my X feed.

I feel like a cold open about time. What? The Young Socialists? Or was it the Republicans?

It was called SlutCon.

Oh, SlutCon? Really?

Oh, my Lord. Look at this website.

Oh, this is the agenda: “How to Be Good at Sex with AI.” Let’s see.

You cannot do this. Cut this out.

“Women venting.” Oh, Sacks, that one’s for you.

I would have planned an orgy, Friedberg here. Friedberg’s giving a talk: “A Fluffer’s Guide to Jobs” by David Friedberg. Is that a panel, or is it a fireside, Friedberg? Which one?

“Reinforcement learning: how feedback small groups.”

Oh, you could rewrite your positions. I see that. “Reunderwriting Your Sexual Positions” by Jamal.

Jamal's the last one—the VIP room.

Oh, VIP room. Yeah, that’s Jamal, for sure. Jamal.

Sacks, how did you hear about this? Don’t tell us that you ran out of bed and went over to the East Bay.

Sacks, are you on the mailing list?

No, they invited me to speak.

My X feed just— the new AI algorithm is trying to find things you might be interested in.

Sacks is also keynoting this. My X feed showed me SlutCon, so I clicked on it, and now it’s been nothing but SlutCon for a couple of days.

Chamath, what did you think about walking around San Francisco? We went on a walk around downtown San Francisco.

Well, we were trying to find a place to get a drink between the Benioff dinner and Sacks’s talk. We ended up at the W Hotel, but I was worried for Chamath because it was his first time on the streets of San Francisco, walking around on his own two legs. How was it, Chamath?

Did you throw out those shoes? The Louboutin shoes?

I burned them.

The soles, I’m sure, were sullied beyond repair.

I burned them. Those are one-use for Chamath. Those are single-use Louboutins. Once I got in the car, I gave Eugene my shoes and said, “Burn these, Eugene. I never want to see these ever again. They’ve touched the pavement of San Francisco.” All I said was, “Yuck.”

Why don’t you just have your domestic staff carry you on their shoulders? I don’t understand.

Why the crowd? It makes no sense.

So Benioff pulls off this awesome convention, this conference, and then Trump does this press conference yesterday saying he’s going to send troops into San Francisco. Did you guys see this?

Yeah.

Well, yeah, but Benioff’s thing happened first, right? Benioff made news because he said, “Send in the National Guard to clean up San Francisco.” But I actually talked to him about it and got a little bit of context.

Which, of course—

Well, no, he wasn’t misquoted, but it happened in a more innocuous way. He wasn’t trying to weigh in on a political issue. Basically, what happened is, I think, he was describing the security at Dreamforce and how they bring in a couple hundred extra off-duty cops to provide security. So then the reporter tried to stir the pot by saying, “Well, would you bring in the National Guard?” And Benioff said, “Sure, if they can be cops.” That was kind of the end of it. Then it turns into this massive story that Benioff is supporting the National Guard coming to San Francisco, which I think he should support, but it’s not exactly what he said. He was saying that for Dreamforce, he would hire them.

Yeah. Who knows exactly what he was trying to say, but the point is—

He could have just been talking.

Yeah. He wasn’t seeking to make the news that he ultimately made, but it’s not a bad idea. It’s actually a good idea. Do you think Trump teed off on that press cycle when he was doing that press conference? Was it kind of off the cuff when he said, “Oh, they should send the National Guard into San Francisco next”?

Well, I don’t know.

Yeah, I wasn’t there for that. But San Francisco is a pretty obvious city to send in the National Guard because we have this downtown area where it’s like that zombie-city area, the Walking Dead part on Market Street. Everybody understands that it’s because they won’t stop the drug dealing. So you’ve got open-air drug markets on the main business thoroughfare of town, which is Market Street.

By the way, let me just back up and say that we have the best mayor we’ve had in decades now.

Daniel Lurie.

Dan Lurie’s great. And I think Brooke Jenkins was the DA that we supported to replace Chesa Boudin after we recalled him, so I think she’s good, too. But they’re still working within the confines of a legal system here that is very left-wing. You’ve got all these leftist judges, and I think they’re somewhat constrained or hamstrung in terms of what they can do.

Again, the evidence of that is that you’ve got hundreds of these drug dealers operating with impunity downtown, and you’ve got this sort of zombie city of people who are so sick with drug addiction that they’re pretty much living on the street. I think it would be a relatively easy thing to clean up if they send in the National Guard.

I would argue that—I’ve been working in San Francisco, and I think it’s gotten a lot better this year. So I looked up, after all this Benioff brouhaha and the National Guard thing that Trump said yesterday, what the city’s publication shows: crime is down 30% citywide and down 40% downtown. Homicides are at a 70-year low this year. There are record lows in the number of tents left on the street. They’ve cleaned up all the tents, and I’ve seen this because I drive to my place in the city and they’re all gone.

For the first time in 7 years, they’ve actually got a net gain in total police officers on the force. Now they’ve got convention bookings up 50%, with hotel bookings up 60% because of it. Here’s what DA Jenkins put out: felony convictions in cases this year—narcotics, 88% of cases have resulted in convictions; robbery, 85%; burglary, 72%; and gun possession, 70%.

All the car break-ins are down to a 25-year low. I’ve noticed it acutely in my experience in the city, so it felt a little shocking to be told, “We need the National Guard.” This is a city on the upswing, in my opinion. I think things are getting much better anecdotally, and the statistics represent it, too. It feels like there’s a little bit of a runaway narrative with San Francisco relative to what those of us who are in the city are experiencing.

Let me thread these 2 stories together and hand it to you, Sacks. The interesting thing, when Sacks and I were ousting Chesa Boudin, was that one of the things we learned was that Chesa Boudin had this famous quote where he was saying, “Oh, well, we can’t get rid of the fentanyl dealers. I’m for drug regulation, but not for fentanyl because it’s a superdrug that kills people very quickly.” It turned out he said all these Honduran nationals were victims.

Yeah.

Which, theoretically, could be true, but it’s kind of like in that—what do they call it?—suicidal-empathy category.

But it turns out this is something where ICE and the DEA have done some work.

They actually started deporting these people last year because the people in San Francisco, pre-Dan Lurie, refused to do it. So the federal government has been taking action on this specific issue. We've also been taking out the boats bringing the drugs here.

You're not seeing a city government try to block federal action either, which is different from what you see in other cities. By the way, this is a good forewarning of what Mamdani's New York is going to look like when Mamdani's New York takes off. Imagine what's going to come into the city because there's now going to be a permissive culture and a permissive society to do this sort of stuff.

Yeah. So this is where we'll get strong alignment and agreement: deporting Honduran fentanyl dealers. What's taking so long? I would ask the administration to take a beat on the San Francisco thing because this city is making major improvements, and I feel very good about the direction. It's not like other cities that are on the downswing, and I also think AI companies are setting up in San Francisco.

Money's coming back. I say the power move by Dan Lurie would be to say to the Trump administration, “Hey, can we get, I don't know, 3 months of National Guard just to be in every BART station and then on these 4 corners in the Tenderloin, just as a deterrent?”

You know what people would do if they did that? People would protest and riot because they'd say, “We don't want federal policing of our city.” It's going to be—

For sure. I think you would get 10% to 20%, even in San Francisco, rioting. I think the rest of the people would be like, “Okay, it's a short duration. It's on these corners for this purpose. Okay, bring it.” That's why I think the Democrats are making a stupid decision.

But if the city's doing its job and the city's improving, why do you need that? It's getting better. This is not the—

Yeah, that's if things are worse and you're in a spiral. It—

—is a backstop.

Sure.

Yeah. Maybe the reason it's getting better is because that backstop is there.

What? The National Guard backstop? They're hiring more police. They haven't done that in 7 years in San Francisco. They're finally adding to the police count, and they're enforcing and getting convictions. It's kind of working.

Well, I'll partially agree with both of you.

So, I agree with Friedberg that we have the best mayor we've had in decades with Daniel Lurie, and I think he is making progress and the city is bouncing back. I think we probably had the lowest low during COVID, so a lot of that is just finally the bounce-back. Then we got the AI boom, which is creating a lot of wealth, bringing a lot of people back to the city, and starting to absorb all the vacancy we have.

In any event, all those are good things. Nonetheless, we still have this problem of this blighted area of Market Street where, of all people, Gavin Newsom proved that we could clean it up. Remember when President Xi came to town from China for that summit—

—and all of a sudden, magically, that whole area was cleaned up in 48 hours—

—and all the drug addicts and all the drug dealers basically disappeared. I don't know where they sent them. So it shows that there is a way to clean this up virtually overnight, and I think somehow the mayor and the DA are constrained in what they can do. You could have a targeted operation here by the federal authorities to go in and clean that up, and I think that'd be a good thing.

What do you think would happen in the city if they did that with respect to the population? Are they going to welcome or not welcome federal policing of their streets?

Well, again, I don't think we should be held hostage by lawbreakers who threaten violent protests and chaos in response to law enforcement authorities doing their job. I don't think we should be blackmailed by that, threatened by that, or held hostage by that.

But they're getting convictions. They're making arrests.

I think that's awesome. I'm applauding their efforts, and I'm not criticizing them. What I'm saying is we still have this intractable problem of this area of downtown where, like JCal said, there's a network of drug dealers.

Just to build on your point, Jason, about the Hondurans, there's actually an article in the San Francisco Chronicle talking about this, where the majority of the drug trade in San Francisco comes from a single town in Central America. So, yeah, why couldn't you just round up this whole network in a very targeted operation and deport them? That would clean up a huge part of San Francisco.

I think there's an opportunity here. If you look at the cities where it's been very successful, Muriel Bowser, who's the mayor of D.C., obviously wasn't thrilled about the president sending in the National Guard, and she said so publicly. That being said, she also cooperated with it, and it redounded to the benefit of all the citizens of Washington, D.C. So I think there's an opportunity here for the mayor to cooperate with the Trump administration, do a targeted exercise, and clean up the streets of San Francisco.

Or he can do it himself. Did you guys see this interview with this woman in D.C. who was describing what it's been like in D.C. over the last 40 years and seeing the changes?

Yeah, she feels safe.

Elon retweeted it. Nick, I sent you the link on Signal. Just listen to what she says.

“I'm always got to be the person to say it, and I don't want to be the person to say it. But Donald Trump has been president for 9 months. Nine months. And I'm seeing videos from Chicago and D.C. where people are like, ‘Yo, I feel safe. My kids feel safe. This is the best thing he ever did for us.’

“So what was happening all these years? I've known D.C. and Virginia to have this type of thing going on my whole life, and I'm in my 40s, almost touching 50. So what was really, truly going on if a motherfucker became president 9 months ago and cleared that out with just a flick of a pen? I'm just thinking, did y'all really ever care?”

Exactly. So I think the point she's making there is we don't have to live this way. This is a choice. We don't have to live in San Francisco with our main drag, our main street, Market Street, basically being an open-air drug market, where you've got hundreds of people who are addicted doing drugs in the street and doing other things in the street.

Again, Gavin Newsom proved that it could all be cleaned up very, very quickly. So, look, I think it's great that Daniel Lurie, Brooke Jenkins, and the new administration are doing a better job, but they're still constrained and hamstrung by their own—

—by the city's—hold on, by the city's—

—liberal mindset, extreme liberal mindset, towards what's possible. And like that person just said in the video, Trump just did it with a stroke of a pen. This is a choice. We don't have to live this way.

It's the biggest unforced error since opening the southern border. These are just the 2 stupidest decisions the Democrats ever made. Obviously, the easiest thing to do to split the 2 approaches here, Friedberg, is to have Dan Lurie say, “Thank you for offering, Trump. We would like to have it for this period of time in this location.”

Then you could communicate to the people of San Francisco, “It's just a test. It's just on these streets.” When it comes top-down and you don't have a choice, and JB Pritzker does what he did, which is, “I'm going to ban you from coming in and reducing crime,” do you hear yourself, Pritzker? “I want to stop you from lowering crime.”

There's something else we've got to do as well, which is stop funding the drug addiction because the city provides something like $2 billion a year to these nonprofits that are supposedly there to take care of this homeless, addicted population. They manage them like their flock because for every addicted person they add to the flock, they get paid more money.

Obviously, this is why we have this problem here. In fact, the homeless we have in San Francisco are not predominantly from San Francisco. They're from all over the country and all over the surrounding area. They come here for the benefits, so we create the magnet for all these people. Part of it is law enforcement, and another part is just cutting off the flow of money that's funding their drug addictions.

You are 100% correct on that. I completely agree with you on that point, and I think forced treatment-center transitions are needed. We have a program—I’ve mentioned this on the show before—but for those who haven't heard about it, here's the absurdity of these programs in San Francisco.

There's a program called the managed alcohol program. $5 million per year out of the San Francisco city budget is spent on this program, where you can walk into a hotel and they have free beer. You ask for a beer, and they give you a beer.

What?

You drink the beer, you go out, you party, you come back, you get another beer. They have all these studies that rationalize this program, which is, “Oh, this is a way to help alcoholics recover from alcoholism when they're homeless and they need this treatment.” But here you can see this is from last year: free beer.

Yeah. So the treatment is beer, and that works for alcoholism, I've heard.

But it’s one of these absurd things where now it’s kind of a joke. There’s also this closed-off part of downtown where it’s a little bit of a “do what you want” area, where they keep folks. It’s like, “Hey, you can use your needles here.” Then folks will get a drink, play music, and go after the thing. It’s really sad how these programs, obviously well-intentioned initially, become some of the most absurd.

If you want more of something, keep spending money on it. In fact, there is a program for people who are addicted to gambling, and they give Chamath free flights and cranberry juice at the Wynn to help him with his craps addiction.

Yeah, it’s amazing. If you keep going, one’s going to land.

You send a private jet and extend a massive credit line. It’s amazing how many gambling addicts you’ll attract to your casino.

By the way, did you guys see that J.B. Pritzker disclosed his income from 2024?

Oh, let’s guess the over-under. Give us an over-under.

Out of the $10.3 million, $1.4 million of it was gambling winnings from Vegas.

What? Yum yum. Okay, all-new respect.

$1.4 million playing blackjack. So he’s counting cards. Yeah, definitely. Definitely counting cards, David Sacks.

Well, no. I’ll just tell you, to win $1.4 million in blackjack, you basically need to be betting $100,000 a hand.

Even I don’t do that. That’s the only way you can win that much, which is a lot of money per hand. I’m not judging. I’m just saying.

I mean, that’s Dana White-level. That’s David Sacks at Caesars.

No, I’m not at that level.

Come on. David plays at the $100-a-hand table. By the way, just to put some numbers behind it—

Speaking of Vegas, aren’t we going next month?

Yes. Yes. Yum yum.

Yum yum. Just to put some numbers on it, San Francisco’s homeless budget is unbelievably high—an estimate of around $800 million, with the lower estimate at $700 million. The low estimate for the number of homeless people is around 8,000; the high estimate is around 20,000. If you split the difference between those two, they’re spending $52,000 a year per homeless person. $52,000 a year. It is a solvable problem.

Well, that is an incentive to solve the problem because, again, a lot of the money goes to NGOs, and they get paid for every addict they can add to their flock. Thomas Sowell had a great line about this: You get as much homelessness as you’re willing to pay for.

Well, and your point before is that other places don’t give these services. Therefore, the homeless slash junkies converse about this, and then they pick the right city.

We don’t have to live this way. I appreciate the mayor. I appreciate the DA, but we don’t have to live this way, and there’s more that can be done.

That is exactly it. You do not have to live this way.

Yes. Do not stand for it, folks. Okay, here we go. Lots of topics. The U.S.–China trade battle is continuing, and this time we have a new wrinkle. Trump and Xi are set to meet in South Korea later this month. They’re going to work out a grand trade deal, and this would be their first in-person meeting since 2019—6 years ago.

But last Thursday, China announced new export controls on 12 of 17 critical rare-earth minerals, effective December 1. Obviously, we need these for EVs, batteries, your AirPods—everything. On Friday, Trump threatened a 100% tariff on all Chinese imports, on top of the existing tariffs, as of November 1, and accused the CCP of trying to “hold the world captive.”

The stock market tanked, then recovered on Monday when Trump told everybody, “Don’t worry about it. It’s going to be all good.” So far this week, things have mostly de-escalated. Friend of the pod, Treasury Secretary Scott Bessent, said the Trump–Xi meeting is still on track, and, quote, “We have substantially de-escalated.” He also said the 100% tariff does not have to happen.

However, on Wednesday, Bessent called the export controls unacceptable and said that it’s China versus the free world. Let’s take a pause here. Friedberg, thoughts on price controls generally? If you want to expand into the overall relationship with China as it relates to rare-earth metals, go for it.

Bessent’s comments on the price controls speak to what he calls a non-market economy, meaning that the Chinese government is intervening and creating an artificial surplus in the market, which drives prices down and effectively eliminates competition. The Chinese push in the ’90s, supported by the CCP, ultimately eradicated competition and made it noncompetitive for U.S. companies, so they exited the market and now the United States finds itself completely dependent on Chinese suppliers.

Particularly as it relates to rare earths, there was this company called Molycorp that ended up going out of business. The asset—the Mountain Pass mine that Molycorp operated—became MP Materials, which Chamath knows well and which did this big deal recently with the U.S. government as they’ve tried to revitalize production at this mine and increase output.

Friedberg, can you explain price floors for people, just in case?

Price floors are when the government mandates that things have to cost at least a certain amount. So they increase the price of inputs and basically make it more difficult for China to undercut pricing, which increases the rationale for capital to invest in starting competitors because now they can have a guarantee of revenue or a guarantee of price.

But the problem is that I would argue the government’s role maybe shouldn’t be in setting price floors because that creates a very bad effect long-term on the market. The government should deregulate and create a more free market in the U.S. A large reason why mining and processing of the ores moved offshore was the regulatory environment in the United States, which made it difficult to compete with China.

We could argue about the environmental laws and how extensive and relevant they may be today relative to the ’90s, but this created a real challenge for companies to compete because it was so much more expensive to operate in the U.S. So I would argue we should be focusing not so much on setting price floors to create an incentive for folks in the U.S. to get involved in the industry, but rather on deregulating and providing perhaps tax incentives and other economic incentives.

That way, you don’t inflate the cost of things, but you create a much more liquid market, drive much more volume, and create more interest in building on this side.

All right, price floors. When we were talking about Kamala Harris running for president, she was asking for price floors and price fixing and making sure the government got involved in this. How is this different, Sacks?

Well, look, as a general matter, I agree that the federal government shouldn’t be setting the prices of goods, and it shouldn’t be picking winners and losers in the economy. But I think this is a different situation because, over the past 30 years, we haven’t really had a free market in rare earths.

What’s basically happened, as Friedberg was saying, is that China has been allowed to dominate this industry. Specifically, what they did starting 30 years ago was identify rare earths, rare-earth magnets, and rare-earth processing as very strategic elements in the supply chain, and they set out to dominate them.

They did that by massively subsidizing their businesses and driving all of the U.S. competitors and global competitors out of business. They were partially allowed to do that by stupid WTO rules that allowed so-called developing nations to subsidize their own industries, even though China is not developing. I mean, its economy is comparable to ours, and yet these WTO rules allowed it to effectively cheat by subsidizing its industries at the expense of our industries.

The result is that China now has a source of leverage over the supply chain. Over the last few weeks, they’ve used that leverage by imposing this vast new export control regime on rare earths. Their idea is to tighten the screws when they feel like it and to use rare earths as a form of leverage in the relationship with the United States.

So we’re well past economics here, and we’re into international relations. We’re into the balance of power. We’re into international security. I don’t think the price floors here are designed to make the government the mediator of what prices should be in the free market. They’re designed to create enough certainty for U.S. investors that they can reinvest in rare earths and reshore the processing and casting of rare-earth magnets.

I just don’t see another way to create the incentives for investors because they know that if they invest a lot of money in rare-earth processing, all China has to do is slash the prices and drive them out of business. China continually uses that tactic. So I think you have to address that if you want to alleviate American dependency on this critical resource, which China currently controls.

This started in 2007 or 2008 with Hu Jintao. What he basically said was, “We’re going to create 6 or 7 national champions,” and he wanted to create them in the critical areas that they needed to dominate over the next 20 years. One of those was rare earths. Another one was batteries and EVs, and yet another one was around the active pharmaceutical ingredients, or APIs, that are the core inputs for pharmaceuticals.

So what did they do? When Xi Jinping came into power, he continued that Hu Jintao strategy. What did they do? It’s exactly what Sacks said: They have this very aggressive form of mercantilism.

How do they affect that? They are willing to do it through provincial balance sheets, federal balance sheets, loan guarantees, and organizations that are quasi-public-private partnerships. They step into markets and can effectively change the spot price on demand. Now, why is that problematic in America? I’ll tell you the example of one of our businesses.

You do these deals that are called take-or-pay agreements. MP has a deal with General Motors. One of my businesses that makes LFP cathode has the same deal with GM. You do what's called a take-or-pay, which means that General Motors steps up and says, “Great, I will buy X amount of volume from you at Y price.”

So you take that deal to Wall Street and say, “I need financing to build a factory or a capability.” Contrast that with the Chinese company that just goes to the government, and the government says, “Here's a bunch of incentives and money and price breaks. Just go build it.” So already, we're a little bit behind, but it's much more free-market and it's more transparent.

You take a deal, you go to Wall Street, they give you money, you start building it, you start to produce it, and then you expect General Motors or Ford or Tesla to start buying this stuff. But as Sacks said, if all of a sudden China sees it, what they do is enter the market and dump an enormous volume of that product into the spot market. What happens? The price craters. Now all of these end buyers are in a really tough position because their competitors can buy at a much bigger discount than they can.

They have to flow that price through their products. Their prices are more expensive. They have less demand. That's the huge negative cycle that China has complete control of. So why is the federal government stepping in to create what is essentially a last-resort buyer of record? It is so critical in these markets because the U.S. balance sheet is the only one that can create a strategic reserve around these inputs, not dissimilar to how we do it with petroleum.

We can nominate things that are critical, all of these things that can now absorb the price shocks that China may otherwise be able to affect. So I think this structure, where we now have these public-private partnerships, frankly, is the only practical antidote to dealing with this issue. Otherwise, we're always going to be prone to this very clever, accurate, and smart mercantilism from the Chinese.

I'm a big fan of this general market structure. It's the only way that you cannot be theoretical and actually respond to the conditions on the ground.

Yeah, we've been doing a little research on this, and here is neodymium.

This is one of the two inputs for magnets—for permanent magnets.

You can see here: $50,000 a ton back in 2020. During COVID, it spiked almost up to $250,000, and then it came back down. So look, you can imagine how incredibly difficult it is, when there's that much volatility, to try to buy capex, to try to plan opex, to try to extract from the ground, go through all of the regulatory rigmarole that Western countries have to go through, and then all of a sudden have a sale price that you can't control.

If you look at the cost—we were just doing some research on it—if you get your AirPods, it's 1% or 2% of the cost of that. In medical devices, it's a lot more. But even in a Tesla Model Y, it's less than $10.

I think this is an inaccurate way of looking at it. You have to look at the actual mass. In cars and engines, you're talking about tens of kilos ultimately, and that's where the real costs build up. Separately, the permanent magnets that go into cars—you can work your way around those.

The real future markets that matter are in robotics, because these permanent magnets are the key input to the actuation mechanism of robots. Meaning, how do robots move? How do the joints move? Through actuation that happens because of rare earths and permanent magnets.

And as Elon said at the summit, he's building all those himself. Right now, only 7% of the known deposits of it have been found. There's plenty of it out there. But to your point, I came to the same conclusion as you, Chamath: We should be building a strategic reserve of these.

If you look, we could be buying them when these things are down and deploying them when we need to do something different. I think instead of trying to time the market, what the United States can be much better at is having a broad forecast of demand, knowing how many million tons of this stuff we will need, and just thoughtfully and methodically being a purchaser of record to create these strategic reserves.

The worst case that can happen is you can do a swap with these customers. So Tesla, all of a sudden, can't go into the spot market. They should be able to come to the strategic reserve and be able to buy lithium. General Motors needs something, whoever needs something—Figure robots need something. I think that mechanism is the only way we can compete effectively on a day-to-day basis with the Chinese.

I would argue that if we let the market work and create the necessary structure to allow the market to operate—meaning the regulatory barriers are diminished to develop mines and conduct the processing of the ores—we can solve this problem. There are 2 sides of the equation when it comes to this kind of rare-earth industry. One is mining the ore.

This Mountain Pass site, which is on the border of California and Nevada and is now MP Materials, has this rare-earth ore called bastnäsite. That's the ore that you get out of the ground. Then you have to process that ore to get the elements out of the ore that you want to use in manufacturing.

I went a little bit deep on the processing steps over the last week because I got really interested in why we don't do this anymore in the United States. There is so much nasty output from this process. You're using very strong acids to do leaching, processing, extraction, and refinement. All of the material gets separated out from the rock, and it's dirty. As a result, there are byproducts and waste and environmental exposure, but also human exposure.

The technology hasn't been deeply developed in 40 years. It's still 40-year-old chemical-engineering technology, and we stopped developing it about 40 years ago in the United States. It's not even technology. It's just metallurgy and chemistry. But there are techniques now that can be applied to do this in a scalable, safer way without any of the environmental hazards.

This is largely why this shifted over to China: They didn't have the regulatory burdens that we have in the United States, which made it much cheaper, and the labor cost was much lower. But with automation and certain systems, we can actually process this ore. It's not super complicated. It's a series of well-understood steps of chemistry and mechanical separation, so we could process ores.

The other important point is that there could be well over 1,000 times the proven reserves of rare earths in the existing proven-reserve catalog using new discovery techniques. The U.S. Geological Survey has this kind of Earth MRI system, where they're trying to do new mapping. There are also several private companies that can use radar and microwave imaging. Then you can use a system of interferometry, where you put multiple beams down underground and get a better read, or a better understanding, of where these ores might lie.

There are unproven reserves in Texas, Wyoming, Colorado, Missouri, and all over the continental United States. So if we decided that this was strategic to us, one of the important points is that we could unlock both new discoveries and new refinement systems in the United States and actually be truly competitive with China if we put the will behind it.

I think this is a good moment for the United States to observe what's going on and not just be focused on China's production. I think we could outpace them and accelerate through this if we chose to.

So, Sacks, where does all of this fit into the broader agreement? Do you think that they're going to go to South Korea and come out with a deal? Do you have a sense of what the contours have to be to make this a success?

The Treasury Secretary, Scott Bessent, who we know, just said that the Trump meeting is still on track. He also said that we have substantially de-escalated and that the 100% tariff does not need to happen. So he says it's still on track.

By the way, I think he's doing a phenomenal job here. He really understands the intricacies of this trade relationship, and I think he's capable of both applying pressure and then finesse. So I think he's doing a great job on this.

In any event, I think the reason why you want to have the leaders meet—President Trump and President Xi—is that this bilateral relationship between the U.S. and China will work best when there's an agreement at the top. What I mean by that is that when you just let the various bureaucracies deal with each other, you're more likely to have misunderstandings or food fights.

Let me give you an example. Over the past few weeks, obviously, you've had the Chinese Ministry of Commerce pass these export controls on rare earths, which created the possibility of greatly limiting access not just for American companies, but for companies all over the world, to Chinese rare earths. They've sort of backed away from that.

Separately, in the U.S., you had movements by our government to greatly expand the export-control list. Basically, there was an interim final rule issued by the Commerce Department on September 29 that said that if a Chinese company is on the export-control list, that restriction should also apply to any of its affiliates or subsidiaries that are more than 50% owned by that company.

That's kind of obvious, right? You don't want a Chinese company like Huawei to circumvent the Entity List by simply setting up a shell organization.

And so I think it was a perfectly reasonable rule to issue, but then China reacted to that very strongly. There's such a high chance of the two sides miscommunicating with each other and over-escalating when you just allow the bureaucracies to talk to each other.

I think what's most needed right now is to have a grand bargain between President Trump and President Xi. That will then calm down the relationship. Obviously, this is going to be a very competitive relationship for as far as I can see, because there's both security and economic competition going on. But I do think that when you have an understanding at the top, it creates conditions for more stability.

All right. Traders on Polymarket think—or know, depending on the situation—that we're going to see cooler heads prevail: a 15% chance of the 100% tariff on China going into effect by November 1. In other words, an 85% chance it's not going to happen. There's a 73% chance of a U.S.-China trade agreement by November 10.

Who do you think has the greatest incentive to decouple: the U.S. or China?

I think both sides understand that they can no longer be dependent on the other for things that they consider to be critical. The U.S. has realized, “Wait a second, we can't allow China to have a monopoly on rare earths.” By the way, it's not just the rare earths and the processing; it's also the casting of the rare earths into magnets.

There's this long tail of, I don't know, 20,000 different kinds of magnets that are made from rare earths and used in the industrial supply chain. If China were to cut those off, then it would greatly impede the manufacturing of pretty much every different kind of electric motor, from cars to other goods.

In any event, my point is just that the U.S. clearly cannot be dependent. I'm sure that China is reaching a similar conclusion. That doesn't mean trade can't happen between the 2 countries. I'm sure it will continue, but I don't think either side wants to be in a position where the other side can use its leverage beyond just economic terms and into geopolitics.

But with respect to the Belt and Road Initiative, don't you think China has been planning for decoupling in a very long-term way relative to the United States? They're more prepared and active in executing the decoupling, whereas the U.S. is still waking up to the fact that there are certain critical supply chain dependencies that we have with China, and maybe we shouldn't be decoupling as quickly as we've stated our intention to. It feels like they're in a pretty advantaged position, Sacks, on both the buy side and the sell side.

They have 20% youth unemployment. They have a demographic time bomb that's gone off. They have a pretty moribund real estate market. They have very little to no foreign direct investment. So I don't agree with you.

Okay.

Well, I would just say that I think what China has been doing—

Meaning, they're just as challenged as we are. They have different challenges.

Yeah. No, it's a good point. I would just say that I think China is a rising great power. Over the last decade, their economy grew to the point where it's roughly the size of the U.S. economy. Certainly, on a PPP basis, it might even be bigger—

Right?

But on an international currency basis, it's still smaller. Anyway, the point is that as China has risen as a great power, it has sought to create its own international institutions.

If you think about what the U.S. has, the U.S. has NATO, obviously, and the G7, the World Bank, the IMF, and all that stuff. What China has sought to do is create its own institutions. They created the Shanghai Cooperation Organization, which is a security-focused organization. I wouldn't call it quite the equivalent of NATO, but it's trending in that direction.

They've probably been the main sponsor behind BRICS, and the Belt and Road Initiative is part of that. They're creating their own network of relationships, their own international order, and it is a challenge to the U.S. order.

During the unipolar moment, from, say, 1991 to 2017, the U.S. was the only great power in the system, and we basically defined all the key international institutions. As China has risen, it has done what all great powers seek to do, which is create their own order.

So, yeah, I think they do want to be somewhat independent of the U.S. They don't want to be dependent on us, and I think we're realizing that we have to act the same way. The shame of it is that when we were running all the key institutions during the unipolar moment, like the WTO, we engaged in so many foolish policies that allowed China to dominate rare earths. We should never have let that happen.

Can I clarify one thing? I think a lot of people don't know this historical artifact, which I think is so fascinating: I don't think China is a rising power. I think China is a reascending power. The reason I say that is, if you go back from 1500 to now, over all of those years, China had the world's largest GDP 70% of the time.

So if you think about it in that context, especially as a Confucian society, I think the Chinese think about this as, “Hold on, this is reasserting the dominance that we've always had.” If you start to look at it through that lens, that's why they view the blending of public-private partnerships and the organizations that go and backstop private companies in China the way they do.

The other interesting thing, which I learned studying China, is how they allocate capital. The capital allocation system is both rigid and incredibly flexible. What do I mean? The way that China allocates money is the following: You have Xi Jinping and his close circle of people. They dictate the priorities.

I think there was an article, Nick—you can find it—but Xi is personally helping to draft China's 5-year plan right now. What happens is that it goes from that central group of 7 or 8 people to the Politburo. When it gets approved by the Politburo, it gets sent to all the provinces, and then it gets sent to the prefectures.

An example was in 2006 and 2007, when they decided to prioritize EVs. What you effectively have is a cascading set of venture capitalists operating against a national priority, and then they are rewarded for it. You essentially have 300 VCs running around organizing human capital in all of these different provinces against all of these different priorities. At the end of that, you get a BYD, a Xiaomi, or these great companies.

I think that's what's so daunting if you think about it. They have—

And they have one of those plans, Chamath, for every industry. They have an agriculture 5-year plan and an energy 5-year plan, and in each one it details their forecast—not just how to catch up, but what's possible. That's why they're charting new avenues of innovation in nuclear power.

It's about them, I think, getting back to where they have been, as opposed to realizing that they could potentially be the world's largest economy. It's more them saying, “We were the largest for so long. We took our eye off the ball, and now we need to get refocused.” I think that's more of a national-pride kind of motivation here. But it's obviously different people, a different country, a different governing structure, and so on.

Isn't that incredible, though? China had the largest GDP in the world 70% of the time since 1500.

It is incredible.

It's stunning.

Yeah. I mean, there were these innovations that drove some of these ships. The technology of ships and fleets of ships drove the Dutch Empire to its prowess and then the British Empire to its prowess. There were these moments of technological innovation.

But you're right: always on the heels was China. This may be China's moment of resurgence across the board.

The person who really saw this ahead of the curve was Lee Kuan Yew from Singapore, basically the founder of Singapore.

And Graham Allison wrote a book about Lee Kuan Yew where he talks about his insights, and there are a couple of quotes here that I think are really good on this. What Lee Kuan Yew said is that it's not possible to pretend that China is just another big player. This is the biggest player in the history of the world.

He also said that the size of China's displacement of the world balance is such that the world must find a new balance. Chamath, like you're saying, China was the biggest power in the world for hundreds of years, and then it was overtaken by the Western powers. Now it's reasserting itself.

In any event, this is why I think that during the unipolar moment, the U.S. should have been a lot more careful about facilitating China's rise.

Because we created a monster.

This is the point Professor John Mearsheimer has made many times. He wrote about this as far back as 2002, saying in his article “Can China Rise Peacefully?” that the answer is no. If we facilitate their rise, we will create a geopolitical competitor, and we'll go from being a unipolar world order to a multipolar world order. That's basically what's happened.

At the time, just remember, when China was admitted to the WTO, the motivation was really about lowering the cost of goods and services for the American consumer, which was an intrinsically reasonable motivation. It wasn't as though there was some nefarious intention at the time, but it was shortsighted in that it didn't do the calculus 20, 30, or 40 years forward to determine: If we then kneecap our ability to actually produce anything, where are we going to be if they do start to decouple?

It was hubris, though.

It was hubris because we thought there was only 1 great power in the system, and that’s the way it would always be. We thought great-power politics was off the table, and we thought we were at the end of history.

It wasn’t hubris; it was greed. The CEOs at the time and the people who were influential in making donations to politicians wanted to make more money. That’s why they did it. They didn’t do it with anything other than a profit motivation, which is greed.

Part of the American mindset at the time was that we had a monopoly on knowledge work, right? And this was key. This speaks to the hubris of the decision: knowledge work is where you create a separation of labor from intellect. There was this idea that America would move fully into a services economy and not have to be in a labor economy, and the labor economy could be outsourced. Ultimately, that didn’t prove true. The internet created a rival to knowledge work—a diffusion of knowledge and diffusion of knowledge capacity to societies around the world—which really did create a different structure than we anticipated at the time.

I think that’s looking in the rearview mirror, Friedberg. I think the CEOs at that time were asking, “How do we increase the profit margin on an iPhone, on a PC, on any product being created?” The way to do that is find the lowest—

Labor source possible. That was China. They just did it out of pure greed. So that was part of it, but there was another factor here as well. I do think the ideology was really powerful. Remember, in the 1990s, the leading philosopher or international-relations theorist—the most celebrated book—was The End of History and the Last Man by Francis Fukuyama.

What he basically said is that we’re at the end of history because liberal democracy has won, and it’s only a matter of time before the whole world becomes liberal democracies. The theory about China was that if they got rich, they would become more like us—that people would rise up and demand a liberal democratic government as they became rich. And so we then justified—and there probably was an element of greed to this as well—helping China rise on the grounds that somehow this wealth would turn them into a democracy.

In fact, all that it did was make them richer and more powerful. It didn’t change their government. It just turned them into a huge geopolitical competitor to the US. The author who actually predicted this—well, there’s John Mearsheimer, who predicted it in the early 2000s. Also, Samuel Huntington wrote a book called The Clash of Civilizations and the Remaking of World Order in the mid-’90s that was sort of the opposite of The End of History. The End of History was celebrated; The Clash of Civilizations was sort of reviled.

But what Huntington said is that what these other civilizations want around the world is not Westernization. They want modernization. And as they modernize and become richer, they’re not going to basically Westernize. They’re going to become competitors to the Western order. And that’s basically what happened.

I thought for sure you were going to say Thomas Friedman’s The World Is Flat, which was, I think, air cover for all this, because you would get the peace dividend if everybody had McDonald’s and everybody was watching the same movies. Go ahead, Chamath. You have some thoughts.

That Huntington book is actually exceptional. The thing with Francis Fukuyama is, he writes in this beautiful way, but he literally bats 0. It’s like you could basically read a Francis Fukuyama book and understand that whatever the opposite of what he wrote is likely going to happen. Anyway, that’s neither here nor there.

But I just wanted to pick on what Jason said, because I agree with him. The thing is, you had a lot of this geopolitical tension, Sacks, which I think you articulated well. The elites were coming together, like, “We’re going to convert China. We’re going to do all of this stuff.” And I think that was a trend, but then what hyperaccelerated it, in my opinion, was something that was very tactical, which Jay Cal mentioned: the compensation plans of most of the executives and CEOs of these public corporations started to get very attuned to gameable metrics like EPS in and around that same time.

And so Jason’s right: The direct financial incentives to basically think about this as a monocultural global world, where you can just export things to every nook and cranny of the world to optimize for the flow of profits back to you—trap them in any country you want, because it doesn’t ultimately matter to your compensation—that perversion also happened at the same time, which I think accelerated what otherwise would have given us maybe another decade or 2 to realize that we weren’t in complete control of our destiny.

So you can see it also in Taiwan. How in the world did 90% of the chips come out of 1 tiny island? It’s just capitalism. It’s the best place to do it at the cheapest price, most efficiently. And then you have this dependency on Taiwan because we didn’t think about redundancy in the supply chain.

Friedberg, in related international news, Trump—or the Trump administration—has put up about $40 billion to help out Argentina. Can you explain what’s going on there?

They’re trying to bolster the credit rating of Argentina with a currency swap. That’s related to the financial condition that the proponents would argue he inherited; the opponents would say he’s making things worse. The US is stepping in to support his free-market principles, which are ultimately going to drive cross-border cooperation. So I think that the swap agreement puts in place a bolstering of support.

I thought this was related directly to rare earths.

To rare earths? I’m not sure about rare earths. All right, Chamath. I saw you retweeting some of this, and there’s been a little bit of a back channel in the last 48 hours. Kind of breaking news here: Data centers are spiking the cost of energy, or people are perceiving them to be doing so, in different geographies here in the United States. And I guess, not wanting to be hated or get into wars with consumers, some big tech companies are standing down plans. Explain what’s going on here, Chamath, if this is a trend yet.

I put this out there when I wrote that because I do think that this is the beginning of a trend. To be very specific, Google was planning a $1 billion spend on a data center in Indianapolis. And essentially, what happened was there was enough pushback that it looked like, if it went to a vote in the city council, they would have voted against approval of the data center, against the necessary zoning and other regulatory approvals. So I think, to avoid what would have been an awkward press cycle, Google just pulled it and said, “Don’t worry, we’re not going to do this here.”

That in and of itself, I think, is an isolated incident, except that in this same week it happened 2 other times. In Wisconsin, there was a proposal very similar to Google’s, but this time from Microsoft, where again those local residents got pretty upset and, at the 11th hour, Microsoft pulled it. And then, in the third example, there was an attempt to get an Amazon data center built near Tucson. That, I think, was also mothballed.

So why is this happening? When I tweeted this out—or when I posted it, rather—the comments were pretty instructive. I think there are 3 things that I took away from it. Thing number 1 is that people are seeing their electricity prices go up in the local areas where the data centers are being built. The second is that people are worried about the water consumption of these data centers, the cost of water, and access to clean water, and that the plans to be net neutral on water consumption may or may not be real. So there’s a concern there.

And then the third thing that many of the people commented on was the actual noise pollution that the data centers were creating. But I took a step back from this, and my thought is very simple: It actually speaks to 1 of 3 very important inflection points that we’re dealing with on the AI side.

One of them we talked about last week, which is that you can’t have 50 state laws, right? It needs to come from the federal government so that there’s 1 set of regulation and there’s clarity. So effectively, what Sacks has been able to execute on the crypto side, I think his intention is to try to do this on the AI side, and I think the federal government needs to be given a chance to do that. Otherwise, locally in the United States, the only winners are going to be lawyers, and then we’re going to create an own goal and China’s going to win.

Second, we just talked about it: The government needs to be the economic backstop to the US supply chain that powers AI, right? We talked about all these things—the metals, the minerals, the rare earths—so that the rents that companies can extract are mostly market-driven and not exploitative, and don’t create air pockets in the funding cycle.

But here’s the thing that nobody talks about. When you look at these data centers, I think what it’s saying, bottom-up, is that these hyperscalers need to get these communities on their side. I think there is no understanding of what a token is. What is a token? I don’t think, if you ask the average person on the street, they know what it is. What they know is that tokens are not jobs.

If AI is going to be embraced by the public, I think what we need to do a better job of is showing the tangible economic benefits, especially in places like Wisconsin and Indiana that may feel like they’re missing out. We can’t have this narrative that AI may take your job away while it’s also doubling your electricity price, because they’re not feeling and seeing the abundance that we think it can offer.

This is exactly right, Chamath. If you think about the average American citizen, they have somebody in their family who drives for Uber or DoorDash. They have truck drivers, Amazon factory workers, and all those jobs are going away. It's just a matter of when. Some people believe we can handle the displacement; others don't. Go ahead, Friedberg.

My idea is that you have to use the balance sheet of these big companies. Use that free cash flow, and it can act as a cushion for a lot of what needs to happen.

So, they can pay higher tariffs for electricity.

They can also, frankly, go and pay for some amount of the electricity bill of these local folks. They could pay for solar and storage if they wanted to. There's a plethora of ideas here where they're probably going to need to step into the market to get folks on their side, and I think they have to do it.

It's interesting you mentioned that because I just had Zach Dell from Base Power on This Week in Startups last week. What they do is put batteries on the side of your house, outside. They're square ones, so they don't feel as industrial as industrial ones, and then they'll fill those batteries when the electricity is cheap. They'll deploy it when it's expensive. They get the arbitrage.

I said to him on the show, “They only charge consumers, I think, $500 or $1,000 to install it, just so they have some skin in the game. You don't have to rip your roof apart and put solar on. It's just batteries.” They said, “Well, why don't you do that and do an entire neighborhood? When they build these 3,000-home infill neighborhoods, why don't you put them on every single one and then create a network for data centers?” And he said, “Stand by. That's exactly what we're working on.”

So it kind of combines your idea. Sacks, you're the czar. What are your thoughts on what Chamath points out, which is the average working American and their perception of AI? You have job loss, electric bills going up, and then you have fake news. The reality is, is this real or not?

These Sora-slop videos I'm seeing make it feel like there's a bit of a trend toward a negative framing of AI innovation, and maybe it's not going to help the bottom half of society. Are you seeing that? Do you believe it? And do you have plans to counter it?

Well, what I see is that we just came off a quarter of 3.8% economic growth, of which 40% was attributable to AI. In other words, if we didn't have this AI boom going on, by my math, our GDP growth rate in Q2 would have been 2.3% instead of 3.8%.

AI is the difference between having great GDP growth, say, around 4%, and modest GDP growth around 2%. It seems to me that everyone who wants to have a great economy should be supportive of this AI boom that's happening.

The job-loss narrative is just that. It's a narrative. It's a story. The media loves to tell it. It's mostly theoretical and mostly anecdotal. You can point to a few jobs here and there, but there's no evidence of widespread job loss. This is just something that people are prognosticating about the future.

In every previous technological revolution that we've had, the jobs have shifted from more rote jobs to more sophisticated jobs that create a higher standard of living for everybody. The best example is that in the year 1900, 50% of the U.S. economy was involved in agriculture. In other words, 1 out of every 2 people were farmers. By the year 2000, that number had dropped to 2%. The difference—the 48%—did not become unemployed. They went off and did other things. They worked in factories, for example, and they went into the services part of the economy.

So I think what's going to happen here is that AI is going to enable people to shift their work to more gratifying and less rote parts of the economy that will increase productivity and standards of living for everybody.

Biology has a really good way of explaining this, and I think this goes to the heart of why the job-loss narrative is a fallacy: humans are end-to-end. What makes them really good and irreplaceable is that they can do an entire job end-to-end, and they can pivot their objective and figure out what their objective should be.

AI is not like that. It has to be told what to do. It has to be prompted, and then when it gives you a result, that result has to be validated to make sure it's not hallucinating. Frequently, you don't really get the result you're looking for, so you have to iterate and reprompt. You go through this loop where you do it over and over again until you get something that's valuable.

In other words, humans do the things at the end. They do the prompting and validation. AI does the stuff in the middle. As biology says, humans are end-to-end; AI is middle-to-middle.

This is why I fundamentally think that AI and humans will be very synergistic. AI will not be a replacement for humans. It's going to do the stuff in the middle that humans don't like to do, and it's going to allow humans to be much more productive.

Right now, I'm very optimistic about this boom. I do think that people who are criticizing it should really think about whether they want the U.S. to have 2% growth rates or 4% growth rates, because that's what we're really talking about here.

Friedberg, the CEOs of Uber and Tesla believe the opposite. They think these driver jobs are going away pretty quickly. The data is showing there's a serious problem right now with young people getting jobs specifically because of AI.

Obviously, GDP is not wages. The people we're talking about who are affected by these job losses don't own equities. The top 50% to 55% of the country own equities. What do you think about this narrative? Do you believe it's a false narrative, a true narrative, or somewhere in between?

Here's how I would think about what may happen, rather than think about someone, quote, losing their job, which is a term you keep using and others keep using.

I think it would be, yeah.

Whatever it is, you're acting like they're passively in a victim state. We have the lowest unemployment in history, and there's an extraordinary demand for new capacity of work by humans to do stuff in the United States right now because of the extraordinary manufacturing build-out that is underway, driven largely by AI and soon-to-be robotics and several other industries.

We just talked about rare-earths mining and processing. There are many industries that are taking off in this country that have not existed, and there is a demand for labor. When there's a demand for labor and there's a shortage of people to do the work, the employer typically says, “Let's go recruit. Let's go spend. Let's go train.”

When you have an innovation cycle like the one underway right now, the hiring for the new jobs happens before the loss of the old jobs happens. You have to have someone build all the Model Ts before people stop becoming horse-and-buggy drivers. Who's going to do all the Model T building? They're going to get paid. They're going to be offered money to go do those jobs.

What typically happens in these cycles is that there's a recruitment exercise, a recruiting cycle, that happens up front. Rather than say all these drivers are going to lose their jobs, another way to frame it is that all of these drivers are going to be offered new, higher-paying jobs before there's lower demand for driving.

They're going to say, “I can go make twice as much money per hour by going over to take this new job.” Then they're going to say, “I'd rather do that than drive. I'm getting paid twice as much, and I get to work on more interesting stuff. I like that.”

That's how this is likely going to go. That is the more positive view on how what you're calling job displacement actually plays out in the U.S. economy in the decade ahead. A recruiting cycle precedes the elimination of old jobs that aren't needed anymore. That recruiting cycle is a net positive for people.

They say, “Wow, that is an amazing improvement in my income, an improvement in my lifestyle. I get to work on more interesting things. My life is better because of it.” Years later, the demand for those old jobs starts to die down, but at that point, everyone's been recruited away anyway.

So, Friedberg, you mentioned the Model T putting all the horse-and-buggy drivers out of business. If we had today's media environment back in the days of Henry Ford, that's exactly what they would have been bemoaning: that all the horse-and-buggy guys were going to be put out of business.

Part of the reason why that would have been a very bad take is because it's easy to see the people who are being displaced in the short term, like maybe the Uber drivers or something like that. But it's hard to see all the new industries, all the new businesses, and all the new jobs that don't even exist yet that are going to be created.

For example, what were the downstream effects of the Model T? First of all, it created this new kind of industrial assembly line, which applied not just to cars but to all sorts of goods. Henry Ford pioneered that, and it had huge ramifications for the U.S. becoming the industrial power that it ultimately became, which led to its winning World War II.

Then, just in terms of the car, you had things like the build-out of the interstate highway system.

The tire industry.

Yeah, just think about the highways that led to—

Holiday Inn, hotels and motels, drive-through windows, McDonald's, and car culture—

And—

And the taxi cab industry itself emerged.

George Lucas got his start as a director by making the film American Graffiti, which was all about car culture in the 1950s. So that led to Star Wars. There are all these butterfly effects of the Model T being invented, which are absolutely massive.

And it would have been impossible to foresee all of that stuff 100 years ago. When you have these new industrial revolutions or these new technological revolutions, it's going to increase the capacity of humans to design all sorts of new products and create all sorts of new businesses. What we've seen is that this lump-of-labor fallacy—that you have all these humans that we don't know what to do with, so you have to just keep them busy—has always proven to be a fallacy throughout human history.

And by the way, those buggy drivers did not lose their jobs. They chose to go into 1 of these new industries where they were offered more money.

I don't think this is the issue at all. I would say that when you look at Arizona, Indiana, and Wisconsin, what you are looking at are people whose jobs are not at risk today, tomorrow, or the next day because of AI. I don't think that's what it is, and I think that they probably, on the margins, benefit. The question is more how they're getting information and how they're processing their point of view so that they show up to these local municipality meetings and essentially cancel progress.

That is a very important thing to diagnose because it is not isolated. I'm trying to ask the question, which I think is a more practical and important one: What do we do to make sure that the narrative and the facts are clear? Right now, you're seeing very different and diverse pockets of people in largely purple and then some red states voting against the exact things we need to bring this abundance forward. I think we need to figure out why and how that's happening and have a version that counteracts that. Otherwise, this will become more slowed. It will not go faster.

Well, you bring up a good point there. There are, I would say, legitimate and illegitimate reasons to be skeptical of AI. The one that you bring up, Chamath, that I think is very legitimate is that we're going to have to stand up a lot more electric power generation. Otherwise, it could drive residential rates up, and that will create a backlash against AI and the build-out of data centers, and we don't want that. So that's a legitimate issue that has to be solved.

However, there are what I would call illegitimate reasons to fear AI. That's got everything to do with the media environment. They're spreading all these unproven narratives, in many cases based on contrived studies, where they're saying that there's going to be 50% job loss in the next few years. You have all the Terminator-type scenarios where AI is going to grow beyond our control—all of these doomer narratives that are being created that are scaring people and creating a knee-jerk reaction at the state level.

We talked about this last week, where there's now 1,000 bills going through state legislatures, mostly because the legislators feel a need to, quote-unquote, do something, not because they know what they want to do. So there's a lot of unnecessary fear out there that's been created by the media and then by organizations and companies who have a regulatory-capture agenda around this.

I hear you. What I'm saying is we need to figure out how to fix this, because you can't have Google, Microsoft, and Amazon cancel essential projects that advance this entire sector and economy forward. We're not talking about some random fly-by-night business here. We're talking about 3 of the 5 principal horsemen of the AI race that we are fielding on the field.

Well, you're looking for suggestions. Do you have any off the top of your head? Do you think it's in education? Do you think it's in explaining the value of AI and how those people in those states would benefit from it, or do you think it's just simple economics—lowering their electricity bill?

I think there are 2 things. The people in Indiana and Wisconsin said, again, electricity prices, water, and noise. I think that we can fix those problems, and I think we—we meaning our industry, and specifically the hyperscalers—need to start using these gobs of free cash that today public markets don't value anyway. It's not like Google, Amazon, and Facebook get credit for their cash. In fact, Facebook's stock price went up when they basically burned through all their cash.

If you really want to take the capex around AI and make it a true supercycle, there's another avenue for where capital can be deployed. I would encourage the CEOs of these companies to wake up, get their act together, and do it. That's number 1.

The second thing is, I think we need to have better spokesmen for our industry. Every time you see somebody talking about all these navel-gazing technologies and then unable to translate them into simple English and product value, it becomes a problem. For example, this week, on the 1 side, we have Anthropic AI doomerism that Sacks had to confront, but on the other side, there were all these people who got freaked out because OpenAI wants to push toward erotica, and all they could think of was the impact on their children.

Again, this is not necessarily true or not true, but the point is we need to have better, more eloquent, reasonable spokesmen who can articulate a vision that everybody can buy into. I said this on X: What are Arabella and Soros going to get behind now to find a bogeyman? You don't have Israel. You're not going to have Ukraine soon. So what are you going to find as a bogeyman? Unfortunately, it's sitting right in our backyard.

Yeah. I'm not sure Altman is the greatest spokesperson for the industry. Just to clean up a little bit here, Friedberg, my position isn't doomerism. I just think displacement is going to be more significant than we think, and the evidence I have of that is just watching what's happening with young people applying for jobs and the increase in unemployment there.

Then look at the people who deploy AI at the highest velocity, which are obviously tech companies. If you look at Alphabet, they had peak Googlers in 2022: 190,000. They've had massive earnings and profit increases and have 187,000 people now—in other words, less.

Then you do the same for Meta, Uber, and Amazon. Meta peaked at 86,000 employees and is now at 75,000. Uber peaked at 33,000 and is now at 31,000. And Amazon, where you'd think you'd have a lot of growth here, peaked at 1.6 million and is now down to 1.55 million.

These companies are not hiring. The gains they're talking about publicly in the deployment of AI are significant, and they all have aggressive plans to use automation—specifically in robots, self-driving, and, obviously, knowledge work—to reduce headcount and increase profits. I think that's going to become the narrative that's going to be very challenging for our industry.

If you look at our own companies, they're saying, “Wait a second, your companies aren't growing, and my kid got a computer science degree.” If you look at the unemployment level of people graduating with computer science degrees who are developers, it's increasing. Unemployment among developers is increasing, and that's just really notable.

So I wouldn't say I'm a doomer, but I am concerned that the new jobs might not show up as quickly as the old jobs, or that the static team-size trend continues, because we've had static team sizes for years now.