Epstein Files Flop, State of the Market, Autonomous Robots, Trump's Gold Card, Friedberg on Jeopardy

Our incredible comedian and celebrity guest got sick at the last minute today and didn't make it. We won't say who it is, but when he comes on this show, you are going to laugh your ass off because he's awesome. I like the comedians. I think their takes on society and culture are pretty interesting.

Do you think it will work in our format? What's your prediction here? We've never done it. I think the quality of the show is best when it's less about people just doing takes and more about the back-and-forth banter.

That's what I'm always trying to do: get the ball to go around the horn and get some real dialogue going here. Sometimes people feel passionately with me.

Again this week on the All-In podcast: David Friedberg and Chamath Palihapitiya.

My name is Jason Calacanis. You can follow me on X at @Jason. He's Chamath, and he's—why can't I make fun of you in the replies on Twitter anymore?

Because I applied it to you.

No, what I did was, my replies were so full of MAGA lunatics and crypto scams that I had no choice. I had 500 replies to every single one, so I put it on subscription and said, “If you want a reply, it's $3 a month.”

1,500 lunatics signed up.

So I did it just so you can troll me, and I thank you for the $3 every month.

It's all going to charity.

Let your winners ride, Rain Man David. Instead, we open-sourced it to the fans, and they've just gone crazy with it. Love you.

Do you want an intro to the show or some warm-up on Jeopardy, or do you want to go right into Epstein? The letter from Pam Bondi to Kash Patel is crazy.

“Dear Director Patel, before you came into office, I requested the full and complete files related to Jeffrey Epstein. In response to this request, I received approximately 200 pages of documents. Late yesterday, I learned from a source that the FBI field office in New York was in possession of thousands of pages of documents related to the investigation and indictment of Epstein. Despite my repeated requests, the FBI never disclosed the existence of these files.

“When you and I spoke yesterday, you were just as surprised as I was to learn this new information. By 8:00 a.m. tomorrow, February 28, the FBI will deliver the full and complete Epstein files to my office. Regardless of how such information was obtained, there will be no withholding or limitations to my or your access.”

My question to you guys is, do you think this is much ado about nothing—that the FBI needs to have the discretion to say no—or do you think this is one of those things where they're not allowed to do what they're doing?

I think it's above my pay grade. I don't know the law around FBI investigations. What if they investigated a bunch of people who were not guilty, and then those people were in the files? Maybe they need to look at them before they do a document dump. Maybe there's an informant in there. I'm trying to think of what happened with the whistleblower papers.

This says something different. This doesn't say, “Let's negotiate what we should release together so that we protect people.” This says, “We're just going to lie to you and tell you”—here's her side of the story. Maybe they have another side. Maybe the FBI has a different side. We have to hear from them, right?

Is this just because people are intrigued by the gossip angle of it? Is it because they want to prosecute people for hurting people? Or is it because they want to cancel people, and this is a nice opportunity to cancel? What's happening?

Interesting question, because this thing has been going on for 20 years. I think this actually has more to do with conspiracy. Is there a deep-state cover-up? That's the question.

I think the question is, if the chain of command requests something, are you allowed to withhold it because you decide, in your own judgment, that the person above you doesn't deserve to know it? I think that's an important question.

The thing about the FBI is that it can withhold information from the president and from other folks if there's an ongoing investigation. Sources are secret and could be jeopardized, and there can be national-security concerns. I'm reading here, so I'm very excited to see these next 2 days unfold.

Listen, it's a breaking-news story. We'll have more to say about it next week when we get the facts.

It has been an amazing 24 hours for my guy David Friedberg. I am so proud of you. If people don't know, David Friedberg was on Celebrity Jeopardy! this week, and he had a great performance.

You can't say that.

I can't say it? Bleep it out. Bleep it out.

I was on Celebrity Jeopardy!, which shows you how much of a bar there is. The bar was pretty low. They're now inviting podcasters, apparently.

Jeopardy! is watched by about 10 million people every episode, right?

They have a huge audience on regular Jeopardy! I think it's 9 million a night or something, maybe 7 to 9 million a night.

Celebrity Jeopardy! has a smaller audience because it's later, but it's still a couple million people. So you were on, and for the last 4 or 5 months, we've all had to bite our tongues. You've been in a full-scale—I don't want to say panic—but you've been wringing your hands about your performance.

Spoiler alert: Friedberg won. Not only did he win, he crushed it. I was watching this, and it was better than watching the World Series of Poker or a Knicks–Warriors game for me.

Here's my favorite moment: the All-In call, as a tribute. He gets the Daily Double, and I'm going to go all in.

Okay. How dare you?

$12,800 for Dave. But you have to be correct. “In African geography, known for its snows, this Tanzanian peak is both Africa's highest and the world's tallest freestanding mountain. What is Mount Kilimanjaro?”

That is correct. $12,800.

What did you think of that moment, Chamath? He goes all in, and these people think they're going to get Friedberg by giving him a question about Africa, not knowing that he's African-American.

Honestly, a 6-year-old should know that category.

A 6-year-old should know that category?

You guys want to hear something crazy? So, 6-year-olds should know that category?

The night before, I was at dinner with our friend Xander and his wife. Their son was taking a quiz on African geography, and they started giving me all the questions. We were actually quizzing at dinner the night before on African geography. It was like a Slumdog Millionaire moment.

I was thinking, “There's no way that's real.” We were literally just talking about African geography at dinner the night before. I thought, “The judges had to have been listening or something.”

But you were nervous. You talked about being nervous and your techniques. Our friend Jason Koon—the serious professional poker player who comes to the game—coached you a bit. What was his advice to you?

I was a little wound up because I've watched Jeopardy! my whole life. I don't watch it regularly, but a couple of months ago I was watching the show and reached out, saying, “Hey, do you guys know anyone on the show?” That's how I got hooked up.

Then I'm thinking, “Wait, I'm really going on? That's awesome.” They said, “You're going on Celebrity Jeopardy!” So I watched the Celebrity Jeopardy! episodes from last season and figured I could get 60% of the answers.

But you get there and don't realize how hard it is to buzz in on time. If you buzz before the light turns on, I've heard this over and over—

How does it work? Because you bought an actual buzzer to practice with, am I correct?

I bought this cheap buzzer on Amazon. It was impressive, but it didn't do anything. I was watching the television at home, watching old Celebrity Jeopardy! episodes from last year, thinking, “Buzz in, buzz in.”

The problem is that when you're there, you're not allowed to buzz until the light comes on. If you buzz too early, you get locked out for a quarter of a second, and that quarter of a second makes a huge difference.

I was actually behind the people I was against for most of the show. I felt like I knew the answer, but I kept missing. I still did okay, obviously.

Wait a second. Let me ask you a question about that buzzing thing. A light goes on—when does the light go on? When the question is finished, or after they show the whole question?

They show the question on a huge screen, so the whole screen gives you the question right away. You can read ahead and read the whole thing. Then you have to wait for Ken Jennings to finish.

As soon as he finishes enunciating the last syllable of the last word, the light turns on. But here's what I found out: there is a delay of about 150 to 200 milliseconds between your eyes and your ears. You actually hear stuff before you see it, which is really interesting for your brain to register. People have different delays, but for me, I'm waiting for the light and then I buzz in, and it's too late because the people next to me have already buzzed in.

Speed of sound versus—

Anyway, I was a little wound up going into this. I called Jason because I realized there's no upside. By the time I show up, I'm thinking, “I'm going to look like an absolute idiot for saying some stupid stuff and getting answers wrong.”

Of course, everyone texted me last night saying, “How did you miss Hoosiers? How did you miss this?”

Since you're bringing up the big miss—the Daily Double.

Oh, man.

You get the Daily Double: $16,000. Crushing, crushing, crushing their souls.

It wasn't my best category.

“$4,000.” Looked great. Even $20,000 if you're right. “Climactic moments in sports movies: Jimmy Chitwood buries a jumper from the top of the key to win the Hickory Huskers the Indiana State Championship.”

“What is Hoosiers?” Sorry. No. What am I doing?

You were right. I'm so embarrassed watching it. Come on, dude.

I know Hoosiers. I mean, come on. Gene Hackman—he died today, and no one sent me a text saying, “I can't believe you got that answer. Great job.” Everyone just sends the text, “How did you not get Hoosiers? How did you not get Field of Dreams?”

It's always, “I knew the answer.”

I realize it's different. As I'm walking into Jeopardy!, I'm thinking, “There's no upside, because everyone will just call you an idiot for the things you miss.”

I mean, let's face it: you're the Sultan of Science. You went to MIT and worked at Google, so there's an expectation that you should run these people over.

I did run them over.

For the most part. Who are you facing in the semifinal?

We don't know until all the quarterfinals are over. That'll be in the next couple of weeks.

When you see a category come up like the Hoosiers one—I think that one was about sports movies, right?

Yeah, it was sports movies, like big sports movies.

Don't you see “sports movies” and catalog probable answers in your head, like Rudy and Field of Dreams? Did you have all those ready to go?

I'm not a guy who gets very nervous. I don't get stage fright or wound up, but Jeopardy! was so weird. You're up there, and you can't focus or concentrate like you normally can.

My brain had these weird brain farts where I knew the answer, but it wasn't coming out. Or I buzzed in and said “Beethoven” instead of “Deion.” I'm thinking, “Why did that come out?”

God, for “Claudia.”

You don't have to—

I mean, dude, don't get me started.

There is weird stuff that happens up there that's really hard to explain. Then you're angry about the buzzing because you can't buzz in on time, and you're on the set of Jeopardy! thinking, “Oh my God, I'm actually on the set of Jeopardy!” You're watching the scores, and it's very overwhelming.

Was there an audience?

Yes, there was an audience—about 100 people.

Wow. That's nice.

Before you go on the show, they have you do a practice round in front of the audience so everyone gets used to it. I deliberately answered incorrectly and acted like an idiot. I acted like I couldn't buzz in and didn't know the answers. I tried to be a little—

Diabolical?

I tried to be a little diabolical.

Sharp elbows.

Then I came out swinging. I felt like I had to get aggressive out of the gate.

I've done that before in a bar fight. I was like, “Guys, we don't need to get in a fight,” and then—bang—I clocked the guy.

Anyway, we're so proud of you. You won; you trounced them. I have to say, the money went to the Humane Society of the United States. How much did you make for them in the end?

If you win the whole tournament, it's $1 million. The winner's charity gets more, and the losers' charities get less, but it's a fixed amount. It's something like $30,000 or $50,000.

Nice. I think they should put the 3 of us on together in a normal episode. That would be crazy.

Oh my God. If the 3 of us were on, I would wreck you.

No, you would not. I will play you heads-up Jeopardy! anytime for money.

Let's play. I say we play poker. We'll play for all of the All-In profits. I'll give you 2 times the chip stack and let's see what happens.

Oh my God. That sounds so compelling. Would you do it, Sacks? Would you do it?

Would you do it?

All the profits?

All the profits. I'd do it.

But I would carve out 25% of the profits for a 4-way tournament that was televised live. If we could get $2 million in sponsors, then we'd be up no matter what.

Grift endless.

Not grifting. I'm thinking like a business.

The whole point is—

I want to do this so that I inflict pain on one of you, or both.

What gives Chamath happiness is hurting others. That's his love language.

His love language is hurting the people who love him.

I like putting you two to the test. I like seeing you two break.

Okay. This makes him feel good.

Did you guys see Brett Adcock's tweet?

No. What did he say?

Friend of the pod Brett Adcock, who's the CEO and founder of Figure, announced today that he's moved his timelines up by 2 years. He's going to beta-test robots in the home by the middle to end of this year.

That's crazy. Crazy, crazy, crazy.

Are you an investor in his company?

I don't talk about my investments, but in this case, no, I'm not an investor, so we're not talking our book here.

I do think Optimus, Figure, and the other dozen or so companies doing this are incredible. If they can make these for, what, $20,000, when do you think it becomes something a middle-class, dual-income household would buy? 5 years from now? 10 years from now?

I think the issue is bounded by 2 things. One is that I'm not sure generalized AI is good enough yet. Brett had a deal with OpenAI, which he pretty publicly canceled a few weeks ago, and he announced his own model. I don't know the details well enough to know whether he rolled it himself or whether it's just an open-source-based model that he's iterating from, but I think the model isn't perfect enough yet to be general-purpose.

The second is a practical issue with the robots. The actuators themselves are good, but they're not great. You can see it in the demo. It's an incredible demo because it shows the value and power of the model, but there's this master-slave orientation that has to happen, where one model is doing most of the computation and the second model—and the second robot—is feeding off of it.

The demo they do, Nick, you can probably find the video, is of them sorting a bag of groceries for the first time totally unsupervised.

Totally unsupervised. It's an incredibly cool demo.

It is a cool demo. The thing you notice, though, is that the actuators are good, but they're not great. The physical dexterity is still relatively limited, and I think that doesn't allow these robots to be superfunctional in the next couple of years.

When they get that figured out, then I think it could be really useful. If you have a robot like this that could sort the groceries, make food, do the laundry, and mow the lawn, it just requires a level of dexterity that's not yet totally possible.

But what you're seeing in this example is the 2 robots basically figuring out how to communicate semantically between one another. That's incredibly powerful, and it's yet another breakthrough that we need.

I don't know. We're probably a couple of years away.

Look at the dexterity there. He's taking the peppercorns—peppercorns?

Please don't misgender the robot.

He's crushing the peppercorns, or whatever you call this robot. It's really incredible. They're figuring out how to communicate. The coolest part of this demo, which I loved, was that they take an apple, and then the second robot figures out that it should go in the fruit bowl, pushes the fruit bowl to the first robot, and then the first robot—

That's cool. There it is. That level of semantic awareness and understanding between 2 models working interdependently is very cool. They're collaborating with each other.

That makes total sense. I can tell you, here on the ranch, I would love to have an all-purpose robot going out there and using the weed whacker, trimming the bushes and hedges, getting me wood, and collecting chicken eggs. There are a million things they could do on a ranch. It would be immediately applicable for ranch work.

If they're working 24 hours a day, it doesn't matter if they go slow. I think this is the category people are sleeping on. I don't know who on our prediction show said this would be the year of robots, but it's been the year of robots for 30 years in the industry. It does feel like this is it.

Friedberg, you're sticking with your prediction, I assume?

Yeah, I am. It's not just this kind of dexterous automation. I also put drones and autonomous vehicles in the same category. There's some combination of mechanical response to a machine-vision system that I think has become accelerated this year.

You need a lot of rare earths to make robots. Where could we ever get those from? Does anybody owe us a little money? Is anybody behind on their payments? Maybe the vig could be a little taste.

I don't know if you guys have seen this, but there was a company in the '90s that was all the rage called Segway. They were going to absolutely change cities and everything, and they never did. It was basically a scooter you could stand on with a balancing system.

Now they're making these lawnmowers. These lawnmowers are really cheap—$1,000—and they work here in Austin. I've seen 2 or 3 of them on people's lawns. This could be like the one you had in your house, the Roomba.

Your point, Friedberg, is that there will be purpose-driven ones to deliver you a burrito, do your lawn, and so on. A Roomba is $300 or $400, I think, and this thing is $1,000. This is going to get crazy.

I do think it's a lot harder to create one of these general-purpose systems in automation than to create a vertical, utility-specific automation system. A device that just does one thing—delivers something to you in the air, drives your food to you, or loads and unloads your dishes—is a lot more straightforward.

I'm not sure the whole idea of the humanoid is the right path. It's ambitious. It is a general-purpose device, and that makes for a technically very hard road map.

I have to imagine some of the bulldozers out there are also becoming remote-controlled. You can get a bulldozer that you don't have to put in a dangerous situation. It's remote-controlled, and it's going to have AI.

I got pitched on a startup one time that would go up into the Tahoe hills and allow humans with remote controls to drive small bulldozers and make fire paths. Imagine a fire breaks out: you send in, or helicopter in, the bulldozer. There's no human in it. It just has a 5G connection or a Starlink connection, and zip, zip, zip, you're doing fire roads in the middle of a smoke-dense area.

It's going to be really interesting when these things get dialed in. They're getting better every day. I'm going to ask Brett to be a beta tester of one of these robots in my house, and then we'll do a segment.

That would be great.

Excuse me, robot, can I get some morels, tout de suite?

Let's talk about Stripe. I thought the report was really good. We had the Collison brothers on last week; they crushed it. Great job to them.

Here's a quick summary. In terms of processing volume, Adyen did $1.34 trillion and Stripe did $1.4 trillion. That's incredible—they're both in almost exactly the same space. Adyen is growing 33%; Stripe is growing 38%. Adyen's valuation is $56 billion and it's public. Stripe is private, at $91.5 billion, so I guess there's a private-market premium.

Employee count is very interesting, especially since we've been talking about Jamie Dimon's rant last week: Adyen has 4,300 employees, and Stripe has over 8,000. Both are profitable. Adyen has $1 billion in EBITDA, which is extraordinary, but Stripe has a higher margin.

Chamath, what's your take on the tale of 2 cities here?

I thought there were 3 takeaways. The first is that the value of Stripe's ecosystem is probably underappreciated. I think Patrick mentioned it, but he said it almost as a passing fact, and none of us picked up on it. In the report, they talk about all the additional products they're able to build around core payments. One of them is the billing product, which is half a billion dollars a year of ARR. That's incredible.

If they figure out network effects inside the Stripe ecosystem, that's interesting. The hub-and-spoke model of payments being at the center, with all of these incremental services around it, is really interesting and underappreciated for Stripe. That probably speaks to why there's such a difference in valuation, because Adyen has less of that ecosystem, or at least it's not nearly as well described as Stripe's.

Second, I go back to what I've been saying for a while: the rise of these stablecoins is really interesting. The stablecoin infrastructure globally and the push for a bunch of national governments to embrace them—inside India, inside Brazil, and slowly inside the United States—is important.

The third takeaway, Nick, is the tweet I sent you about the nature of the AI ecosystem relative to the rest of SaaS. It came from Stripe's report, which showed the time it takes to get to $5 million of annualized revenue. The average SaaS company took 37 months. By 2024, the top 100 AI companies got there in 24 months.

That's efficiency in the market, right? That's why we're all looking at AI and saying we could see a lot of our economic issues come from growth. The growth is very clear: you can do more with less, and you can generate more revenue with AI. The trend is pretty clear.

I can give you a little factoid from 8VC. We got to $5 million of revenue in 3 months.

Really crazy. A couple of big ones in there.

It's just a very different selling motion from what I've historically seen. The ROI is so obvious in terms of the efficiency it creates and the cost savings you can generate relative to traditional enterprise software. It's a more straightforward sale. The ROI is clearer, the revenue is bigger, and it happens faster.

There's also a sense of urgency in the market right now. People feel like they have to adopt this new technology fast because of competition, because the gains are so clear, and because in a slowing economy this may be a way to accelerate revenue.

I'll be honest with you: at least with our 8VC customers, I haven't seen that yet. We're in 8 or 9 segments of the economy—big segments of the economy. It's more about the frustration they have with what I would call the software-industrial complex.

You can see it with what's happening to Salesforce and other big companies. These renewal cycles are getting harder and harder to justify, so people are willing to take some bets and see if there are different ways of dealing with the problem.

I think the real opportunity is if you can find a repeatable pattern to help these companies replace the big software spend they already have.

That scales really quickly. The only way to do that is by using AI in 2 ways. One is AI inside the machinery of what you're using yourself to make things. Those are products like Cursor, which fully accelerate development.

The second is AI within a very specific product that customers actually need and that also creates efficiency. There are 2 different places to use it.

The problem with using it in both places is that, in the first one, you can manage the errors. It's very straightforward: code either compiles or it doesn't. Even if you're using something like Cursor, which is an incredible product, there are no errors at the end because the thing either works or it doesn't.

The problem is when you use these models in actual work. If you're in a regulated environment, it gets very complicated. If you generate a hallucination in a health care business and it causes a patient record to be incorrect, there are huge consequences.

That exists in regulated finance, real estate and construction, power, and aerospace. Imagine if an LLM helps you design a better plane, but there's a tolerance error that's not well understood. That could have horrific consequences downstream.

We're working with all these people to try to figure it out. It's a very difficult technical challenge, but I thought the Stripe data was really interesting because it validates what we're seeing. Growth in this industry is unlike anything I've ever seen before.

Back to the stablecoins, here's a look at Tether. They're at $143 billion in Tether outstanding. Who knows what's reality there? They've got a shaky history. Then USDC—Jeremy Allaire is at $56 billion already, and that's only been in existence since, really, 2021.

I do think Stripe's main business could be sitting here in 5 years, Chamath, with $300 billion and getting 3%, 4%, or 5% on some coupon. They could be making $10 billion to $20 billion in pure profit if they have a stablecoin that's widely adopted.

I think the best way for Stripe to do this is to build it and facilitate payments between existing Stripe customers. As I said last week, these are all ultimately ledger entries. The more you can commoditize them into a simple ledger entry inside 2 systems of record at 2 companies, the better the product feature.

Stripe has the scale to do that now, and to your point, it could have an enormous stablecoin business. At the same time, it's probably better off embracing what's already been built. It may be disruptive to try to launch yet another one.

They bought that other company, so I have to think they'll launch their own.

Bridge is the facilitation layer. It's the rails.

But this is where brand comes in. If you have a trusted brand among developers, and there are 3 choices, are they going to take Tether? People say, “Maybe it's a little sordid. It's offshore. I have some challenges there.” Am I going to use USDC? “Okay, I haven't heard of it, but they sound interesting.” Or am I going to use Stripe? I'm going to go right to Stripe.

It's kind of like the IBM or Microsoft of payments. Nobody gets fired for picking Stripe.

Not anymore.

Not anymore, right?

Let's go through the market update. A lot of people are trying to figure out whether we're going to have a market collapse or a boom. Let's look at some numbers and have a first-principles discussion.

The S&P is up almost 2% so far this year. The Nasdaq 100 is flat. The Dow is up 3%, so it's a pretty good start to the year in those index numbers.

But if you look at the Magnificent 7, some of them have had serious compression. Tesla is down 27%. I do think they had a big Trump–Elon spike. Google is down 10%, Amazon 9%, and Microsoft about 8%. Meta, Apple, and Nvidia are up to varying degrees.

Coming into our taping this week, Bitcoin is down 15% over the last month. That also got the Trump bump.

Adding to all this confusion, unemployment is still at historic lows—close to 4%. If you look at the deportations that were promised, they've been modest to start. Obviously, they're just getting started and need some money to deport people, but they've only been deporting 500 to 1,000 people a day. We haven't heard many numbers about the last couple of weeks, as DOGE has been the center of attention.

They'd have to get to 2,000 or 3,000 people a day to have low millions—let's say 2 million or 3 million people deported—for it to have any impact on unemployment.

Finally, CPI is up 3% year over year. We had gotten down to that nice 2% handle, and now it's back up a little bit—not insignificantly. In September, it bottomed out around 2.4%, just in time for the election. That same month, the Fed cut 50 basis points. Then it cut another 25 in December, and since then inflation has been growing modestly but steadily.

Put it all together, Chamath. What do you think? Then we'll go to Friedberg.

I tend to be in the Steve Cohen camp. It's not like the bottom is going to fall out, but there's a lot of room for concern.

When you take a brew of tariffs on top of that, we have slowing immigration, and now you have DOGE. That's austerity. We think growth is going to slow to 1.5% from 2.5% in the second half.

I'm actually pretty negative for the first time in a while. It may only last a year or so, but I think the best gains have been had. It wouldn't surprise me to see a significant correction.

Nick, I don't know if you can find that clip, but he had a very precise summary of how he saw the world, and I frankly agreed with everything he was saying. He can probably say it better than I can.

I have a couple of very specific thoughts. The first is that you're starting to see compression of the Magnificent 7 toward everybody else. This is the forward P/E of those companies, and what you're starting to see is everybody else capturing some of the ground back. People are processing what the real upside of the Magnificent 7 is.

If you look at the other chart, though, the Magnificent 7 is really priced to perfection. You have to believe the world stays the way it is; otherwise, you're going to have some amount of mean reversion. I think the stock market is a little expensive on the margin and not particularly attractive.

Second, the bond market has basically said, “We're going to give you credit that DOGE is going to work and that tariffs are going to work.” We've had meaningful compression in the 10-year yield, which is really interesting. I think it's very good for Bessent and Trump.

We've got to refinance $10 trillion in the next 6 months, so you could see this thing maybe get under 4% if we get a good string of data.

The real problem, though, is that if you look back and ask what this resembles, the example I would give you is the United Kingdom in 2010. The deficit as a percentage of GDP was 10%, and the UK government embarked on a multiyear austerity plan. They said they were going to get the deficit as a percentage of GDP back in line. By 2016, it had gotten to 3%, which is where we're trying to get. Right now, we're a little under 7%, and we're trying to get to 3%.

It's interesting to ask what happened there. The bond market gave the UK government a ton of credit, so it kept rates relatively low and brought them back down.

That's what's happening here. The stock market has gone sideways to a little bit down. Let's see what happens. But the big thing is that in the UK, all of this created tremendous dissatisfaction, and you had Brexit.

The question I have is: if we go through a prolonged austerity program and frustration among the American population builds, what's the release valve? The release valve there was voting to leave the EU. Here, it's not obvious to me what the release valve is.

Electing Trump was step 1. I don't know if there's something even more populist than Trump, other than—I think he is the mechanism for implementing the austerity.

I think people want this austerity. The question is what happens when the actual byproducts of that austerity are felt by people for 6 or 7 years. I don't know what the answer is.

Certainly, people are in favor of DOGE and downsizing the government more than I think anybody anticipated. The statistics and polls are showing that it's incredibly popular.

Friedberg, when you look at this—not conflicting, exactly, but a lot of different data about what's going on—what do you see in the numbers? What does your instinct tell you? Part of this is getting used to Trump again. He says a lot of things. Some of them are scary, some are trolling, and everything in between.

The big question in Trump's actions is around tariffs versus the tax cuts that are being proposed versus the spending cuts. Those are the 3 levers. There's a very serious sensitivity to the economic outlook for growth and inflation based on how far each of those levers is pulled and how they relate to one another.

Is Trump actually going to pull forward the cuts that he has talked about, or that Elon has talked about? How real is that? There's a whole spectrum of opinions on it right now.

On one end, you're looking at the House and Senate reconciliation process for the budget proposal they've put forward, and you scratch your head and say, “Are we really cutting enough relative to what economists and others are telling us we need to do?”

Meanwhile, you've got Elon and Trump saying, “We're cutting, we're saving, we're going to get to $1 trillion a year.” But that's not necessarily showing up in the budget. Is it showing up in DOGE's actions? To be determined.

On the tariff side, there's a spectrum of how far these tariffs are going to go. The United States, up until the 1800s, was entirely tariff-driven in its federal-government revenue. It was a way of protecting industry here. Over time, tariff rates came down as we introduced an income tax, which started at 3% after the Civil War and went to 5% for high-income earners. In the 20th century, that totally flipped. Now we have almost no tariffs and a 50% income tax for the highest bracket.

Can we actually revert to a tariff-driven income model for the federal government? What is the economic effect on growth for corporate America in a world where taxes get cut for companies and individuals, but we make all of our money from global trade? Does the increased cost of global trade hurt companies more than the benefit of paying lower taxes? That's the big economic argument underway right now.

It's funny: it seems to fall along political lines, believe it or not, much like everything else.

Democratic-aligned economists say, “My party is doing—”

Exactly. Democratic-aligned economists say the tariffs don't make sense, reduce economic growth, and have a negative effect. Republican-aligned economists say the tax cuts will more than make up for the reduction from the tariffs.

That's the big unknown right now. There's a wide spectrum on the spending cuts, a big spectrum on what's actually going to get done with the income-tax cuts, and a big spectrum on how far the tariffs will go.

Those 3 things have very wide ranges, and they all interact to determine inflation, economic growth, and government deficits over the next decade. We don't have a clear picture yet of how they interact. They're being hotly debated, and there's high variability. They're changing day to day.

Trump says one tariff one day and another tariff the next day. Yesterday there was a whole bunch of confusion about tariffs; today there's a discussion about how far the tax-cut extension will go in the reconciliation process.

I encourage people to use my 72-hour rule and look at what happens 72 hours after Trump says something spicy. A lot of the time, he just says a lot of things.

But the House and Senate have both put forward their budgets. They go through this reconciliation process, and there's a lot in there that leaves a lot to be desired. If you're an absolute fiscal conservative trying to get us to a 3% deficit as a percentage of GDP, you're thinking, “Wait a second. Does this do enough?”

The tax cuts are $4.5 trillion over 10 years, approximately $450 billion a year. If you're trying to catch up, how are we doing tax cuts?

Remember, you can't make those statements as fact. A lot of those 10-year projections are based on someone's estimate of the economic effect of the tax cuts.

Some people say that if we make these tax cuts, the economy will grow faster than the CBO economists estimate. The CBO economists are trying to be conservative, so there's a lot of debate about how much this will really cost.

People say, “Trump is talking about raising our deficit so much over the next decade.” But there's a different point of view: if you assume the economy will grow because of these cuts, then that's not actually true.

Then there are all the wild cards around the Gold Card.

Yeah, we're about to get to that.

Are the tariffs going to generate $1 trillion a year? Are they going to generate $2 trillion in revenue? Nobody knows. It's just a negotiating position.

So that's my main point: this administration is all over the place. The cuts are great, but the bond market tells you a lot. The 10-year yield peaked at 5% 2 weeks before the election, then peaked again in the second week of January at 4.78%. Now it's down to 4.26%.

It's come down by a full half point in the last month, which tells you a lot about expectations for inflation and growth over the next decade. It's a reasonable sign that we don't think there will be rampant inflation over the next decade based on some of the policy decisions and actions being taken by this administration.

I would say there's some indication that, if you try to decode the enigma of the 3 things we're talking about, it's generally deflationary to some extent—or at least it's not inflationary.

Are you optimistic, net-net, Dave? Are you optimistic about this next 4-year period?

I'm honestly pretty uncertain, and I'm pretty unhappy with both the Senate and House budgets. Personally, I don't think there are enough cuts or enough action in there.

It's weird, because you hear Elon talking to all the members of the Cabinet, and he's pretty clear: “We've got to save this government. The government is in a debt spiral. We have to fix this problem,” yada yada. Then it's sort of business as usual.

When we were in Washington, D.C., that was exactly my observation of every senator, representative, and member of Congress we met with or talked to at a cocktail party. It was the same: “I've got to get this for my people.” That was the goal.

We've turned this federated republic into a whole bunch of elected representatives showing up in D.C., scrambling and grabbing money for their constituents. That's what they were hired and elected to do.

It's a really unfortunate circumstance that nobody looks out for the best interest of the U.S. dollar over time and says, “We've actually got a limitation. That limitation should be less than a 3% deficit-to-GDP ratio. That's our budget. That's our maximum budget.” Start from there and do a buildup.

Chamath, you've been talking about this. I don't know if it was a couple of weeks ago when you were tweeting about the Great Reset theory and the third or fourth turning. Do you want to encapsulate your thoughts?

You have to figure out what the goal is. One goal could be that the Republicans want to have consistent political power. That's a reasonable goal. The Democrats want that, too.

A different goal would be to do what Friedberg said: “We're going to take the lumps because we're going to defend the dollar and the credibility of the United States. We're going to make sure it's structurally sound and take the pain necessary to reset it.” That could be a goal.

I think the reality is somewhere in the middle. You can't be in one camp, and you can't be in the other, because I don't think you can get anything done.

The thing I've been thinking about a lot is: when will somebody figure out what the great coalition is that preserves political power, whether that's the Democrats or the Republicans?

The reality is that you will have a consistent majority if you get 3 cohorts of people together. Cohort number 1 is people who don't have many assets and are working- and middle-class. They don't necessarily own homes and don't necessarily have investments in the stock market, so they don't particularly care about what's happening there. That cohort dominates.

There was a clip of a discussion at Harvard this past week about the different political coalitions that voted for Trump versus Kamala Harris. The most important takeaway was that if you make $100,000 or more a year, you're a reliable Democratic voter. If you went to college, you're a reliable Democratic voter. Everything else is a reliable Republican voter.

The thing to remember is that the “everything else” bucket is growing faster than the first bucket.

You have this coalition of the asset-light working and middle class, along with patriotic businesspeople, patriotic business owners, and technology people who care about innovation. MAGA has been able to bring them into a coalition.

My point is that if that's the consistent, reliable coalition that cements political power over multiple elections, we've seen this before. Republicans can go on a 3-term or 4-term run, and Democrats have as well.

It's bad news for the stock market and bad news for asset owners because it doesn't reward the constituents, going back to Friedberg's point. If you're going to feed your constituents, and your constituents don't own stocks or homes—or are so wealthy that they can be inoculated from a massive drawdown in those asset categories—what do you think the winning strategy is?

That is my rough working version of what our version of Brexit is. If you have many years of austerity, what does it really result in? If you want to cement political power, I think it requires walking down these asset markets in a meaningful way. That's stocks and real estate. I don't see any other way around it.

That's a total theory, and I could change my mind as I get more data. I'm just trying to work through the possibilities and the distribution of outcomes. That's where my head is right now.

I think it's a good mental model. Politicians want to stay in power. How do they stay in power? The populace has to want to continue to back them, and they have to understand what backing strategies reward asset owners.

When asset owners are a shrinking minority, that's not a good idea.

Well, 60% of the country owns assets, but 80% of those assets are in the top 10%. It is definitely weighted heavily. People do have some exposure through their 401(k)s in some cases, and 61% of people own a home, but I think it's a good framework.

The good news, from my perspective, is that every time we have a great technological revolution—whether it was the iPhone, the internet, or now AI—that tends to make the most impact on the economy. Based on what I'm seeing on the streets, entrepreneurship is on fire right now.

No, that's not true. I think you're confusing that with the fact that everybody has an iPhone. You talk about this a lot. It hasn't lifted average hourly earnings that much. In fact, we've had massive wage suppression.

It has rewarded the employees and stockholders of Apple, Google, and Meta, but that's not everybody.

I'm talking about the United States and our place in the world and our economy when compared to other countries. I still think that if we lead AI, we will have the best standard of living and the best overall economy in the world.

But it does polarize the win to Apple shareholders in the case of the iPhone, or Google in the case of the internet. It makes the entire populace more efficient and makes the United States more efficient, since we led both of those revolutions.

I don't think it does. I think it supremely benefits a small cohort of people. That's why the denominator goes up, but does it affect individual people in measurable ways on a broad-based basis? I think the data has statistically proven that it does not.

That's why we have the populism we have today. It has disproportionately rewarded equity holders. That's obvious. Wage earners have not had the same escalation.

What do you guys think about the Gold Visa?

I love that Gold Visa. This is incredible, because I literally tweeted 6 months ago, “We should just sell citizenship for $500,000 a pop,” and he added a zero.

I'll give you a prediction. Within the next few months after this gets announced, you're going to hear about founders taking $5 million of secondary in a round to make sure that, if they're non-Americans, they get their visas.

100%.

Check out this prediction, Nick. We now have a Polymarket to trade: how many Gold Cards will Trump sell in 2025?

I think what is the bet?

You can have zero, 1 to 100, 100 to 1,000, and so on. The levels are 1, 100, 1,000, 2,500, and 5,000. You can buy the level you think has the probability.

There is currently an 8% probability on Polymarket that by the end of 2025 there will be zero Gold Cards sold. There's a 25% chance of 1 to 100, a 17.7% chance of 100 to 1,000, and so on. The most probable level is 2,500 to 5,000, at 29% probability.

I'm taking the over.

What do you think this is by the end of 2025?

They have to get the program up and running, and then people have to buy them. So it's by the end of 2025?

Yes.

People are really just betting on when he can get the first one done.

Exactly. How many does he get done?

I know. I'm going to take the top 2. I might take 5,000 and above.

Would you put real money on that?

You should do that.

You can't trade it because it's not available to Americans. But if there were a way to do it, I might.

This is not unprecedented. It's called the EB-5, which I talked about. Noncitizens can invest a lot of money, but it's a bit of a scam. I got pitched on it. People said, “We can get you limited partners for your fund. Here's how it works.” They invest in some fugazi real-estate project, and you have to create 10 full-time jobs.

There are a bunch of scams going on around the EB-5. I said, “As your president, I'm going to sell these citizenships and get 100,000 people to do $500,000 each.” I said we'd sell them like Taylor Swift tickets.

I have to tell you, I think Apple, Meta, and Microsoft would buy 1,000 to 10,000 of these. Let's say you were able to buy them and swap them out: if somebody left and went back to their country, you could still use it. You'd still have the visa.

These would become incredible for recruiting talent. If you had to get the CEO of a company over here and could offer them that, you could buy their company. Is that how it's going to work, or is it tied to a person?

There are 2 different issues. One is whether you could swap these between people. The president could make it work that way if they wanted to. Corporations could essentially receive a season pass that they could swap between users. Those things exist in the world as a concept, so the president could decide to do that.

The second piece is how valuable they are. Are they worth $5 million, or are they worth $1 million? Which price would sell the most?

The way this is proposed by Trump and Howard Lutnick is $5 million for essentially a green card. You get permanent residency in the United States and can live here permanently.

They're getting rid of the EB-5 program after this. That's their proposal.

Here's the math. How many people in the world have a net worth above $100 million?

We know there are 4,000 or 5,000 billionaires globally, according to the estimate.

I think there are a lot of hidden billionaires—in Russia, China, and even America. I would guess there are at least 10,000 or 15,000 billionaires in the world.

People with $100 million?

Yes.

Probably 50,000.

That's what I said.

The reported number is 28,000, 40% of whom are in the United States, which means there are 17,000 outside the United States.

Who knows if that number is real? Come on. How many Russian oligarchs have $100 million?

Let's use whatever fudge factor you want. There are 17,000 reported $100 millionaires outside the United States. Do you think that's the cutoff for people who would spend $5 million? What percentage of them would buy a U.S. green card for $5 million?

I think people with $20 million who are overseas in Venezuela or the Middle East would spend $5 million if it were a path to becoming a U.S. citizen. If you advertise that, over a lifetime, you could make twice as much money living here, I think that if you had $20 million, you'd give 25% of your current net worth to get into the United States.

Donald Trump said you wouldn't have to pay any tax on foreign assets.

Right.

Easy peasy. It's a true green card.

That is not what a real green card is. With a real green card, your global income is taxed.

That's true.

So it's worse. This is way better.

If I had to do it again and this were available to me, I'm not sure I'd spend $5 million.

I'm not sure there's a million buyers. I think there are probably 10,000 maximum buyers for this thing.

I would take the over on 10,000.

I wouldn't take the over on a million. I think you're right on the million number.

What did Trump say? He said there are a million.

No, the most probable level is 1 to 2,500. That's probably right. But you didn't ask for the total; you asked for year 1. It takes 6 months to get anything done.

This poll is dumb. The real question is how hard they will be vetted. There's a lot of gray money around the world, so the question is whether you can bring it into the light.

I know many people in India who are extremely wealthy in ways we don't understand. Their wealth is literally in cash and gold. If they wanted to raise their families in America, how would they take their assets to JPMorgan? Would they show them the Gold Card, and JPMorgan would say, “Great”?

If there's a workaround to the KYC, AML laws—honestly, Friedberg, you could sell 2 million of these things.

I don't think so. If you literally have to go through the existing frameworks on OFAC, AML, KYC, and all that stuff, it's probably in the tens of thousands.

I have to say, this is one of the greatest proposals ever. It's fantastic. Combined with DOGE—if he gets this done, if he gets DOGE done, and if he gets accredited investing done, I'm voting for his third term.

Let's redo accredited investing. If he gets those 3 things done—

Explain to us your personal interest in accredited investing. What's the grift connection? I'm not sure I'm fully tracking.

It's not a grift connection. I feel like there are a bunch of people stealing money through crypto scams, and all of that would be solved if people could take a test to become an accredited investor.

Currently, only about 6% or 7% of the country is accredited. If people understood diversification and how different vehicles work, convertible debt, and so on, you could take people who are gambling in the stock market and allow them to invest in private-market companies.

I believe that would create more upward mobility. A person who's an Uber driver or an HR person working at a company could put $500 or $1,000 into a new product or service they're using instead of betting on the Knicks or the Jets, God forbid.

They could put that $500 into the new product or service they're using—LinkedIn, for example—and that would allow more startups to get created. So many people contact me after reading my book and say, “I want to invest in startups,” but they can't. I think they'd be much better off putting $100 or $500 into a startup than wasting it at roulette.

I'll take the opposite side. I'll tell you why. I think you're right that it would be great, but they should buy the S&P 500.

They can do that today. They can get a Robinhood account. It's 11% a year. Perfect.

That's the protectionist, paternalistic approach we've had. It makes them safe, and their $1,000 becomes $1,070 the next year and $1,150 the year after. Great. They learn one lesson: the rule of 72 and compounding interest.

When you start betting on startups, you learn how entrepreneurship works and how product-market fit works. So, sure, put 80% into an index fund and 20% into private companies. They would learn more.

When Uber wanted to give Uber drivers access to buying shares, they weren't allowed to. Rich people can buy whatever they want. They can make whatever bets they want. They can be in private equity and in all these things that have the chance to go 100x or 10x, but poor people can't.

All I'm saying is that if they're educated and take a course, let them take a little bit of risk intelligently and learn about entrepreneurship.

I grew up blue-collar and didn't have exposure to how private-company formation worked. I didn't understand any of this. I had to battle my way to learn all of it.

If you had a course and people could go just as easily as they go to PrizePicks—which I bet on every Knicks game—as they go to Coinbase or a private-market company and invest, that would be better for upward mobility.

So you're right: people are going to lose money. But they're going to learn.

What do you think, Chamath? Settle the difference between the 2 of us.

I think both are true. We're all much better off owning indices—or at least that was true.

The problem with these indices right now is that they're not really well-balanced indices. The rules have changed because these companies have been smart enough to lobby people like S&P, and S&P has allowed the thresholds to creep up.

When you're buying the S&P 500, you're not really doing that anymore. You're buying the S&P 7, and the rest of the 493 is about 60%. If that's what you want, that's fine, but we'd have to fix the ETF market to make sure there was more transparency and balance.

These weighted indices are basically just the Magnificent 7. That's neither good nor bad. I'm just saying that's what it is.

I'm not talking about diversification. I'm talking about vetted, mature, real companies versus what we've seen time and again with people who aren't sophisticated or experienced when they enter a new market.

Any market has a process of adverse selection. Predatory practices, predatory pitches, and people saying, “Invest in this; it's a great deal; this is the new thing” show up. Most people aren't able to vet those opportunities and end up being taken advantage of.

That's the problem. They're being taken advantage of in every crypto scheme right now.

All I'm arguing for is more education and a path for people who want to do it. Show 5 hours of education and give them 50 questions to demonstrate that they have an above-average knowledge of how private companies work.

The balance we have to strike is that there are a lot of people on the outside looking in with no assets. Second, there are a lot of young people who want high-alpha opportunities, as crypto represents.

It's easy for you to pull the ladder up from under you because you're already rich. For people who aren't rich, if you went back to when you were poor, how would you have reacted if somebody above you said, “I'm going to tell you what you can invest in”?

Would you have said, “Okay, that seems reasonable. I know you're looking out for me”? That's the question.

I don't disagree with the goal. I think there's a reason we have securities regulations and securities laws that public companies have to follow. Private companies are more lax, and that's where the distinction is.

That's not what I'm saying. Now that you're rich, you want rules for everybody else. What I'm saying is—

Don't mischaracterize me, Chamath. That's not true at all. I'm obviously a free-market guy. I don't care what poor people invest in. I'm pointing out the consequence of what would happen.

Go back to when you were poor. How would you have reacted?

I would want to invest in everything. I'm not disagreeing with the notion. I'm pointing out what will happen: predatory assholes will show up and rip people off. That's what happens in every cycle.

If you're educated, that's why the education component matters. You'll learn something, just like people are learning right now not to bet on the Jets ever.

I'm telling you that's what's going to happen. Then Elizabeth Warren is going to get on television and say, “We have to fix this.”

Nobody cares what she has to say.

That's how this goes. I'm pointing out the cycle.

I agree with you that it is the cycle, but how do we fix it? How do you allow people who don't have assets to have assets that work for them?

Even as you guys know, we looked at the data. Even the best venture-capital firms in Silicon Valley, with the smartest and most sophisticated people investing in private assets, weren't able to beat the Nasdaq over the years.

That's true.

I'm confused. What you were saying before is that people should only be allowed to invest in the S&P.

I didn't say that.

That is what you said.

I said, “Here's the consequence.” I'm predicting doom.

Got it.

Here's the other possibility: eBay, Etsy, Airbnb, and DoorDash could say to the people who are part of their networks, “For every 100 rides you do, we'll give you $100 in shares. For every 100 nights you book, we'll give you $1,000 in shares of Airbnb. You can buy extra shares if you want.”

The government could say you can spend up to 20% of your yearly income, averaged over the past few years, on investing in startups.

Some number of people who built those networks—whether it was Google's network, eBay's network, Uber's network, or DoorDash's network—or who were part of Tesla would hit massive home runs. They'd move from the bottom third to the middle third.

Then some of those people would say, “I got really educated. I looked at and understood Tesla and Uber, so now I'm going to bet on this AI self-driving company and this other company that makes robots and delivers burritos.”

The entire group of people in the United States would get savvier about entrepreneurship and capital allocation. That's a good thing.

I think what's going to happen is not much of anything. The rules are going to stay exactly where they are, in favor of the top 10%, because I think this argument between the 2 of you is exactly the reason it can never change.

Maybe the alternative comes back to what I said before: the only alternative left is to debase assets. If you debase assets and make them much cheaper, theoretically there's less money to lose per quantum of investment.

Maybe that's the right way to think about it.

We have to get more people owning equities in this country. That's the high-order bit. If you feel like you have more agency in your life and you're smarter and savvier, that's the American dream, and we've lost the American dream.

Half the country doesn't feel like it can ever get into the top half. They don't feel like they'll ever be able to buy a second home or even a first home.

You have this helplessness in one group of people saying, “I need a handout,” while the other group says, “Got any stock tips? Where are you making money? What can I place a bet on?”

We're sitting here at a rigged game. We all get to play in one casino, and everybody else gets to work in the casino. I just want the people who work in the casino to be able to place some bets and maybe become owners in businesses.

Let's talk about the U.S. Postal Service. Trump is going to issue an executive order to dissolve the leadership of USPS, and the Postal Service is going postal about it in some ways—not literally, but they're angry. The proposal is to absorb the agency into the executive branch.

The post office has been operating for 250 years. Trump plans to fire the governing board and place the agency under the control of Commerce Secretary Howard Lutnick.

For context, the U.S. Postal Service lost $10 billion last year on $80 billion in revenue. It can't figure out how to make a simple profit margin or even break even, and it employs 635,000 workers.

Howard Lutnick did an interview with Fox News yesterday in which he said one idea he took back to the president was for the Postal Service to do the census, which would save $4 billion a year.

Give him another idea: nonfarm payrolls and GDP data should be collected by USPS, too, because they touch every business. You could get much more accurate information instead of sampling with all this error and craziness.

I'm surprised we don't see even more dramatic revisions. That is probably errors on top of errors. I don't trust the GDP data or the unemployment rate. We all know this stuff is crazy and wrong. I just don't know how wrong it is.

This is where Stripe's data might come in handy. I tweeted about this, and it became one of my most popular or controversial tweets. It got 3.5 million views without an Elon retweet or anything.

I had a simple concept. First, the Postal Service goes down to one time a week. Easy peasy. Once a week, there's nothing coming through the post office that's that important.

Second, starting next year, all citizens have to opt into receiving postal mail by paying $1 a year. You have to sign up and give them a credit card or something. I'm thinking 80% of people won't bother because it's all flyers and garbage anyway.

What people don't know—because I was in the magazine business and knew all about this—is that we had a magazine rate and a media rate, and all these marketers subsidized it. This is the ultimate marketing and publishing grift.

Magazines, newspapers, publications, advertisers, and catalogs pay almost nothing. They should double or triple the rate or remove the discounts.

Then take all those buildings and put them into the new sovereign wealth fund. Redeploy the buildings, get some money out of them, give every postal worker 2 years of severance—or 1 year, half a year, or a quarter year—retrain them, and let the private markets handle this.

There was a tweet from a woman who got leaked data from one of her friends or colleagues in the government. They broke down 7 or 8 leases and real-estate arrangements, I think inside Veterans Affairs. The numbers are astounding.

Half the office space isn't being used, and the other half is being underutilized. It's bonkers. They're going to be able to sell 75% of the stuff.

Don't blame me here, but I think it's a really good opportunity.

Our friend Jeff Bezos should come on the pod. Jeff, sit in the Sacks chair one time. That would be fun. He's making some big changes at The Washington Post. He's lost a fortune running the thing, and it seems like he's getting engaged and going into founder mode, dare I say.

He posted to his X account and emailed everybody that the editorial page is going to be run differently. He said:

“While newspapers once had a mandate to publish opinions from the broadest possible spectrum, the internet now mostly covers that. I'm confident that free markets and personal liberties are right for America. I also believe these viewpoints are underserved in the current market of ideas and news opinion.”

So he's going to focus on those 2 pillars: personal liberties and free markets. This seems awesome, and I could get into why it's brilliant on a publication basis, but I'm wondering what your thoughts are with him getting more engaged with a publication he was incredibly hands-off with.

I was surprised that he wrote this. If you want to write about personal liberty, one of the tenets of personal liberty is free speech. But he's effectively said that certain opinions aren't allowed anymore.

I don't think that's the solution to The Washington Post. All it does is polarize the readership even more.

I looked at Google Trends. The overwhelming majority of Washington Post readers are in Washington, D.C., and then Maryland and Virginia, the 2 surrounding states. It's very much a Beltway paper.

I think he's trying to have a direct influence on the ideas that people inside the Beltway read. As the owner, he's allowed to do it, but I wasn't a fan of the idea because I think the Elon plan is much better.

Here's the fire hose: go at it.

You have to find the people. Despite all the conspiracy theories, I don't think Elon suppresses free speech in the least. In fact, I think it's a literal free-for-all inside X.

It is a free-for-all. We have Nazi diamond pendants coming from Kanye.

The difficulty with X, which I think will be the next set of features he'll have to figure out, is how curation happens. You're curating, I'm curating, and other people are curating.

For example, when I go to an account that I like, there's no easy way to mass-follow a bunch of the people that account follows. I can't just copy it or start with a profile.

Those are all features that would allow you to take on all kinds of opinions right away and filter from there. I think that would be useful.

If I owned The Washington Post, I would have been even more extreme on the free-speech part. I would not have sanctioned speech.

Newspapers historically always had a point of view. They picked a side. Fox, MSNBC, and cable news have picked a side.

This will make the publication viable by picking a side and saying, “Here's what we stand for. This is our belief system.” I think it will make it viable in one way.

You're right: he wants to have a certain influence. That's why people buy these things. That's why people have historically owned them. They have a point of view.

The idea that it didn't have a point of view previously was probably a mirage. Some people felt there was objectivity, but I like him being more engaged with it and tightening it up.

All right, for the Sultan of Science, Chamath, our sick friend, the comedian, get well soon. We can't wait to have you on. It's going to be a hilarious time. And for David Sacks, who's very busy, the Rain Man in Washington, D.C., saving the world, I am the world's greatest moderator, and we will see you next time. Bye-bye. Love you.