AI泡沫、稳定币热潮与《追梦之路》｜BG2 with Bill Gurley and Brad Gerstner

1. Gurley: “I’m Jumping on the Crypto Train”

I'm applauding the innovation. I'm jumping on board the crypto train, and I hope the incumbents aren't able to strangle this thing in Washington.

Hey, man. Great to see you.

Good to see you, Brad.

What an incredible weekend of college sports. I have to bring this up. Texas had that big upset of No. 6 Oklahoma. You had to be pretty stoked about that.

It was a good game. It was fun. People who haven't been to a neutral-site game—such as Florida–Georgia or Texas–Oklahoma—should know that they meet every year in the middle of the Texas State Fair. The stadium has several hundred thousand people outside of it.

When you get inside, right on the 50-yard line, one team's fans are on one side and the other team's fans are on the other. It's loud, and it goes back and forth. It's unlike the experience you get when there's a home team and the crowd is all rooting for one team.

Oh, that's cool. My Hoosiers—my Indiana Hoosiers, Bill—upset the No. 3 Ducks, going to 6–0.

I have a lot of Duck fans in my friend group, so I'm going to refrain from celebrating with you. But I have to say, I'm used to celebrating Hoosier basketball, but rarely Hoosier football. Curt Cignetti has done an unbelievable job turning that program around.

My 90-year-old mother was watching that game and sending me play-by-play. So congrats to the Hoosiers.

Well, as long as we're calling out college football teams, then we can move on. UCLA started the season 0–4, then upset Penn State at home with almost no fans there and won big again this week. Apparently, there was a coaching change after the 0–4 start, so this could be the biggest turnaround in the history of college football. Go Bruins.

Pretty incredible.

2. Two Years of BG2

There's so much happening in the world today, and we're going to unpack a few of those things. We're going to follow up on some of the issues that I raised in the Jensen Huang episode: the latest AI announcements, all this bubble talk, circularity of revenues, quality of revenues, and AI regulation.

But we're also going to do something today that we don't often talk about on the pod, and that's life and career. Bill, you have a huge book coming out: Running Down a Dream: How to Thrive in a Career You Actually Love. I'm so excited for this book, so we're also going to talk a little bit about that today.

We're coming up on the 2-year anniversary of this pod. I can't believe it. Time has flown by. When you and I talked about doing this, we said our mission was that we really wanted to talk about markets, investing, capitalism, and companies, but through the eyes of the investment analyst.

You and I, more than anything else, are analysts. We try to find the biggest problems, opportunities, and challenges in the world, study them deeply, and compare notes nonstop. Occasionally, as an analyst, that leads you to a big investment idea. Sometimes it leads you to a podcast, maybe writing an article, teaching a class, or, for you, writing this book. Sometimes it even leads to a major policy initiative, like the Invest America Act, which actually became law.

I think you would agree with me that the response over the last 2 years has been amazing—more than either you or I expected. But that also creates its own pressure: to show up, to deliver those unique insights, and this takes a bunch of time.

3. Bill Stepping Back to Write His Book

So, given that—and I don't want to bury the lead here—you have some huge upcoming projects you want to work on, and you're going to step back from co-hosting the pod. I'll still talk you into joining on occasion, maybe to be a guest, but you're freeing up time to work on your big passions, like this book, and to go deeper into these topics that people have heard you talk about here: U.S.–China relations, regulatory capture, and the dysfunctional state of U.S. health care.

For those interested, the pod's mission remains the same. I'm going to keep the same name. We're going to keep chopping it up with analysts I respect, sometimes with Bill, and covering topics that matter—like last week's pod with Jensen Huang or upcoming pods I have with Sam Altman or Sacha.

This is a moment of really unique consequence. We both recognize that. We're grateful to have the opportunity to open-source these conversations that are truly shaping the future. I know I speak for you when I say we do it for the love of the game. It keeps us sharp and keeps us on edge. It's a privilege to get on here, chop it up, and share something back with the tech ecosystem that has given us so much.

So, Bill, do you have anything you want to say? It's been a good 2-year run.

First of all, thanks to you, Brad. This has been great going back and forth. I had 2 primary initiatives coming into it. One, as you mentioned, was to stay sharp, and the other one was to share and give back.

4. Gurley’s Next Mission

I've been writing my thoughts on the tech industry since I was a sell-side analyst, so it's coming up on 30 years, and I've always enjoyed thinking out loud. I think it makes us better as analysts and helps us to understand. But I also like to share with people, and there's no question in my mind that this got bigger than I ever anticipated it would.

I've been chased down in international cities recently. I started talking with someone, and they had no idea what I looked like. But the minute they heard my voice, they said, "Oh, you're the guy from the podcast." So it has been popular, and I know there are going to be people who are upset with me. All I can say is, I'm sorry, and I apologize that I'm not going to be doing it anymore.

5. Runnin' Down a Dream

I came across this quote that was really inspiring to me: "Life begins where your comfort zone ends." There were a number of people who helped push me to write the book. It's taken up quite a bit of time in the last 8 years. It's been a very long project. We'll talk about it more later, but I'm feeling a calling to go work on—or at least attempt to work on—some of these bigger issues.

I want to create a platform for that. I want to create room for it and move a bit away from the space that I know quite well and love quite a bit, but push myself outside of my own comfort zone and hopefully have an impact on things that really matter.

You and I talked about this throughout the entire time I was working on Invest America. I certainly encourage you and push you to do this. I think you have an enormous amount to contribute.

Listen, you and I chop it up together every day. I know where you stand on a lot of these issues. I'll bring those opinions to bear for our audience, and I certainly know that you'll have the burning need to come on as a guest on occasion and share some of those views.

6. The AI CapEx Bubble?

But, in the spirit of analysis, let's dive in. This AI money bubble and the Jensen Huang episode—let's start by talking about that.

We've had a flurry of announcements, including another announcement this morning between OpenAI and Broadcom, where OpenAI is going to be building its own inference accelerator, amounting to well over $1 trillion of incremental CapEx. That's above and beyond what we already knew was going to get built out.

I know that you have concerns about the level of CapEx—the absolute level—and I know that you also have meaningful concerns about how it's being financed. Why don't you walk us through your major concerns?

I think anybody who's been a student of financial history has studied different types of activities that, historically, have created red flags. The reason that any AI person you talk to would know what you mean if you said "circular revenues" is because someone has used it in the past in a way that wasn't good.

I had an exercise, which I tweeted—we can put it in the show notes, and people can find it—but I described 6 different transactions that have happened now that I would say are non-normal. I described those things to ChatGPT and asked for its analysis, both as an accountant and as a financial investor.

The AI itself would find its way toward company names like Enron and WorldCom and those kinds of things merely by describing the type of transaction. I think that suggests, if we believe in intelligent AI, that that's just what historically has become the best practice and way to think about these things.

I've told you before, I think you have highlighted that some of the multiples are actually not that high. I think this is part of the reason: There are red flags that people are looking at.

If you peel that back a little bit more, one of the things that you and I have talked about is the very nature of round-tripping, or circular revenues. I think there's a continuum.

On 1 end of the continuum is a true sham transaction. There's no underlying demand for the product. I send you $1 billion, and you send me the $1 billion back. That's clearly a sham transaction because there's no underlying demand.

7. Sham vs. Real Transactions in AI

On the other end, I have massive demand for my product, you have plenty of places where you can go get capital, and we just happen to have an investment relationship in addition to that, and I'm buying your product. Those things happen all over the place in our economy. It may be something to pay attention to, but it's certainly not even close to being illegal, and it frankly doesn't cause me a lot of concern about the quality of revenue.

Then we have things in the middle. You can ask a question like, "Would this much revenue or this much product have been purchased but for this investment?" I think that, at a minimum, calls the quality of those revenues into question.

So, when you look at that, do you discriminate between the types of transactions that have been announced? You raised this question first 18 months ago about the credit transactions that were occurring with a hyperscaler. So, maybe just unpack a few of the different types of transactions.

Yeah, and look, I think it started at the very beginning. I think that's one of the things that's causing this: it's become part of the competitive landscape and the competitive dynamic. I think there are many boards and many CFOs who have been put in a position where they say, “Well, if we don't do it, everyone else is doing it. You might fall behind.”

8. Microsoft–OpenAI Credits & Cloud Economics

It started, from my perspective, with the original Microsoft–OpenAI deal, where credits go in as an in-kind investment, and then those credits are used back against Azure and Microsoft cloud services. In that case, as I said back then, I'll say it again now, that's revenue—a cashless transaction. There's no cash, but it becomes an income-statement revenue item for Microsoft. I don't think that's ideal from an economic standpoint.

That practice has now, I think, happened at Amazon and happened at Google. I think they've made investments in other AI startups with the same kind of thing. At the very least, it drives usage of their product versus someone else's. In the worst case, it creates revenue that might not have existed had it not been for that deal, or at least not on those terms. But anyway, it started there. It's become quite competitive now.

There's an interesting podcast on Plain English, which is in the Bill Simmons family, with Paul Kedrosky. He makes the argument that part of the reason these transactions are taking place—I think this is a credible argument—is because some of these players have already put so much CapEx and so much debt on themselves that they don't want to take the next step. So, in that case, you have reached some level where the company's saying, “Oops, I feel uncomfortable going further than this.”

The transaction that comes to my mind when I think of that is one where Meta agreed to cover the debt risk on a facility where they don't own the debt. To me, that's classic off-balance-sheet financing. If they own the risk of it, just because they don't own the paper, I don't see the difference, really. But, like I said, this is happening in a lot of different places.

9. Nvidia’s Investments: Healthy or Hype?

One of the things that, again, I've talked about a little bit on All-In and other places: if I look at the Nvidia deal as an example, Bill, Nvidia has the opportunity to invest, though not the obligation to invest. OpenAI has the opportunity to use their chips, though not the obligation to use their chips, as evidenced by the fact that they just announced their own chip this morning, and they just cut a huge deal with AMD.

In the case of Nvidia, you're not talking about a highly levered business. It's a company that's going to generate $450 billion of free cash and is taking a small fraction of that over the next 3 years and investing in companies that it thinks are good-returning investments. Google and Google Capital have been doing this for years, et cetera.

Again, I think in those cases, you can say for certain that maybe more of their product is being consumed than would have otherwise been consumed. We saw this announcement last week where they're investing in xAI with respect to the new round. Most of these companies that they're investing in, I think, have the economic wherewithal to raise the capital in other places. Elon could certainly raise it in other places. OpenAI was well oversubscribed, so they could have raised it in other places.

But here's what I think people should be on the lookout for: where would I have more concern? Imagine there's a chip that has only 1 customer. There's not a lot of demand for the chip, and that chip manufacturer gives a customer $10 billion, and that customer turns around and buys that chip. So, there are no other potential customers, and the buyer would not have had the ability to buy it but for that capital. That, to me, raises big red flags.

I do think in this overall ecosystem, the reason I'm happy you're bringing it up is that one of the things we need to do to keep the wall of worry there, to keep the excesses from emerging, is to call them out. I'm not concerned as a shareholder in Nvidia with what I'm seeing Nvidia do today. I like how they're deploying their cash on their balance sheet.

But I do think that as you go further and further out the risk curve—further and further to these startup neoclouds, or further and further to startup chips, et cetera—where people, to your point, are a little bit more desperate for capital, don't have the balance sheets, and don't have the market leadership position, I would not be surprised at all in this moment to see more of those yellow flags emerge.

There are a couple things that I would say in response. One, there's a reason we know about a lot of these things, and that's because some auditor somewhere made them disclose them. They felt that it was abnormal enough to require disclosure. Second, I listened to you and the All-In team talk about this issue. I do think investment is riskier, or more risk-seeking, than the customer loans it was compared to.

Cisco got in trouble just with the customer loans, because they were giving loans to startups that really didn't have the wherewithal to pay them back. But that's really the issue for me, though: when you switch from a loan to equity, you no longer have to pay it back. So, in some ways, it's easier on the purchaser than if they had a loan.

But here's my bottom line. I think what this overall situation does, first of all, is that I think it's driven by competition at this point. The first step into the gray zone was way back at the beginning, and so now I think we're fairly pregnant with it.

I think it's a competitive dynamic. I think it increases the chance that we go over the top, that we end up over-provisioning. I kind of felt like that was unavoidable anyway, but now I think it's higher. I think it maybe pushes out when we find out that happens, because you've just created more virtual leverage on the whole system, and you might be hiding some of the signs that would tell you things are slowing down.

10. CoreWeave, Hidden Leverage & Demand Risks

I'll give you a great example. One of the more peculiar of all the deals—and this was disclosed in a CoreWeave filing—was that Nvidia has promised to buy any of CoreWeave's service capacity that they can't sell to anyone else. That is very unusual. That's not the same as making an investment. That could easily help CoreWeave with their creditors in getting more debt financing.

But it also means, as an investor, we don't know what's going on with real demand for CoreWeave, because we probably won't be told if they start moving into the world where they're offloading to Nvidia or not. If you'd have said to me, “What would you look for to see if we've reached a point where things are slowing a little bit?” you'd say, “Well, let's look at one of the pure plays.”

And so now that's money. I could see how that could be, but I would expect that every analyst on every CoreWeave call for the next 8 quarters—maybe thanks to you just raising the flag—is going to be asking the question: “Do you see any slowing? Are you having to send any of your demand to Nvidia as a result of this?”

One of the things I like about this as well: these are public companies, both CoreWeave and Nvidia. It is a disclosed transaction. It's not like this stuff is occurring in the dark of night. People can ask questions about this with regard to demand.

And I will tell you, the amount of money that is being spent to track every single part of this supply chain, from Taiwan to the United States—I mean, look at Dylan Patel's business at SemiAnalysis. The thing has exploded. The amount of money people are spending just to stay on top of this is enormous.

The second they see something that smacks of any leakage in demand, boom, stocks fall and warnings go up. So, I think it's a good point. But let me transition, because I do want to talk about this question of demand.

On the one hand, there's this question about quality of revenues. The other question is, are we overbuilding? So, let's show this chart again. This is basically the $3 trillion of build-out expected over the next 5 years. This is the CapEx chart that we've shown here before.

11. 3 Trillion CapEx: Are We Overbuilding AI?

To put that in perspective, Bill, that's about 60 gigawatts, right? Because we now are normalizing everything to gigawatts of data center. So, that's about 60 gigawatts. It's not all incremental; a lot of that is replacement or upgrade. Keep that in mind.

The second is this chart of Nvidia revenues. This is the Nvidia sell-side forecast, okay? The forecast this year is for about $200 billion in revenues, growing to about $350 billion in revenues over the next 5 years.

So, this year, that means that they're selling about 4 to 5 gigawatts worth of compute. Again, most of that's incremental, but it's not all incremental. That would grow to 9 gigawatts of compute—9—in 2029–2030.

So, that's $350 billion or $400 billion. That's the Nvidia consensus revenue forecast, right? I asked Jensen on the pod about this, and I said, “What is the chance that we get into a glut over the course of the next 4 or 5 years?” We'll play the piece, but he basically said there's zero chance over the next 2 to 3 years because all the build-out will go to the biggest hyperscalers with the biggest balance sheets in the world, and they're building it to run their core businesses.

We haven't even gotten into the full amount with respect to these new generative AI workloads. So, what is the percentage probability that you think we'll run into a glut in the next 3, 4, or 5 years? Until we fully convert all general-purpose computing to accelerated computing and AI. Until we do that.

Yes. I think the chances are extremely low. Okay, so here's a question I have for you: Did you hear anything in the last couple of weeks that caused you to believe that we're on the verge of some bubble bursting, that we're greatly overbuilding, or anything else? Or are the flags just up and now it's a wait-and-see?

Yeah, sure. It's funny: They had Howard Marks on CNBC this morning. I'm a huge Howard Marks fan, and they asked him this question, and he said, “Look, multiples are too low for this to be a bubble. You can't be on bubble watch if the multiples aren't high enough.” And you've been making this point for a long time.

12. The MAG5 CapEx Surge: Too Much, Too Fast?

I would say you'd have to be a fool not to notice that these numbers you're talking about are so remarkably unprecedented from anything we've ever seen before. They are massive. I've talked about seeing the Magnificent 7 go from being massive cash producers to where many of them are taking the majority of their free cash flow and putting it into CapEx. It's totally new, and clearly everyone believes that this wave is maybe bigger than the previous waves we've seen that have led to so much value creation.

So, all that's happening. I like to believe that it's okay to recognize that the market's great and still think that these transactions shouldn't happen this way. I'm able to keep both those things in my head at the same time.

I think it's a super fair point.

By the way, you just mentioned it, so we'll include this chart. This is Mag 5 CapEx as a percentage of their operating free cash flow, Bill. And if you look at it in 2025—that's this year—they'll spend about 66% of their operating cash flow on CapEx.

Yep.

Right? And if you look at the consensus forecast for their CapEx relative to their operating cash flow, this is the peak, around 66%, and it has it going down to about 45% or 50%. Now, embedded in there is that they're going to keep growing their operating free cash flow at 15% to 20% a year, right? So, there's still room for them to grow with that coming down.

But I think that's another thing to keep your eye on: How much are they spending? And by the way, just to give you an order of magnitude, Bill, in 2023 their total CapEx was 156 billion, and this year it's 379 billion. Right? So, a radical step-up, and your point's a good one.

And just remember a couple of years ago, in 2022, when Meta stepped up its CapEx spending on Reality Labs, the stock got obliterated because people said, “What the hell are you doing? This is all about free cash flow per share.” Including myself—I was saying, “Let's get fit here. Let's drive more free cash flow out of the business.” So much so that the CFO sent me a hat that says “free cash flow,” right?

13. Meta’s CapEx Comeback & AI ROI

So, they got real about free cash flow, but there's a difference between investing that free cash flow in data centers and AI and investing it in Reality Labs. As an investor, let me just tell you my own perspective: The reason Meta's stock is doing great, notwithstanding going back to high levels of CapEx spend, is because now the investors understand it and believe in it. We're seeing the benefits, right, in the earnings of the business. They're growing the earnings of the business. They don't have to hire a lot of new employees, and so it's fundamentally different than the CapEx that was going into Reality Labs, where investors were saying, “Hold on a second, we're going to spend $100-plus billion over the next 5 years. We don't even know what we're building.” Right? We don't know what it'll be worth at the end of the day.

So, I think for now, at least, there's enough belief in the byproduct of generative AI because people are using ChatGPT, they're seeing the utility in the enterprise, that they're willing to tolerate these companies giving over half of their free cash flow to these build-outs.

I do think another dynamic is the race condition created by the competitive dynamic. And it appears from where I sit—and you don't need to comment, because you're an investor and maybe have more information than I do. You probably have more information than I do—but it appears to me that OpenAI, through all these partnerships and announcements, is trying to create escape velocity, you know?

14. OpenAI’s Race for Escape Velocity

That could be against the model providers; it could be against a hosting provider, depending on how you think the market plays out. It could be on the consumer side; it could be on the API side. But it creates an interesting stress test for anyone else in the ecosystem to say, “Are you going to give chase?” Because all the numbers you laid out there are gargantuan, and it'll be interesting. That's my opinion. It just feels like they're daring people to follow them, and I suspect a bunch don't. I mean, it may work.

Well, you've seen this before, I know. We've talked about it many times on here. This was in Lyft. Ultimately, I think a couple of things to remember: These announcements are frameworks. They allow people to begin working, but they're not etched in stone. These are not contractual obligations. Everybody's got to deliver their parts. If demand comes in lower, then these people are not going to—

Oracle wouldn't have to be contractual for it to be RPO. You may know more.

Well, for sure—I shouldn't say that they're all frameworks—but I know, for example, in the case of AMD, they're going to have to deliver a workable chip, or you're not going to build 6 gigawatts' worth. Yeah. Right? In the case of this Broadcom announcement, obviously they have to build a workable chip.

So, I think your speculation—and, again, it makes sense to me—is a pretty big one. If you said, “What are the advantages of getting out there and locking up all of these deals?” I can't imagine it doesn't help a lot with recruiting. All the best researchers in the world want to work at the place that has the most compute, and so you want to lock up the compute. I have to imagine it helps with the supply chain because now you're locking up that supply. So, I think your speculation is a pretty big one, a pretty good one.

But at the end of the day, if you add up all these deals, I tried to do this, and we may be off by a bit. I'd encourage people who have a better estimate to let me know. But if you add them all up, it looks like to me OpenAI would be on the hook for, like, 150 billion of CapEx in 2030.

Okay? And so, the question is, Bill, how much revenue do they need in 2030 to justify 150 billion in CapEx? Well, I think you would need at least 150 billion of revenue, right? At a minimum. We just talked about Meta and these companies spending 66% on CapEx. But if they had 150 billion of revenue, then the question is: Is it plausible they could have 150 billion of revenue in 2030? And I would argue as an investor that it's more than plausible that they could have 150 billion of revenue.

But I think it makes it very, very difficult for anybody else other than the hyperscalers. Obviously, Google's going to be there. Obviously, Meta's going to be there. Obviously, Amazon can be there. But it makes it very, very difficult for anybody else in the ecosystem, right, who believes this is a game of scale, where compute competes. So, I think your point's a good one.

All right.

15. AI Regulation: State Patchwork Madness

Maybe shift a little bit to an area of passion for both you and me, which is AI regulation. We've talked on the pod many times about the concerning patchwork of these emerging state regulations that, under the guise of doing good—and maybe they're even well-intentioned—cause a hell of a lot more confusion at best, and at worst they set back our leading frontier labs and really hamper us in the race to stay in the lead in global AI.

Well, it's gone from being more theoretical to now more problematic. I tweeted over the weekend, in particular, about this Colorado AI Act, which is now passed into law, signed into law. It defines something called algorithmic discrimination by outlining these 12 protected classes: of course, age, color, religion, but also limited proficiency in the English language and reproductive health. And it basically said if the algorithm provides information, right—if the chatbot provides information that's used to discriminate—then there's liability back at the frontier-model level, right?

And just this morning, Gavin Newsom signed SB 243, which mandates safety protocols for AI chatbot companions and gives any consumer a private right of action to sue these companies for any emotional harm that comes out of a chatbot. I kid you not. You can't make this stuff up. And you said recently that China is so competitive with the United States because it's run by engineers and America is run by lawyers, and that's the greatest risk we have.

16. Why Federal Preemption Is Critical

Talk to us about the need for federal preemption and, again, let's dive back into your concern about these 2 laws that were just passed. As an aside, I just consumed Jonathan Haidt's book The Anxious Generation, where he talks about what he believes are some social harms caused by some of the apps in the internet ecosystem. And I do think a lot of the passion for writing some of these states' laws comes from that place.

There are local congressmen who feel like they should have been out in front of social media more, and so they want to get a jump on this. And I think some of that comes from there.

You run this massive risk of trying to regulate a brand-new technology at a state-by-state level. You could ask yourself—and, by the way, I’ve said this a lot about policy—the intent of the policy is different from what happens once the policy is implemented. People can come in with great intentions, and this goes back to my speech at All-In on regulatory capture, but you can end up with the exact opposite outcome of what you intended because you just don’t know enough about the way you write the regulation.

Right now, a lot of people believe we’re in this global competition to see whose tech stack for AI is used on a global basis. If we implement 50 different state rules that these companies have to jump through, while competitors competing in the broader world don’t have any of them, there is zero chance that’s not going to create mud and slow down the U.S. players. There’s just zero chance.

I’m certain the people writing these laws don’t understand that there might be some global consequence of what they’re doing. I can remember when Obama was excited about removing some of the state-by-state requirements on hairstylists and whatnot because it makes it such that they can’t move between states. It’s kind of ridiculous that they would have different laws and different licenses. If that’s a problem, this is really a problem.

From a global competitiveness standpoint, I would certainly hope that they’re able to federalize this and preempt it. I don’t know if there’s too much water under the bridge or not. I don’t know enough about what it takes in Congress to make that happen. But I think this is bad for the U.S. I think it’s bad for innovation broadly. It’s going to make it harder for startups to do things because they’re going to have to worry about all this stuff.

Yeah, it’s way worse for little tech, right? Smaller companies don’t have phalanxes of lawyers they can send out to comply. I think the other thing is, listen, we already have the Civil Rights Act, we have the Fair Housing Act, and we have the Americans with Disabilities Act. Of course, we don’t want discrimination. But this just seems like broad overreach.

17. The Colorado & California AI Acts

I don’t even know how you comply or enforce it. It just ends up bogging down the entire system in uncertainty and litigation. Again, it’s important to say this isn’t even about whether or not AI should be regulated. It’s just a question of who should regulate it. What we’re saying is that there is ample opportunity for this administration and Congress to get together and write legislation, to the extent it needs to be written, to provide a national framework.

These are inherently interstate technologies. There’s no way to keep them in a single state. Write a piece of national legislation that allows us to continue moving forward very quickly, but at the same time addresses any of these concerns. I frankly think we need a moratorium on all state laws. Postpone all state laws until the federal government has time to act.

And if states are going to pass these laws, Bill, then I wonder whether or not OpenAI or some other company should consider blocking the citizens of those states until it’s resolved at the national level. Somebody needs to get the attention of these states that they can’t do this on a state-by-state level. It’s bad for the companies, it’s bad for the country, but hopefully we’ll get action out of Congress soon.

I think there’s good momentum. We almost had it passed as part of the big, beautiful bill, and I think there’s a lot of movement afoot in order to do it. I wanted to highlight it because I think it’s one of the high-priority issues facing the new Congress. All this AI all the time stuff, Bill, and you pinged me and said, “Hey, I want to talk about stablecoin,” right?

18. Stablecoin Surge: $18T in Settlements

We have this parallel development in the world. So I think if we have 3 major trends in the world, AI is clearly the largest supercycle going on. The reindustrialization of America is massive—all these critical supply chains. I would say the 3rd one is the digitization and tokenization of finance.

You have to fill out pages and pages of crap. Sometimes you get a verbal. We are so behind. I’m going to go out on a limb and say I hope whoever agitated for this to happen in the Trump administration is someone who is kind of caught in a regulatory-capture position, getting lobbied by somebody.

19. From PIX to FedNow: The Future of Digital Money

I think if the Trump administration studied this, they shouldn’t be critical of Pix. They should be envious of it. We should have done this a long time ago with FedNow. But we may be on the verge of stablecoins just being able to do this anyway. The rails have tons of transactions on them, as you’ve talked about.

We have this interesting situation where Coinbase and Circle have done this deal where Coinbase will allow you to earn 4% on your stablecoin balance. To get that kind of return at another bank, even a neobank, you have to have your direct deposit go there. Here, whether it’s $10 or $1 million, you put it in stablecoin with Coinbase and start earning 4% daily.

On top of that—and this gets back to the Pix thing—you can transact immediately out of that account. You don’t have to move it from your savings to your checking to get it to do ACH. You can send stablecoin immediately, in milliseconds, and it’ll cost you a few pennies. The rails are there. They’re ready, and it’s working.

I think the UI is a little difficult, but there’s no reason why that won’t get better and faster. So I look at this, and I just wonder what the team at Meta—they might just be kicking themselves. With all the money they spent on that coin and everything they wanted to do on WhatsApp, they should be running back at it. I don’t know.

Well, I think so. Maybe they should remember that the guy who did the Libra network, David Marcus, has started Lightspark now. It’s a startup. Maybe Meta should go buy Lightspark and bring David back in-house, because all the things they talked about are now what’s happening, Bill.

And let me tell you one thing: I think you’re onto something big here. One of the challenges we still have is that anybody who looks at our current system knows it’s dreadfully behind the rest of the world. We know it’s the result of regulatory capture by not only our issuers, but the banks and everybody else who benefits from the status quo.

If you look at Visa and Mastercard today, I think they’re doing something like 50,000 transactions per second. I checked with our good buddy Vinny Lingham, and he said that on both Solana and ETH today, they’re still under 4,000 transactions per second. They’re trying to come up with solutions to actually make the rails have the functional throughput and efficient settlement required to really become a consumer product.

20. Coinbase, Circle, and the 4% Yield Revolution

But I think you nailed the other one. Patrick Collison had a tweet on this that I replied to, which is that when the GENIUS Act was passed, there was massive lobbying by the banks to prevent the crypto companies from paying interest on stablecoins. The settlement was that they could pay rewards, not interest.

The way in which it’s manifested itself—because Coinbase is not the issuer, Circle is the issuer, and they did this deal—is that Coinbase is promoting it as though it was interest. From a consumer perspective, a reward and interest, if it’s 4%, are indistinguishable. No doubt, right?

Obviously, Brian Armstrong of Coinbase has been out on X arguing his side of the argument. So he’s clearly either getting opposition or expecting opposition on the regulatory front. When I see whether it’s Visa or Nasdaq or any of these people running at tokenization, I always worry.

If you look at the history of the debit card versus the credit card, it was supposed to be disruptive. It was supposed to be an alternative that would change things, but they just run at it and strangle it and mix it up a little bit, and then it’s not as disruptive as it was. That’s their go-to move.

When I read this thing on Pix again, I’m going to read this out loud: “As part of its aggressive economic and political campaign against Brazil, the Trump administration is investigating Pix, accusing the payment system of unfairly undercutting U.S. financial and technology companies like Visa and Apple.”

I mean, that’s the most absurd thing I’ve ever heard. Undercutting Visa? Do they realize Visa and Mastercard have, like, the top 2 operating incomes in the history of American business? There’s no one that needs less protection than these guys. If anything, there should be an investigation.

Dinner at the White House?

21. Why Visa & Banks Fear Stablecoins

Cabal? I don’t know. It’s so bizarre to me. I’m thrilled to see this kind of disruption. I think what’s possible is super interesting. I suspect all the big guys should be paying attention to this—Apple, Google, Amazon, anybody that might have payment on their rails.

I’ll go out on a limb, Bill: you’re going to see the hyperscalers. You’re going to see Amazon and Meta and these guys back involved in the stablecoin business. At the end of the day, we know money is a network-effects business.

The challenge of Circle and some of these stablecoins, from a consumer perspective, is that Visa and Mastercard are universal.

So you’ve got to get to all the merchants. Well, who has all the merchants? Amazon and Meta. Right? And so I think they’re in a great position to partner with or do some of these things themselves. Clearly, they have the instinct to do it. That’s why they did Libra in the first place.

It’s also amazing for innovation. One of the reasons why I think those bigger companies should run at this is, if you look at the history of the Pix-like alternatives I mentioned in the UK, China, and India, the startups that do financial innovation scale up way more aggressively and successfully on those rails that are cheaper and faster. If anything, having more rigid, high-friction, high-transaction-cost rails makes it harder for a startup to think about using one of those technologies.

And so the success of WeChat Pay and Alipay, which, as I described from my China trip, are universal—they’re the only way people pay in China—happened because of that government instant-pay product, not in spite of it. And it’s the same thing: I talked to the CEO of Nubank, and he said Pix was huge for his business. So it’s probably bad for a laggard bank, but for a bank that embraces it, it just becomes a better feature. And I think the same thing about Coinbase and what they’re doing here.

So I’m applauding the innovation. I’m jumping on board the crypto train, and I hope the incumbents aren’t able to strangle this thing in Washington.

Hear, hear. Hear, hear.

As we move toward the end and talk about my book and what I’m going to do next, I do want to share with you that both my book and the next project are outside of what I’ve spent my career doing. And as I mentioned, that’s kind of moving outside my comfort zone, but it’s also trying to have an impact and give back in areas that I don’t know as well, with a hope toward having an impact.

22. Brad on Invest America (Trump Accounts)

I’ve said this to you before, but I’ve been inspired, frankly, to go do this based on your success with Invest America. When you first told me about it, I had doubts that you could get it done—real doubts. And I’ve watched other people in your shoes try to do these types of things over decades and be unsuccessful, so you made it look easy. I know it’s just getting started, but I wanted you to know how much that inspired what I’m going to go do. Could you give us an update on where things are?

Well, that means a lot, Bill. Maybe, first, just as a reminder: I think you and I agree that we kind of have this battle for the soul of America when it comes to capitalism right now. And that’s fundamentally because too many people feel left out and left behind. Sixty percent of people will never own assets that compound. Likely Mamdani is winning the mayoral race in New York City, and they’re doing it by being anti-capitalist.

But if you look at these 2 charts—I’ve shown them many times before—they just show you that free-market capitalism is the most productive force in the history of the world. This first chart just shows that GDP on a global basis went parabolic at the exact same time that capitalism was really introduced and started taking off. And as a reminder, GDP is important because it’s that surplus for humanity, with a fixed amount of labor and capital, that then leads to better schools, better hospitals, drugs that save lives, and all the things that make our lives better.

You can look at this chart that shows the results: fewer mothers die in childbirth, average life expectancy is extended, quality of life is higher, and literacy rates are higher. Bill Gates expounds on this in his annual letters. So it’s not just an investment account. This is really a much, much bigger battle over where we want the country to go.

I was very concerned, as you know, a few years ago that we were headed down this path. And the path is that you can’t have so many people left out and left behind. So I think the answer to socialism, which has not worked for Europe—Europe is in a disaster relative to where it was 30 years ago on a global competitive basis—and China, as you well noted, has pulled itself out of poverty by leveraging capitalism. So the answer to this drift into socialism is more capitalism.

The Invest America accounts, now known as the Trump accounts, are more capitalism. They make every child a capitalist from birth—a private owner—and give them $1,000 in a 401(k)-like account that they own and control, which their family has on their phone. And so I think that’s a game changer, but you’re right: we just got it passed.

So where are we now? Treasury Secretary Scott Bessent has to implement this. And by the way, this is one of the largest consumer launches in the history of government. As of today, there are 65 million kids in the country who qualify for an Invest America account: every kid under the age of 18. And every kid under the age of 2 will automatically get $1,000 in their account.

23. Implementation Timeline & Treasury Update

You’ll probably hear about a launch starting maybe in early December. We’ll launch the website, and people can sign up for this. Remember, the accounts have to be funded and established by our 250th birthday, July 4, 2026. That’s only 9 months from now.

I can tell you I’ve been blown away by Secretary Bessent, Assistant Secretary Luke Pettit, and the team at Treasury working with the White House. They’re attacking this the way I would expect a Silicon Valley startup to attack the problem. They’ve gotten a great group of technologists, with Joe Gebbia, of course from Airbnb, helping to design the front end of this.

We’re on the verge now of some major announcements, where people can start signing up their kids. The goal is that once we launch this, all these kids will have these accounts. They’ll be able to roll them over into their favorite bank, whether it’s Schwab or Fidelity or JPMorgan or what have you.

And starting on July 4 of next year, Bill, we’ll have as close to automatic account creation as possible. You have a child, the child is born, they get a Social Security number, and they get an account seeded with $1,000. From a kid’s perspective, it’s going to look like, “I own a little bit of Microsoft, and I own a little bit of United Healthcare and Nvidia and whatever.”

We’re going to be able to teach this. In fact, your buddy Tim, who’s teaching financial literacy—you know, we now have 30 states requiring financial literacy. We’re going to have this embedded in the schools. Every kid’s going to have this on their phone. At any rate, it’s going incredibly well. I give them a very high score, but we’ve got to get it done.

So let’s talk about your book.

24. Bill’s Book: Runnin' Down a Dream

I’ve been a huge fan of the speech you made on this, but I love the title, “Running Down a Dream.” What’s the thrust of the book? I remember that lecture, but what really compelled you to write it?

So, years ago—and this is probably going back 10 years—I was reading a lot of biographies, and I noticed certain patterns among people with extraordinary careers. As VCs, we see a lot of patterns in businesses and pattern recognition, and I just saw patterns with people. In the back of my mind, I always wanted to do this presentation. I kept notes on it like I would an unwritten Above the Crowd blog post.

I had an opportunity—I got invited, had an opportunity to give the presentation to the MBA class at the University of Texas—so I worked on it and put it together, made it nice, and gave that presentation. They ended up putting it on YouTube, and many people have come to me and said that it changed their lives and encouraged them to do different things.

Certain people in the media industry noticed. David Senra, who has the new podcast where he interviewed Daniel Ek and Mike Cole, is a big fan of the presentation and talks about it a lot on his podcast. James Clear, who wrote “Atomic Habits,” maybe one of the best self-help and personal-development writers out there, retweeted it and put a transcript on his own website.

And then a few people who are influential in my life started prodding me: “Hey, you should turn that into a book.” Eventually, I got convinced. We talked to publishers, they were interested, and so I started working on this.

It took a long time. The thing I would say about it is that I hope that time equates to quality. I was out there really wanting to make it great. And so the book has an interesting, novel architecture. We combine what I call profiles—stories of success—with principles, tools of success. They alternate. You get a story, maybe about someone you didn’t know, how they started at the very bottom and became successful, and then the types of things they did.

I do think that the principles, these tools that are in the book, are things people can use. And I really wanted it to be great. It’s done. I still need to record the audio version. All the podcast fans of ours tell me I have to do it. It has to be my voice, so I’m going to do it. I hope people love it. I hope it changes their lives for the better. And what I really want to encourage is people to take a chance and do what they really love.

25. The Great Career Reset: Passion vs. Grind

I think it’s such an important topic. One of the things parents are asking me so much these days is, “What should my kids do?” Particularly given that there’s a lot of anxiety in the world right now about future careers. So maybe just talk a little bit about why this is so important now, because I think the timing is really profound here.

So, in the introduction chapter, we unpack a lot of this. And I don’t think anyone would be surprised when I read some of this, but Gallup does a career-engagement study.

They’ve been doing it for a long time. I think in 2023, only 23% of people said they were thriving or engaged at work, and 59% were unsatisfied. That’s just a big universal survey.

Everyone seems aware that we’ve moved to this gauntlet that we’ve put our kids in as they approach college and go through college. In The Coddling of the American Mind, Haidt and Lukianoff called it a résumé arms race. We’ve really taught them to be grinders, but Angela Duckworth highlights that if you have persistence but not passion, you eventually recognize you’re in a grind. When you come out of that, you’re in a really tough spot.

We’ve got these kids on this runway. We’re telling them they have to pick a major even in their applications. They’re 17: What do you want to do with your life? What do you want to do? They really don’t know.

One thing we stumbled upon doing research for the book—I was working with a researcher—was a survey asking people, “If you could start your career over again, would you do things differently?” In that survey, 70% of people said yes. We did it again with Wharton, just to make sure we had a true academic survey, and they surveyed a lot more people. That number was still 6 in 10. Six in 10 said if they could start their career over, they’d do it differently.

26. Bezos’ Regret Minimization Framework

There’s a great book I read while writing my book called The Power of Regret by Daniel Pink. He’s a well-known author, and he has this thing he calls boldness regrets. He said, “One of the most robust findings in the academic research and on my own is that over time we are much more likely to regret the chances we didn’t take than the chances we did.”

He says, again, that the surface domain—whether the risks involved are education, work, or love lives—doesn’t matter much. What haunts us is the inaction itself. I think that ties really nicely with this idea that if people could start over, they’d do something different.

There’s a great video that Pink references that we can put in the show notes, where Jeff Bezos is asked about the decision to leave D.E. Shaw and start Amazon. He said he used a regret-minimization framework. He said only a nerd could do that, but he imagined himself being 80 and asked whether he would care that he left D.E. Shaw, maybe forgoing a bonus, or whether he would care more that he didn’t take this chance—this instinctive chance that he felt like he had to take.

Immediately after thinking about it in that way, he wanted to go do it.

Hey, Bill, I have Bezos’s regret-minimization framework taped to my computer monitor.

There you go. I didn’t even know that.

Literally taped to my monitor. Powerful.

Yeah. That’s what this is about. That’s what this book is about. That’s who it’s for. I want more people to take a flyer and go do what they love.

We have a phrase I use in the book: “Life is a use-it-or-lose-it proposition.”

Totally. One-shot deal, man. How do you plan to promote it? It comes out in late February.

I’m just getting started in that process. If people have ideas they want to share with me, please reach out and let me know.

I’ll host a book-launch event, Bill.

27. Who the Book Is For

Okay. We’ve got a lot of fun stuff planned, but I need to record the audiobook. I know that’s going to take a lot of time, and I’m excited.

There are a handful of people who have read it—maybe 50 people at the publisher and elsewhere—and nearly every one of them tells me that they immediately thought of 3 or 4 people they want to give it to. I hope there’s a viral component to it because people have that reaction.

If you’re feeling stuck in your career, you should read it. If you’re a teenager or young adult who feels overwhelmed by people telling you, “What do you want to major in? Where do you want to go? What are you going to do with your life?” I think this book won’t put more pressure on you. I think it’ll actually relax you and give you a framework that feels a lot more personal and makes you feel a lot more in control.

If you’re a parent who wants to help a child on that journey, I think parents sometimes overly push kids into the lawyer-doctor-banker framework. I’m not sure that’s healthy, especially in this AI world, where those jobs may be at risk as well. Maybe if you’re an administrator or someone in the type of role that guides people in career decisions, hopefully you’ll like it as well.

I think it’s going to be hugely impactful, and it’s the type of stuff of consequence. I’m thrilled that you’re doing it. I’m thrilled that you’re taking the time to do it.

I know that’s not the only thing you’re thinking about. You have these other big topics that you’re thinking about as well, Bill. We’ve talked about them here, and I’m sure we will continue to: regulatory capture, U.S. health care, nuclear, et cetera. What are you thinking about with respect to those things? Are you going to write a book on every one of them?

I could, but it’s not my goal. My goal is to just spend more time on these really big problems and see if I can be helpful in any way. I’ve spent a career breaking down and analyzing different situations.

Two of my favorite podcasts we’ve done—and I hear about this from the community as well—are the one we did at Diablo Canyon and the one about my China trip. Those types of projects are very rewarding for me, but they were also learning expeditions. I went out and put in more hours for those episodes than for others.

Right. The nuclear one in particular—one of the things people ask me is, “Why would I want to go do this?” We were a small part of a movement to change the mindset on nuclear energy. There were people who put a lot more effort into it than we did, and I’m not trying to take credit for it.

But the fact that Steve Pinker was out there, Elon Musk, our stuff—it eventually happened. Seemingly overnight, we went from a very negative mindset toward nuclear energy to recognizing that it’s very clean energy and something that can really help save the planet.

That type of meme flip—if I could go achieve more of those in these other areas—I would consider it a win. That’s what’s motivating me. I’m really looking forward to it. I’m fired up and nervous at the same time, but that’s what I’m thinking about.

There’ll be more to come on that in terms of what the actual platform looks like. I’m still working on it. For now, I’m going to sprint into February to make sure the book does well.

I couldn’t be more stoked for you. This has been a total blast. You and I have been chopping it up for a couple of decades, but doing this over the last 2 years together and pounding these out has been a lot of fun for me. I’m sure we’ll continue to chop it up every day, and I’m sure you’ll find some topic that you can’t live without exploring—

Can’t resist talking about.

Exactly. So we’ll get you back on. But I’m going to give you the last word, Bill. It’s awesome to hear about all of this, and I’m super excited for the book.

I would just end the way I started, Brad: thanks to you. It’s been fun working together on this and doing it every week. It does force you to stay fresh. You have to read everything you possibly can, which I’m sure is super helpful to you as an investor.

And thanks to all our listeners. I’m sure some of them are going to feel like I’m letting them down, and I feel the weight of that. But hopefully they’ll recognize that I’m going to go try and put my work effort to good causes.

As a reminder to everybody, these are just our opinions, not investment advice.