Our Seven Trillion Dollar Future: Dave McClure & Aman Verjee Burst the AI Pessimism Bubble

00:00:31 Andrew Keen: Hello, everybody. It's Monday, August 17, 2026. Yesterday, we did a show, That Was the Week show, Tech Summary, with my old friend Keith Teare based in Palo Alto, And we talked in part about whether or not we were in an AI bubble. There was an interesting piece this week in the New York Times about the way in which, the AI companies and the big tech companies seem to be conspiring to, if not create a bubble, certainly the illusion of, massive economic growth when they're perpetually feeding themselves. And two people who have been on the show before to talk about economic bubbles are my guests today. They're the partners of Practical Venture Capital, two very seasoned investors, entrepreneurs, Dave McClure and Aman Verjee. Dave is talking to us from West Virginia, from Morgantown in West Virginia. And, Howdy. Aman is in, Atherton. We were joking before we went live, guys, that you couldn't be further apart geographically or symbolically. Aman, isn't, Atherton the most expensive suburb in America?

00:01:48 Aman Verjee: You know, I have not seen the latest rankings. I wouldn't be surprised if we were, we were up there. It is a lovely it is a lovely place to live. Yes. Right here right here in, Northern California.

00:01:59 Andrew Keen: And how's Morgantown, Dave?

00:02:01 Dave McClure: It's not quite Atherton, but this is my hometown. I'm back here visiting family for a little bit.

00:02:06 Andrew Keen: Maybe that should be its tagline, not quite Atherton.

00:02:11 Dave McClure: Yeah. Over the years.

00:02:14 Andrew Keen: Yeah. In terms of the manifestation of economic bubbles, I know, Aman, your new book on, economic bubbles is out this week, A Brief History of Financial Bubbles. I wonder whether the evidence or the forewarnings of bubbles, the auguries of bubbles are better seen in places like Morgantown or Atherton. In other words, in wealthy neighborhoods or less wealthy neighborhoods. Maybe, Aman, you could, address that first.

00:02:46 Aman Verjee: Well, I think the, I think the nature of the bubble tends to tends to run in asset classes. It tends to be a big the way that I define it, and maybe I think Keith defines it a little bit differently your guest from yesterday, is just a big run up in some kind of an asset price. And then, over a very short amount of time, it runs back down again. So it's relatively symmetric. It's a symmetric, up and down in a price of an asset, and it doesn't last. And it's, usually typical of a asset as opposed to, you know, income or job activity or other things like that we can have knock on effects. So where you have assets, where you have wealth, that's usually where you see the behavior, the excitement, and some of the things that go along with the bubble a little bit more than what am I what am I call Middle America. You know, places where, you don't have the same amount of assets and where stuff is tied to more, like, the real economy.

00:03:36 Andrew Keen: Dave McClure, you've been through this before. We all have. We're not new to this rodeo. We were all around in the nineties when the dot-com—

00:03:42 Dave McClure: Yep.

00:03:46 Andrew Keen: —boom busted. Are we in a bubble, Dave, when it comes to AI?

00:03:51 Dave McClure: I don't think we're in an unusually large bubble. I think that, actually, there's several companies that really are making quite a bit of revenue, if not also profit. And so I think the, you know, increases in valuations, at least for NVIDIA's for certainly for real, public tech companies, I think, are generating substantial revenues and profits. And even for the private companies, I think we're seeing SpaceX and Anthropic and OpenAI grow revenue at pretty, crazy rates. And, you know, there might be a question about whether that revenue is profitable or not. I think more and more we are seeing that it is profitable. You know, that might be a question of, you know, amortization over time, how much the CapEx that's being spent is, is profitable or not in the future. I think there's probably a lot of other companies that are draping themselves in AI sort of magic sauce and, you know, getting valuations that are perhaps optimistic. And those companies, you know, that's where I see there's the potential popping of a bubble in the future if they if they aren't really AI companies, and I don't think all of them are. Then we might see some, you know, decrease in those valuations in the future. But I think there's definitely some very, very large companies. There's incredible growth in revenue, certainly in Anthropic, you know, historic growth, and more and more potential for profits as well.

00:05:16 Andrew Keen: Aman, you and, Dave were back on the show, as I said, in, December of last year. So Yeah. Seven or eight months ago. And we talked about, quote, unquote, this was the title at least of the show, the naked truth about Elon Musk and Sam Altman. You guys know both. You're both, part of the old certainly, I'm mine. You're part of the old PayPal mafia, so you know these guys very well. And I know, Dave, you also know them very well. It's interesting that the one guy missing from that is the Anthropic CEO, Dario Amodei, who at that point seemed a minor player, a bit player in the Sam Altman, Elon Musk drama. Now he's the main he's the main guy. Is that fair, Aman? Does it speak of how quickly things change? Eight months is like eight centuries in Silicon Valley.

00:06:09 Aman Verjee: Yeah. Absolutely. I know we, we didn't spend a lot of time talking about him specifically, although we did talk a bit about Anthropic. And I think, like, Dave, I share the view that while there are a lot of companies that are draped in that AI magic, And a lot of them won't make it, and a lot of companies will, you know, will, will go by the wayside. The, the macro environment, I was still very positive on. And what we did talk about was there will be lots of lots of folks who are lots of great entrepreneurs building great businesses and that, that rotation will change. I think about nine months ago, the consensus view was that SpaceX seems like it's on track to have an IPO in 2026. Although, we're talking about a trillion dollar IPO back then. They had just raised at $400 billion, or doing around $800 billion. And, Sam seemed to be the king of the world, and OpenAI seemed to be the had the lead on the consumer side. And then I think what happened was Anthropic earlier this year kind of found a niche with small and medium sized businesses, and they focused on coding agents. And that became the source of revenue for, really, the entire, AI application layer and the model layer. And their revenues are on track if you they just filed, this weekend or at the end of last week to go I guess they have been on file to go public, but they leaked their evaluation numbers at $2 trillion for the IPO.

00:07:28 Andrew Keen: Yeah. Quote, unquote, leaked. I'm sure it was Yeah. Very an a very unintentional, intentional leak. Yeah. There was a piece, yesterday in Reuters about apparently, their valuation hinges on between $190 and $200 billion, 2028 revenue forecast, which—

00:07:46 Dave McClure: Which might be conservative, actually.

00:07:49 Andrew Keen: Yeah. You think?

00:07:50 Aman Verjee: They could be at a 120, then if at the end of this year and don't have to cancel the year and year [unclear].

00:07:57 Andrew Keen: So, Dave, what does this tell us then? Eight months ago, we talked about Elon and Sam. I mean, they're not they're both very significant now, of course. They're companies. SpaceX went public. OpenAI is about to go public. What does this tell us about the market, Dave, that suddenly in this sort of Silicon Valley wacky races, Anthropic seems to have pulled ahead?

00:08:23 Dave McClure: Yeah. Well, I think probably sometime after when we spoke, between Q4 and Q1, things really changed, at least considerable change in the fortunes of OpenAI and Anthropic. I think OpenAI probably had a lot of challenges. It seems like they're past a fair amount of that, now. But I'm reminded of the famous quote, there are years when nothing happens and weeks when decades happens. And I think—

00:08:49 Andrew Keen: Yeah. Lenin, I think, said that.

00:08:52 Dave McClure: Yeah. So I think somewhere in February, March, April, there were probably several weeks where, a decade worth of progress happened. But I think, Aman is correct that, you know, it seems like Anthropic made a pretty good bet on focusing on coding platforms and agents. I was kinda hearing that might have come from watching Cursor and seeing what was happening at Cursor around the same time. And, notably, Cursor was just acquired by SpaceX. I think that closed on Friday. And, you know, I think across the board, we're gonna see acceleration of revenue in all three companies, SpaceX, OpenAI, and Anthropic. And so that's why I really don't think it's as much of a bubble as people are worried about. Again, there's possibly other, you know, companies where they might, you know, stumble and fall and, you know, we'll actually probably see some of that in more traditional areas that are SaaS companies. I think the SaaSpocalypse also happened during that time frame possibly as a result of all the progress that was happening.

00:09:52 Andrew Keen: And SaaS, you might explain to our nontechnical—

00:09:56 Dave McClure: Well, SaaS was really cool ten years ago. It became decidedly a lot less cool.

00:10:02 Andrew Keen: Cloud companies. Is that another way of putting it?

00:10:05 Dave McClure: Cloud computing, generally speaking.

00:10:08 Andrew Keen: And so like Salesforce?

00:10:11 Dave McClure: Yeah. But I think the concern or issue there is that the cash flows from SaaS companies are now perceived to be a lot less durable as a result of disruption by AI companies. And so we're seeing, you know, tremendous acceleration of revenue by Anthropic and SpaceX and OpenAI, but we're also seeing them encroach into, you know, areas of business that you might not have expected previously. So design and legal and medical and other areas. And as they make announcements about each of those verticals, you see knock on effects in those traditional verticals. I say traditional, but these are really, you know, tech software businesses in those verticals, that get their, you know, potential future growth rates, cut at the knees, and that's probably having a pretty big impact on how people think about investing.

00:11:01 Andrew Keen: Aman, you, as I said, you're part of the PayPal mafia, so you know Elon quite well. Is SpaceX an AI company? I'm not really I'm not really sure I know the answer to that.

00:11:12 Aman Verjee: You know, at the start of the year, I would have said not really. We actually underwrote it, and we made an investment in it as a as a firm, in part just because the we knew the founder and we knew of the company historically, but they're really about putting rockets into space. And if you'd asked— if you'd asked me back then, I would have said, Elon has a separate xAI AI company, which was primarily Twitter, which is primarily Grok, their open source platform model. And they were working on things that were tangentially interesting to SpaceX, but not integral with the business. And then they merged with, xAI at the beginning of the year. And if you look at their revenues, at least, Elon is guiding to $100 billion in ARR at the end of the year. And probably he hasn't said this explicitly, but I think somewhere between $50 and $60 billion of that $100 billion is now from the AI business. And only and probably less than 10 will be the space launch business. And then there's another, you know, $40 billion in Starlink, which is the connectivity business, which integrates both. So I think it has now become on the roadshow, they did use the word, they had a very you might remember the S-1 that they wrote and then leaked. And then I think for the first time, maybe in history, your wife would know better because she's the she's the securities lawyer. I'm not I don't remember a company putting forward looking projections into an S-1 or [unclear]. I was told by my lawyers when I was a chief financial officer, don't you ever do that. You will get you will get sued. But in that three to eight page document, they did put out three year revenue projections, and they talked about the revenue track of being more than $300 billion annually for that company. And over half of it is in AI. And if you just count the number of times they use the words artificial intelligence in their prospectus, I think it's over 1,100 times. I didn't count it myself. I used I use AI to count it, but—

00:13:03 Andrew Keen: They probably use AI to write it.

00:13:07 Aman Verjee: They, they describe—

00:13:09 Andrew Keen: Probably Grok. I hope they use Grok rather than Claude.

00:13:11 Aman Verjee: Well, they probably the thing that they do, though, if you think of the AI stack as being models and applications at the top, so things like Cursor, which is part of now SpaceX and generating meaningful revenue. There's LLMs like OpenAI, and let's not count Sam out yet. He will be back, I think, with another—

00:13:26 Dave McClure: Not— not at all. I don't think we should at all.

00:13:30 Aman Verjee: And then you've got all the infrastructure. You've got the compute that's required to power it all. And it's the entire the entire Anthropic business runs on somewhere between 1.3 and two gigawatts of compute and capacity. And for that, they're supporting, what, $70 billion in revenue. SpaceX is now one of the premier builders of AI compute. They as a part of the acquisition, they have this computer called Colossus in Tennessee. They've built Colossus two and three. They're building out these data centers, and they are now supporting almost two or will be at the end of the year supporting almost two gigawatts of capacity. And if you listen to the earnings call, their first earnings call, their CFO, Bret Johnsen, talked about going to from two gigawatts of capacity between five and ten. I think Elon said closer to 10 than five. So let's give them the benefit of eight gigawatts of capacity. They could be powering much of what Anthropic and what maybe OpenAI and with Grok and what some of these open source models do.

00:14:23 Andrew Keen: So didn't Elon say a few weeks ago that eventually SpaceX be worth more than the entire Earth or something slightly Elon-esque like absurdly Elon.

00:14:34 Aman Verjee: He says a lot of things.

00:14:36 Dave McClure: Elon's sort of notorious for saying outlandish things and setting, you know, things like your targets.

00:14:43 Andrew Keen: Did I really Yeah. Aman said that, you guys invested early at Practical Venture Capital in, you know, in SpaceX. Are you still confident that you hold on to your shares? I assume you've made a tidy profit out of that. How early did you invest?

00:15:02 Dave McClure: We got into a position, I think, probably in '22. I'm on I'm trying to remember the exact time frame. I said about '72. [ed.: likely "'22"; McClure has just placed the position in 2022]

00:15:08 Andrew Keen: Shut everything down and go and live, go and play around the island, son?

00:15:13 Dave McClure: I mean, it's not the entirety of our portfolio, but it is now the biggest position in our second fund.

00:15:19 Aman Verjee: You can't sit on your wins there, Andrew and Dave. You got to quote Bill Belichick, it's on to Cincinnati. But Dave has had some pretty big wins. I don't want him going anywhere yet.

00:15:26 Andrew Keen: At least on to sort of going. West Virginia. I are you still confident? I mean, holding on to it? The markets it seems like the public market's ambivalent at the moment about SpaceX. It seems to go up and down.

00:15:40 Dave McClure: I will say I was a lot more skeptical probably six, nine months ago than I am right now. I think if we looked at the recent earnings report that came out from, from SpaceX, they kind of, crushed the numbers pretty much across the board. I don't think we were surprised that they were doing well with Starlink. That's always been, you know, sort of what's been working at the company. And I think there's, you know, maybe some caution about the CapEx spending that's going on in the AI business. But what was really, pretty amazing, I think, if you look at the moves that have happened in the last six, nine months, you know, from the acquisition of the xAI business, which a lot of people were pretty skeptical about, you know, whether that was a real business or not. They bought it for $250 billion. That was largely a private transaction that Elon approved as, you know, the owner of SpaceX. And, what happened after that was, you know, a lot of, you know, the people that were at xAI left. There was some concern about whether xAI was really, doing well or not. They subsequently, you know, made the deal with Cursor to acquire Cursor. They set up two deals, very big deals, renting out compute to Anthropic and Google. And all that happened probably in a matter of maybe three to six months. I'm not sure of the exact time frames there, but pretty much since the beginning of the year and, you know, by June, all those moves have been made. And I don't think I've seen anybody pivot a company, that quickly at that size in such a short period of time. And if you'd ask me, you know, whether the company is worth $1 trillion at the end of last year, I would have said probably not. If you'd ask me today, do I think the company is worth $2 trillion, I'd say, I'm pretty sure it's gonna get there by the end of the year. If we if we believe the projections of revenue that, Elon is talking about right now and where they expect to get to by the end of the year, I do feel like the valuation is sustainable. And, really, that's a pretty big change in my opinion in the last six, nine months.

00:17:49 Andrew Keen: When the facts change, one's opinion change. Dave, I know you're—

00:17:52 Dave McClure: I don't think people really understand how much of a pivot that was. Like, from thinking about xAI as, you know, a frontier model competing with Anthropic and OpenAI, seeing that kind of blow up in a way, and then acquiring Cursor, renting out compute to, you know, previously what you would consider to competitors. You know, now all of a sudden, SpaceX valuation is actually based on Anthropic's progress.

00:18:23 Andrew Keen: Yeah. Which also speaks of the circular nature, certainly the somewhat incestuous nature of this economy. Maybe, Elon's middle name should be pivot. I know, Dave Yeah. You're probably you and Aman don't always agree on politics. You're probably more in my camp. You're not—

00:18:39 Dave McClure: That's the statement of the century.

00:18:42 Andrew Keen: Yeah. You're not a big fan of Elon's politics. I'm not sure if Aman is—

00:18:46 Dave McClure: I am not.

00:18:47 Andrew Keen: Certainly, you're less of a fan than Aman. You and I are less of a fan. But I wonder, how do we square the fact that Elon is clearly a financial or entrepreneurial genius, and at the same time, his politics are incredibly dubious. Do these things is there something unnatural about that, or do we just have to accept that there's a good and a bad Elon?

00:19:13 Dave McClure: I'm not gonna make any assessments of good or bad. I'm just saying that's not, you know, my preference that he'd be involved in politics, and his political moves are not my favorites. I sometimes share a similar view of my ex-boss, David Sacks. I think he's—

00:19:25 Andrew Keen: Right.

00:19:26 Dave McClure: Brilliant investor, brilliant entrepreneur. I don't agree with his politics.

00:19:31 Andrew Keen: Is he still the AI czar or just—

00:19:34 Dave McClure: Well, he's no longer in the administration. He's back, I think raising a new fund, in fact, for Craft. Full disclosure, we're LPs in Craft as well. So, you know, I think you have to sort of maybe, distance the art from the artist and look at the work. I think I know. I own a Tesla. I've, you know, been a subscriber on Starlink, and I do think that Elon's products are terrific, and I think he's an amazing entrepreneur. I don't ascribe to his politics, and I'll just leave it at that.

00:20:10 Andrew Keen: Aman, what about Sam? Let's talk about OpenAI. There was a headline last week in the Financial Times. Yeah. OpenAI upheaval mounts. I mean, it seems to be mounting now for months as Sam readies the IPO push. So many senior people have left. You nodded when, when, Dave said that we shouldn't write OpenAI or Sam off. But should we be a little concerned? I mean, you're a you're a SpaceX investor. I assume I mean, you your business is in the secondary market, so you must deal with a lot of OpenAI equity. Why shouldn't we be concerned with all this upheaval at OpenAI?

00:20:53 Aman Verjee: Oh, I think you should be. I think they have gone through they one of the reasons I was a little more ambivalent on them, at the start of the year, although I thought they were they did have that leadership position in the, in the consumer side, was the that there was a lot of management change. And, the number of people on that management team that have been moved off or, you know, sidelined or moved into different roles has been, has been startling— in addition to just all the board drama from a couple years ago, you know, they've had folks like, Kevin Weil resign. Bill Peebles left the company. Brad Lightcap was the COO. He's doing something else.

00:21:33 Andrew Keen: Yeah. He announced his departure last week.

00:21:35 Aman Verjee: Yeah. Sarah Friar, the CFO, who I like a lot. You can just tell slight, you know, disjointedness between her and Sam as they're talking to investors. I don't think there's anything, you know, critical there. But when you lose founders like Ilya Sutskever and Mira Murati and John Schulman and Andrej Karpathy and then your CFO and CEO [ed.: likely "COO" — Lightcap, named a moment earlier], I think there's a lot of new people. I guess what I what I'd what I'd describe it as is a it's a new team. And when a new team comes together, they're just, you know, they're still what's the term? Storming, norming, and forming? Whatever my, whatever that HR aphorism is. And so they're still learning how you know, like, who's got the ball, who's gonna play, who's got that coverage, who's gonna, you know, who's gonna, find that open space. Oh, wait. I thought you were supposed to pick that up. I didn't I didn't, you know, I didn't I don't remember how that feels.

00:22:19 Dave McClure: I think we made the, the note on our own podcast that there's been more changes at the OpenAI board level than there's been drummers at Spinal Tap.

00:22:28 Andrew Keen: Yeah. Well, Aman, you were on the show last year as well.

00:22:32 Aman Verjee: Yes.

00:22:32 Andrew Keen: And, one of the things we talked about was what Yogi Berra can teach, Silicon Valley about booms and busts. I wonder what Yogi Berra would say about OpenAI and all these changes.

00:22:44 Aman Verjee: He'd probably come up with something I'm a big baseball fan as you can see. He'd probably come up with something very, very witty about, you know, predictions being difficult, especially about the future.

00:22:52 Andrew Keen: Right. Exactly. But, I mean, Aman, you're not you're not a bum thrower. But in your own way, you're suggesting a deep concern about OpenAI?

00:23:01 Aman Verjee: I think you gotta be concerned about especially in an IPO where I've been through them before, and you just have to have the team hitting on all cylinders. And when there's a lack of coordination, if you even if you have a bunch of superstars, you know, it always takes a couple of games for the United States Olympic basketball team to beat France or whatever. I think there are 2024, '25 was just a lot of change, and then 2026 was a lot of their storming, norming, and forming. And then they missed the, the they missed the coding trend where we're just generating all the revenues. So they're now on top of that. I think they're new, you know, GPT-5.6 Sol and Luna and Terra are A-plus models. They still have a very strong lead on the consumer side. I would bet at the end of the year, you're gonna see them at around, let's say, $40 to $50 billion in ARR. Probably not profitable.

00:23:48 Dave McClure: Already there, Aman. I think they're already the numbers that we're Yeah. Yeah.

00:23:52 Andrew Keen: Is there room, though, in this market for three—

00:23:52 Aman Verjee: Not yet. [ed.: brief interjection; speaker attribution uncertain]

00:23:56 Andrew Keen: —dominant players?

00:23:57 Aman Verjee: No. I was gonna say it's a crowded lineup. We've seen we've seen great companies that have been around for a while doing it well, like Databricks just kinda bow out. And, see, we're not gonna play the IPO game right now. We have SpaceX, which is sucking up $75 to $85 billion of capital. That's already putting a big dent in the IPO-able market. Let Anthropic go next. If those two trade well and I think so far SpaceX is fine, you know, at or above its issue price volatility, but we'll see how Anthropic does. If we close out the year with some momentum in the IPO market, it'll break all the records for money raised. Then I think 2027 is gonna be a better environment for OpenAI to go out with a call it a $100 billion guide on their revenue number, a better definition around their business model on the coding side, some new consumer products, maybe not being as capital intensive. They'll have to figure out how to continue that growth rate. I think '27 is gonna that's probably their time. But, you know, it kinda goes back to you have to you have to find the right time, and the team's gotta be performing. And I think they've had a year now to gel with the news new the new team and the new managers. And, I think that if they can execute that plan, I think they'll be just fine. The valuation won't be where Anthropic's is, but I think they can get above a trillion dollars in the market at the right time.

00:25:05 Andrew Keen: Do you agree, Dave? I mean, it I get the sense just as an outsider that we were I mean, most people were expecting an OpenAI IPO this year, so Anthropic goes first. Will it seem like a defeat to OpenAI if they're third?

00:25:21 Dave McClure: I'm sure that if you were, you know, sort of watching a year ago, you might feel like, you know, OpenAI has fallen from some great height, you know, to second place and potentially third place if you're looking at SpaceX. But at the same time, I think Sam is a pretty savvy operator, also, you know, an amazing entrepreneur. I think the company has probably found its footing. It still got some, you know, things to figure out. Particularly, I would say, you know, a lot of the financial support that we're seeing this week, this past week from NVIDIA across the board, not any one company in particular, just broadly. That's gonna help stabilize a lot of the CapEx concerns and debt financing concerns that might have been troubling OpenAI. So my guess is I'd be I'd be very surprised if they don't have an IPO at least by Q2 of next year. In fact, I would guess it'll probably happen in Q1, and I would expect that'll be at least at a $1 trillion number. And probably you know, my guess is it'll probably be closer to 1.5. So we'll see what happens, but I have much less concerns about OpenAI than I did maybe, again, six months ago based on some of the recent progress and pivots that Sam has made. So, you know, I'll probably say this for a little bit of a hot take, but I'm gonna bet that Anthropic will be a $3 trillion company, SpaceX will be a $2 trillion company, and OpenAI will be a $1 to $1.5 trillion company by Q2 of next year.

00:26:41 Aman Verjee: Nice. Hey.

00:26:43 Andrew Keen: Aman, you're noting that those are—

00:26:45 Aman Verjee: I don't think it's a problem to go third, by the way. Do me a favor. Yeah. You and your you and your audience, do me a favor. Put these IPOs in order of timeline. Yahoo, Lycos, Excite, and Google. Which one do you think went fourth?

00:27:00 Dave McClure: Google was fourth for sure. Right.

00:27:03 Aman Verjee: Don't worry about don't worry about—

00:27:04 Dave McClure: They did fine.

00:27:05 Aman Verjee: They did fine. Only I've got Lycos not so much. So it's better to do it right than to do it first.

00:27:10 Andrew Keen: Okay. Well, then let me rephrase that given you threw out Excite, and even Yahoo. Who's the Excite in this market? We've been through this before.

00:27:21 Aman Verjee: You mean yeah. Go ahead. If you have a take, Dave.

00:27:24 Dave McClure: I'd have to think about that a little bit. I mean, it's not really the same parallel, but I think you're talking about, you know, people that might have been trying to solve problems with the previous generation of technology, and I think that's, you know, SaaS companies. So I think that's, you know, those are the companies that I would be concerned about in this market. I think again, you know, let's just kind of state for the record again. OpenAI is probably gonna be doing $40 to $50 billion right now, could be doing, you know, $100 billion by the end of the year. A company doing $50 billion in revenue growing at a, you know, two x or better per year, and, you know, we can talk about profitability there, I think is gonna do just fine.

00:28:02 Andrew Keen: Well, but the present Dave, isn't the profitability key? Because they're not profitable, and it requires more and more money for data centers, massive infrastructure. So however much—

00:28:13 Dave McClure: Question is, will it be profitable in the future? Right? Right now, growth is more of an issue than profitability, but the real issue is whether they could stand up the debt. You know, and I think that, you know, there were some interesting sort of ways that Sam was trying to play that game before, and Oracle maybe coming in, and, Stargate, I think, with SoftBank. But now that you're seeing across the board NVIDIA making moves to bring in a bunch of financial service providers, backstop the entire, you know, industry with $500 billion worth of capital, I really think that we're probably in okay shape. I mean, you could you could maybe be concerned about whether Anthropic's growth and revenue continues, but I don't see that being a problem.

00:28:57 Andrew Keen: So given NVIDIA's role, is there an element of circularity here, Aman? You're nodding that it's obviously in NVIDIA's interest that these companies survive because—

00:29:05 Aman Verjee: Yeah.

00:29:12 Andrew Keen: They're the ones who buy NVIDIA chips. Is there a circular element here?

00:29:18 Aman Verjee: Yeah. There's a there's a part of circular financing, though, that is, quite normal and reasonable in asset backed finance. So when I was the CFO of Sonos, as for instance, we were building hardware, consumer electronics. And we would work with suppliers, manufacturers, some of them are to China. And the way that we did it is we bought these things from China, and we had ninety day payment terms. And so the contract manufacturer would just be extending financing to us. And then we'd sell those Sonos systems through Best Buy and other, you know, channel providers, and we'd be getting thirty day payment terms. So we had a negative cash conversion cycle, which just means you're getting money from your customers faster than you're, you're paying off your CMs. And as long as, and that's not you know, if you think of, like I just heard the CNBC, special with Jensen and the six private equity players, and I got, you know, much more comfortable, I think, than I did, than it was before with those financing arrangements. But it's like buying a car. You can go to you can go to buy a, I don't know, a Corvette from General Motors. You can pay cash, but who can pay? Who can afford a $100,000 for a Corvette? You can get a loan from a bank, a third party bank on reasonable commercial terms. Maybe put $10,000 down, finance the other 90, or you can get a loan from GMAC, General Motors Acceptance Corporation. And when they extend you the money, as long as it's on commercially reasonable terms and third party terms, they can also securitize and sell those bonds and market. And I think what we've created is an asset class to Dave's point that is an AI infrastructure asset class people want to invest in.

00:30:50 Dave McClure: Asset class. Yeah.

00:30:52 Aman Verjee: Yeah. And that and CapEx has exploded in the last three years, but it's building real assets with real value. I think Intel and other companies have been depreciating their chips over three years, four years, but it's like your smartphone. You know, your old smartphone, it does kinda work fine. It just sort of obsoletes. These GPUs are like cars. They don't just obsolete in two or three or four years. I think they are and NVIDIA sees this. I think they are financeable over eight, nine, ten years, and NVIDIA is now pulling in other third party players to finance it. So the red flag would be is NVIDIA putting in money to their customers, and then their customers pay them their own money. And then NVIDIA but not on commercially reasonable third party terms. NVIDIA is like, hey. If you buy a certain number of, you know, GPUs, we'll forgive 10%, 20%, 30% rebates. Will we end up with bad debt because the asset doesn't work or it's not financeable? I think the demand is real. I think the assets are financeable. The fact that Goldman Sachs and Apollo and Brookfield and KKR. I can't remember all the six of them, but there were some major players who are putting their own money in. And if as long as that's a really good positive signal, I think, that shows that you shouldn't be that worried about the circular financing any more than you are with, like, aircraft leasing or, you know, general—

00:32:04 Andrew Keen: Buying iPhones using—

00:32:06 Aman Verjee: Well, that would be different.

00:32:07 Andrew Keen: You were buying, you using Apple financing. Dave, you're in West Virginia.

00:32:12 Aman Verjee: Exactly.

00:32:13 Andrew Keen: You grew up there. So if things are relatively rosy in the financial garden in terms of a crash, what about the impact on America itself, on places like—

00:32:13 Dave McClure: Yes.

00:32:24 Andrew Keen: Morgantown in West Virginia? Lots of articles like this recent one in The Wall Street Journal about a jobless boom. Are you concerned with this, that it's fine if you're Elon or Sam or Dario or the three of us. But for most of America, no one's really seeing the benefit of this current boom.

00:32:44 Dave McClure: Well, I think they are seeing the benefit in certain ways, but that's probably mostly if they are an equity holder. And I do think there are some concerns for Middle America and certainly maybe the lower third of America. If those jobs that they're performing are not, you know, AI based jobs or jobs that are positively impacted by, in fact, they could be negatively impacted by AI, then we do have some concerns. And, you know, we probably already see some concerns about the cost of electricity in these communities. You know, there might be concerns about, you know, other potentially negative impacts of those spending as well. I think people are generally optimistic that technology historically has created more jobs than it's destroyed, but it certainly provided displacement of jobs. And there might be, you know, retraining that's needed. There might be people that can't, you know, necessarily, you know, take on those new jobs if they are, you know, much more involved with AI. I don't know that I wanna make a prediction there. I would just say that it is reasonable to be concerned about whether people who have previously been doing, you know, service provider jobs, in other areas are gonna be able to keep up. That said, I would say, you know, a lot of the business here in West Virginia is in areas in health care related areas, probably in energy related areas, and I do think we're gonna see both of those growing, in an AI world. So I don't know what the outcome will be, but it is it is prudent to probably be at least cautious and monitoring what's going on.

00:34:19 Andrew Keen: Aman, there was another interesting piece last week in The Wall Street Journal about, America becoming a nation of Luddites. This is the summer of twenty twenty six. Lots of pieces about the rejection of data centers in rural America. What's your take? Is this summer fair for newspapers because not much news is happening? Are you concerned about this shift to tech skepticism both on the left and on the right in American politics and this hostility to AI? Is this gonna, if not, bust the AI bubble, if there is indeed a bubble, at least slow the growth?

00:34:59 Aman Verjee: It's, I would say I'm concerned. I think that it is probably not great that for to Dave's point, I think it's technology these technology revolutions aren't new, and it's something I cover in the book. We saw we saw, I think, more technology happened faster in the nineteen twenties, certainly in the eighteen eighties with the railway boom, probably in the nineteen nineties with the Internet. I think all those were periods of extreme change due to technology that happened more quickly than what AI has done so far. The optimism, though I remember the Internet optimism where it was gonna make life better. We're connecting people. You can go back and read the press in the nineteen twenties about just life getting better with technology and electrification and the assembly line. And, even in the era of the Luddites, there were far more, I think, optimists than pessimists if you just survey people. And the economic growth was happening, and so people could see the benefits in real time. This is one of the few revolutions, maybe not the only one, but one of the few revolutions where the leaders in the technology, AI in this case, are so dour about it. You hear Sam talking about it and how it's gonna threaten jobs and, you know, create idleness. Dario has been on the record. I think in about a year and a half ago, he said that AI was gonna destroy half of knowledge— entry white collar workers within one to five years. His prediction is already wrong because we're a year and a half into his prediction, and we haven't you know, it doesn't seem like we have a 10 to 20% unemployment rate, but that is what he said. And when you start predicting things like that, I think it's natural to get the, political hackles get up, and people get concerned about employment. They're concerned about data centers. We've seen the evidence, though. Data centers bring jobs. They bring skills. They bring, you know, manufacturing talent to these communities. They increase the tax base. They aren't putting they're not they're, self financed, so they're not stressing the tax base. You can tell them that you have to pay for your own power. You can't you have to be a contributor to the grid, not a detractor from the grid. They can, you know, they can play by those rules. And yet there is a brewing political backlash from, I think, both sides, and that is that is kinda concerning. But we the thing to remember, I think, was what Dave McClure just said before is it's not a net destroyer of jobs. It's a it's it will displace jobs. That's happened before. We've seen entire job categories eliminated within ten years. Back in the nineteen seventies, the number one job that women had in the workforce who didn't have a college degree was secretary, typist, stenographer, clerical work. We have between four and six million of that job category in the early seventies. Within a decade, those women were displaced. Those jobs were displaced. They either entirely went away or dramatically changed, and yet the labor force participation for women in the early seventies in The US was 52%. And today, it's, like, 77%. So the jobs not only just didn't get you know, the jobs did just get this did get displaced, but a whole new category of jobs arose that you couldn't even imagine in the seventies, whether it was IT, you know, programmer or computer consultant. My daughter wants to be a YouTube influencer. I don't I don't really even know what to make of that, but that is not—

00:37:54 Andrew Keen: Everyone's daughter wants to be a YouTube influencer.

00:37:57 Aman Verjee: I'm not sure.

00:37:58 Dave McClure: What's—

00:37:58 Andrew Keen: Like that?

00:37:58 Dave McClure: I think there is going to be an incredible amount of growth in, solopreneur or small business entrepreneurship that's backed by AI. And I do think for people who are entrepreneurial minded and this is not necessarily they'd have to be technical. I think this is just people who are can go after interesting businesses. And instead of hiring five to 10 people to do those businesses, maybe they hire a couple people and build a business with AI. I actually think that could be a very big growth area, in the future. Now whether that's gonna provide enough jobs for people who are no longer driving trucks because we've got, you know, AI driven trucks out there, I don't know. I do think that factory automation and robotics automation are going to displace a lot of jobs that maybe people associated historically with manufacturing businesses. You know, I think that's very possible that we're gonna have a much lower, you know, ratio of human capital required for these businesses in the future. But that doesn't necessarily mean that all those people will be jobless. It really just depends on, you know, how creative they are about finding new jobs and whether we can provide some help for them in as they transition.

00:39:07 Andrew Keen: Well, when you say whether we can provide some help for them, we, of course, is not the three of us. It's the government. Keith Teare, I know you know his work. He's—

00:39:17 Dave McClure: Sure.

00:39:18 Andrew Keen: Always very ambivalent about regulation, the government. What's your position, Dave? Should I mean, do you agree with Dario and, and Sam that some of this industry needs to be regulated? Keith seems to think that they're calling for regulation in order to, well, secure their own market dominance. But isn't there a need for more regulation?

00:39:39 Dave McClure: I think you just bridged to two different questions. One is, you know, are we gonna provide some sort of social safety net for people who might be displaced from jobs? The other one is, you know, are we gonna regulate these companies? And I'm not a big fan of overregulation, particularly when I think there's a lot of competition happening outside The US and other countries. We should be aware of whether regulation slows, you know, the overall country down. But just in general, I'm less of a fan of that. But I do think we have to be aware of how society changes and, you know, historically, we would have technological change. It's been over the course of a generation or two, you know, the railroads, the steam engine, electricity itself, you know, even the earlier incarnations of the Internet. I think the thing that's happening now is the changes in AI are so rapid and so dramatic and so fast. You know, do we have enough time to adjust? You know, when those changes take place over a matter of, like, you know, three years instead of thirty, you know, how quickly can society adjust, to that type of change?

00:40:42 Andrew Keen: Aman, you've just published your Brief History of Financial Bubbles. You've looked at everything from the railway boom, the tulip craze, the subprime, boom, the subprime bubble of two thousand and eight. How would you historicize AI in terms of this supposed abundance of intelligence? What's the most comparable technological leap forward?

00:41:09 Aman Verjee: I would probably point to two. Of the 10 bubbles that I that I looked at, there were two positive bubbles, and then there were about six or maybe seven destructive bubbles. And the two positive bubbles turned out to be eighteen forty five in The UK, the railway boom back in, your, your home country in, 1845 to 1849. And the other one was in The US. Primarily, it was the tech boom from '97 to 2000. And in both cases, there was an overbuild. They built stuff too much too fast. I had a demand, and the demand eventually came for the railways and passenger rail or for the Internet. But the investment boom came and went, and it resulted in a decline in the asset prices. But it left the world with a valuable technology, which I think is something maybe closest to where I don't really think we are in an AI bubble for reasons, I think, similar to Dave that we can talk about in terms of evaluations and just the real businesses. But to the extent that there's an overbuilt happening, it'll leave the world with a new, more valuable technology that I think will paid benefits. Just as a comparable, you know, before the Internet, there was a if you had bought all the stocks founded during that Internet boom from about '90, '94 to 2000, you would have owned NVIDIA and Amazon. They were founded in '93, '94. Google in '98. PayPal, where Dave and I worked and met in '98. EBay and Salesforce and Broadcom and, and Amazon. And even if that even if you had held your stocks from that point all the way to 2020, you would have made money. Now you would have a big up— upturn in 2000, 2001, and you would have lost 70% of it in [unclear]. But if you had held your stocks through that whole period, you would have been just fine because the value created the long term, they'd actually justify the amount of money going in. There was just a timing difference. And similarly, you know, you referenced the Luddites who were in, what, 1811 in The UK. The railway boom happened in 1845. There was a long boom in right after that bubble burst in the rail in the railroads from 1850 to 1873 called the long Victorian boom in The UK. And it was a time of rising incomes for everybody, technology driven leadership at the end of the industrial revolution that left the world far better off than it had been. And that's the thing to remember is that I think— your— Dave's point was right on, if you can get people to embrace that positive change, understand that there will be transitions, whether the government provides training, whether the private sector does it, Your schools have to teach AI and teach our kids how to use AI in this new world. As long as that happens, then, you know, the long term, I'm very optimistic about.

00:43:37 Andrew Keen: We did a show recently on 1873, the long depression of eighteen seventy three, which, is seen now, I think, as the first global economic depression. Should we be fearful of that, Aman, and of a repeat of 1873, of the entire global economy falling into doldrums?

00:43:59 Aman Verjee: I don't if it we should always worry, I think, about these downsides. The causes of that particular bust, I think, had a lot more to do with the financial sector. They were The UK was on the gold standard, and so was The US. It was right after the US civil war. If I if I look at the period from 1870 to about 1900, let's take just focus on The US. Right after the civil war where you had all this technology happening, it was a period of tremendous economic growth. The population here in The US essentially doubled. There was a huge inflow of immigration between 1870 and 1890. Agricultural production doubled, and, mining and manufacturing went up, I wanna say, eightfold and sixfold, respectively, although it may have those numbers inverted. And so you look at the state of economy all through that whole period, and it was a period of consolidation. You had big companies like Standard Oil and, you know, you had the Rockefellers and the Carnegies, and you had, large companies getting tremendous amounts of power. But at the same time, what happened to freight rates between 1870 and 1890? They declined. What happened to steel prices from 1870 to 1890? They declined. What happened to the price of oil? 1870 to 1890? Went down by 60%. And then you had all the trust busters and all that stuff in the in the early nineteen hundreds. But, even the result of that power, that consolidation, that entrepreneurial capitalism, that concentration of wealth was unquestionably positive on average for median people, for people in the workforce, working class, for women, for minorities, for just about everybody in those countries. And so there may be periods of dislocation, two, three, four year recessions, ups and downs, but the most important thing is what happens over the long sweep of history. And the technology has a very important role in improving that growth because it comes down to productivity and how capital and labor work. And every because—

00:45:39 Andrew Keen: Keynes is famous. He said in the long run, we're all dead, and I'm not sure how long run we're talking—

00:45:45 Aman Verjee: About that. He was he was right. Well, it may be it may be, you know, you have to think we Dave and I run a venture fund that just has a seven year term, and, a lot of VCs think in ten to fifteen year increments. So, I don't think it's gonna I don't think it works. You know, capitalism doesn't work quarter to quarter. It doesn't go up in a straight line up into the right.

00:46:04 Andrew Keen: We're trying to be very reassured. You guys are very calm. I guess you have to be as investors. But, Dave, there's lots of I mean, just today, Michael Burry, a particular pessimist, walks— warns of the market bubble. Yes. Times run a piece on—

00:46:20 Aman Verjee: He's called nine of the last two bubbles.

00:46:22 Andrew Keen: Right. So a clock is always right twice a day, a broken clock. You know, New York Times seems to run a piece every day about AI hype running into reality. And then there are real manifestations of certainly exuberance as the Leopold Aschenbrenner story, a seeming young genius who lost a lot of other people's money. Is there an argument that this is, a boom? I mean, what do you make, Dave, of the Aschenbrenner story? I'm sure you saw it. He got a run-in on that one.

00:46:55 Dave McClure: Yeah. I think that was more problem of leverage, not of, the boom per se. He probably got caught between a lot of volatility in the market, and I think he was running at four x or more leverage. And at one point, you know, had to find a backer, sell most of his public stocks. However, he did, as far as I understand, hang on to the Anthropic position that he had. So it wasn't a problem of the AI stocks that he was holding. I think it was just the volatility in chip stocks that caught him with his pants down for some period of time.

00:47:23 Andrew Keen: And—

00:47:24 Dave McClure: Yeah. I have a feeling he'll probably come back, with new backers and, you know, the guy's only 25 years old, I think. And I believe his, his fiancee I don't know if he's got married, but his fiancee—

00:47:35 Aman Verjee: Married now.

00:47:35 Dave McClure: Is, Dario Amodei's chief of staff. So Yeah. He's probably got pretty good insights as to what's happening in the AI world. I have a I have a feeling that guy will be just fine. Again, I think it's hard to say, like, what is going to happen. There's certainly gonna be a ton of change. There's definitely possibilities that change will impact people negatively. Frankly, I'm probably more concerned about the overall US economy from just a deficit perspective and budget perspective more than AI bubble. Like, I think we should be concerned that we're running, you know, pretty tremendous deficits and outspending our national budget on a regular basis. I think now the largest single line item in our national budget is the interest on our deficit. So that's probably a way more of a concern. Unfortunately, I don't think either party has a good track record in reducing that deficit. And we are seeing pretty increasing crazy cost in health care. But I am pretty optimistic that the productivity gains from AI might be able to help that situation. I think we're gonna see tremendous, you know, even though we're going through quite a bit of challenges with the situation in Iran and Ukraine. I think we're gonna see tremendous advances in health care, in energy, and physical automation over the next five years. That really will benefit society very broadly. So I'm again optimistic in the long run about the changes that will, you know, benefit humanity, but the upheaval in the short term could affect certain groups negatively. And those are real concerns that we should be watching and paying attention to.

00:49:13 Andrew Keen: Yeah. I think that goes without saying. Let's have a final word with you, Aman. Your Brief History of Financial Bubbles is that you put a lot of work into it. I wonder why the this what this history tells us about competition between nations. Some people don't think that the fight between the Chinese and the American economy is a zero sum game when it comes to AI. Others believe it is. Can America and China both win in the in this new world? And, is there equivalent frothiness in China as there is in The US?

00:49:54 Aman Verjee: Oh, those are those are deep questions to end on. I think there's a lot of benefit to the countries working together. There is a conversation, I guess, in the Valley more broadly about, you know, cheap Chinese open weight models being, imported and used by the Americans. And then somehow that undercuts American leadership and should be should we be concerned about that? Are we rivals or, you know, are we, something worse than that, like enemies? I will just say that my daughters both are Mandarin fluent. They both speak Chinese. I'm a, by the way, as big a proponent of America and American democracy and, all the principles that undergird this country as anyone. But I would hope that a part of that means that you can educate your kids to learn about different cultures and that you can work with them. If we're if my kids now go to and they go to Taiwan every summer for, you know, for a summer and school thing in Taiwan, and they come back full of stories about what they love about Taiwan. And it's the night markets and the culture and, you know, some of the, the, stories that they tell and the holidays and appreciation for all that stuff. If we grow if we grow like that to be trade partners to work together, we've benefited tremendously from importing stuff from China and abroad. I'm a I'm a free trader and always have been. We should specialize in what we do and export what we do and what we know that makes China better, that makes us better. I think that's gonna that's the secret to creating wealth for everybody. How you divide the pie and, you know, who gets trained for that future, those are all fair questions and how we prepare individuals or different groups for that, you know, for that long arc of history is a is a fair point. There may be areas where we disagree with China on, on politics and certainly are. There may be territorial disputes over the Taiwan Strait. But, you know, my hope is that for the most part, we can be partners in progress on all of that stuff and work together in what really matters, which is business and culture. You know, they love a lot of our movies, in Taiwan. If you go to Europe, people are always talking about Taylor Swift and buying American brands and then lining up to go to Starbucks. And, and then you can disagree with Donald Trump. And you can be angry about something that, you know, one of our politicians said on, on Twitter, and that's all that's all part of it. But, well, I think there's a lot of benefit to being able to work with China and their industrial capacity and their the energy population and their innovation. We have so many of our best scientists in China come to educate— get educated in The US and go back. You can see that happening. There's a lot more to gain than lose if we do this right.

00:52:17 Andrew Keen: Very cheerful, refreshingly cheerful, Aman Verjee. His new book is out on, A Brief History of Financial Bubbles, put a lot of work into it, Dave McClure talking to us. Let me end with you, Dave. You're cautiously optimistic. As I said, it's refreshing. There's so much pessimism around these days in politics and economics. What's, what's your secret sauce? What keeps I know you can't speak on behalf of Aman, but what keeps you so cheerful?

00:52:47 Dave McClure: Yeah. By the way, I wanna make sure that, people wanna check out Aman's book. That URL is bigbubbletrouble.com. So if you wanna find trouble.

00:52:56 Andrew Keen: You're his agent too. Do you have time to be his agent?

00:52:58 Dave McClure: You bet. You know, I guess I would say as a as a hillbilly who grew up in West Virginia and found my way to California and entrepreneurship, I'm, you know, an example of the American dream and, you know, somebody that came from not too much. You know? My family was, you know, immigrants from Italy and Ireland and England, you know, three, four generations back. I had family who worked in coal mines, you know, here, probably two or three generations back. You know, it's continued to be, you know, a great opportunity for a lot of people. I again, I might disagree with a lot on some of the directions of our current administration of politics, but, you know, America as a country has always created entrepreneurial opportunities, particularly for immigrants. And I just hope that continues to be the case. In fact, most of these AI companies that we're talking about, a lot of the people working there are immigrants and people who come from other countries. And probably half of the, you know, unicorns in Silicon Valley, if not more than half, have foreign, you know, founders involved with them. So I continue to be optimistic about, you know, the progress we've made. Yes. There's problems and challenges for us to overcome. But, you know, I generally feel like we're headed in a positive direction at least, you know, going forward.

00:54:17 Andrew Keen: Yeah. And, Aman, last time we talked, he talked about the importance of immigration to keep America prosperous and optimistic. Real pleasure to have such cheerful guys on the show in such supposedly dark times. Aman's, new book is out. Aman, as you said, you can buy it online, A Brief History of Financial Bubbles. Very important, interesting new book, and great to have you guys. We'll have you back on by the end of the year when—

00:54:42 Aman Verjee: Great.

00:54:43 Andrew Keen: As you predicted well, I'm gonna call you guys if, if OpenAI doesn't go public this year. You're the—

00:54:49 Dave McClure: All three companies will be public, and the collective market cap will be north of $6 trillion, maybe 7. Love that.