Let Them Eat Intelligence: A Silicon Valley Eulogy for the Working Class

00:00:31 Andrew Keen: Hello, everybody. It's that time of the week again. That Was the Week, our weekly technology summary with my old friend Keith Teare, publisher of That Was the Week newsletter. Keith has some big news this week, at least according to his headline, humans create intelligence. And now he says everybody can access it with, you know what, artificial intelligence, Keith argues, and seems kind of self evident, but it's an interesting idea. He says that, he believes that humans need access to intelligence because human knowledge should not be, and I'm using I'm quoting him exactly, should not be trapped inside experts, institutions, or companies. In other words, what we're talking about now when it comes to AI is a radical democratization of intelligence. Everybody has it, and it's a theme that, is in a lot of the articles he connects with this week. Noah Smith, for example, also, like Keith, believes, I think, in, the democratization of intelligence. He talks about the end of the age of heroes. So, Keith, we're living through historic times. Is that fair? Am I mischaracteri— you— you sometimes accuse me of mischaracterizing what you've written or said. But are we talking about the radical democratization of intelligence?

00:02:03 Keith Teare: I think you gotta distinguish between, individual contribution to intelligence and collective intelligence. We're not gonna ever democratize individual contribution. That is where experts live and belong. And, you know, some individuals contribute more to collective intelligence than others. That's all—

00:02:24 Andrew Keen: Like you and I.

00:02:25 Keith Teare: Exactly. So that's always gonna be true. But the sum total of all intelligence made up of everybody's contribution and that and everybody just doesn't mean living humans, but all humans ever. That is really the meaning of the word intelligence when thought of as a collective, a word for the collective sum of all progress today. And it's that needs to be distributed. And up until now, that would have been impossible. You know, it would have been encyclopedias, libraries, books, maybe search engines. What we have for the first time with AI is the ability to aggregate, distill, digest, and then leverage intelligence. And it's now cheap enough that billions of people could get access to it. Ultimately, it should be cheap enough that everyone gets access to it. That isn't my wish. It's almost like the inevitable end game of what AI is. It uplifts every individual to the level of what previously only a wealthy individual would have been able to have access.

00:03:37 Andrew Keen: But it's not just wealthy. I mean, this idea and I'm gonna quote it again. Human intelligence should not be trapped inside experts. I mean, an expert, of course, historically, at least in the modern age— to become an expert, you had to go to university. You had to work hard. There were more or less successful experts, more or less authoritative experts. You said earlier that there's still gonna be experts. What kind of experts? If, in your view, human knowledge shouldn't be trapped inside them.

00:04:14 Keith Teare: I think the same way there's always been experts. I mean, look. If two different people look at the output of, an AI conversation about, let's say, viruses, an expert in viruses is gonna have a very different understanding than somebody who isn't an expert in viruses. So experts don't go away. In fact, experts, if anything, get elevated.

00:04:39 Andrew Keen: I mean, you can't have it both ways. Noah Smith doesn't try to have it both ways. In his end of the age of heroes, he talks about essentially or he predicts the end of mathematical heroes. I'm not sure he's particularly unhappy about that. He said few of us really did get to be heroes in the age of science, and now nobody's gonna be. And he said it's not and I'm quoting his end. He said it's not so bad, really. Never— most— and I'm quoting him. Most people never get the chance to be heroes. Truck drivers don't. Financial advisers don't. Executive assistants don't. They have to find meaning just from being regular people, from taking care of their kids, having friends, having hobbies, joining civic organizations, getting involved in politics, and so on. It's not such a terrible fate. So is this not such a terrible fate, Keith, the fate of all of us in this age of radical democratization of intelligence?

00:05:43 Keith Teare: Yeah. I kinda disagree with, Noah Smith in one detail. If we're talking about the sum of human intelligence, I agree with him. We can all benefit from that. And, the lowest of the low and the highest of the high equally can benefit from that. Where I disagree with him is there will still be heroes because individual intelligence doesn't go away. It get— actually gets amplified. And so think about math, his area of focus, math. I don't— do not believe AI will mostly solve new math problems. There's no evidence yet that AI can, create new knowledge, new intelligence. What it can do is understand and distribute existing intelligence, and it will still take individuals to go beyond. And there will be individuals that do. And in fact, they'll do it more because the AI will level them up to, the understanding of what is already understood and known. So their starting point for innovating will be much higher. And so there will still be heroes. I think he's wrong about that.

00:07:03 Andrew Keen: Well, in a sense then, we're starting all over again, a fresh slate. And this was another theme that you pick up on, in your newsletter this week. You connect with an interview that Yascha Mounk made of, the great MIT economist, Daron Acemoglu, who has a new book out. The book is called, What Happened to Liberal Democracy, and it's really a book about the opportunity in AI. Acemoglu also had a good interview— a good article in The Atlantic summarizing his book. Like you, he recognizes this radical democratization of intelligence with AI, and he talks about I'm not sure if you read The Atlantic piece, but he talks about what he wants to see as a radical pro worker AI. Do you share what Acemoglu is arguing if you believe that we still live in an age of experts?

00:08:05 Keith Teare: No. I completely disagree with it. I put it in because I disagree with it. I think the, let's call it fetishization of workers. It's almost like the middle class worship of the working class as some kind of romantic iconic group of people that should be preserved is, both paternalistic, but also just historically inaccurate. The history of the working class has been that it is it grew in order that it could shrink. It grew through the industrial revolution, and he's quite right about that. And wages became stabilized and standardized to some extent through things like minimum wage. But the idea that the wage represents some kind of progress is only true if you look backwards at feudalism or early capitalism. The wage is, symptomatic of a power relationship between, the employers and the employees. And what technology is doing, he calls it severing that relationship. It's completely removing that relationship. The whole idea of labor and capital is now open to question as whether that's the future organization of society. And that's a good thing. And, unfortunately, a lot of the left, you know, you they make permanent what is temporary. And the wage capital relationship is temporary.

00:09:50 Andrew Keen: So what happened to Acemoglu? Let's be concrete here. He says, Silicon Valley's attention needs to be shifted towards pro worker AI, and I'm quoting him here from his Atlantic piece. He said, starting at the international level, The US and Europe should create AI agencies that can generate incentives such as grant programs or public competitions for technologists to develop models that are designed to expand workers' capabilities. You would disagree with that?

00:10:20 Keith Teare: Completely disagree with it. The what's happening now is the end of the working class. Not the, preservation of the working class or the looking after the working class, which is why I call it paternalistic. The working class is going to be replaced because there won't be one. There will be robotics. There will be AI. There will not be traditional wage labor jobs. And that's a good thing looked at from a progressive point of view when you want to take, you know, my aspiration as a working class kid growing up in a council estate in England was to not be working class. I didn't want people seeking to make me permanently working class. I want— I didn't want to be working class. And now for the first time in history, it's happening.

00:11:11 Andrew Keen: How many of the kids you grew up with in Yorkshire ended up in Palo Alto?

00:11:20 Keith Teare: Why is that relevant?

00:11:21 Andrew Keen: Well, because it's a bit of a lottery. I mean, the if many of them did well, I mean, you've always said to me that most of the people you grew up with, what are they all doing now?

00:11:33 Keith Teare: Well, they're doing better than they were doing, when I was a kid. I can tell you that because life in general is better than it used to be. But, that's kind of a micro point. The macro point, which is I think, more fundamental to his argument is that, as automation kicks in, human labor diminishes. That's inevitable.

00:12:00 Andrew Keen: And— so what happens— okay. So you don't agree with Acemoglu. So what about Smith's point about heroism has to change? Do you agree? I mean, how are we supposed to value people in this new age where intelligence is, at least according to guys that are optimists like you, is, is abundant?

00:12:21 Keith Teare: Well, that—

00:12:21 Andrew Keen: There's no scarcity of intelligence.

00:12:25 Keith Teare: Actually, you're now reinforcing why I chose the title this week, humans create intelligence. There is no such thing as intelligence separate from humans. And heroes, defined by Noah, are those humans who contribute more to the collective intelligence than others. They don't go away. They absolutely don't go away. AI doesn't replace innovation. It's just a tool for innovation. And so this whole idea of a counter position between and the word AI and the word intelligence are not the same word. Intelligence is 100% human, and AI is a tool created by humans to distill, digest, and distribute intelligence. That's a very, very important distinction and what both Noah and, I can't pronounce his name.

00:13:25 Andrew Keen: Acemoglu.

00:13:25 Keith Teare: Acemoglu are doing is they're papering over that distinction between AI and intelligence. Intelligence is only human, and therefore, the idea that humans will cease to contribute to it is just not true.

00:13:41 Andrew Keen: Yeah. And the title of the interview with Mounk on— on Persuasion is why liberalism has become a victim of its own success. And I wonder I mean, this is a bigger point. I wonder whether liberal ideology, the dominant ideology of the last two or three hundred years, which is built around the self and the ability of individuals to create and recreate themselves. I wonder how this is gonna be impacted by AI, whether we become more collective. It's— it seems to be the central question of the long— the central long term question of the AI age.

00:14:22 Keith Teare: I don't— I actually think the opposite. I think we become more individual. I mean, I can just tell you my own experience of this. So my hard drive ran out of space a week ago, and, and it's a Mac. Right? So normally, you can't upgrade a Mac's hard drive, you'd have to go and buy a new one.

00:14:47 Andrew Keen: Or you clean it out.

00:14:49 Keith Teare: So, what I decided is to take the step of upgrading the hard drive in the Mac, which is doable. It's just difficult. And to do that, to wipe out all of the stuff on the hard drive. And so I said to my, to Claude, please back up, the AI part of the hard drive in such a way that we can put it back afterwards. Almost like a brain transplant from robot A into robot B. And, and over the course of forty eight hours, it completely backed up. Then I upgraded the hard drive, then it restored the Claw— the OpenAI— the OpenClaw brain, if you will, back to the new hard drive.

00:15:38 Andrew Keen: So okay. I can— but we'll still— what's your point? Your point is that everyone's gonna be able to do that?

00:15:42 Keith Teare: Is that— as an individual, I was super individual. I was able to do something that I couldn't have done on my own. And so the idea that the self is undermined, no, the self is completely expanded.

00:15:58 Andrew Keen: I think the fear would be and, you know, we seem to repeat this conversation every week. The fear of people like Acemoglu is that the reality on the ground, headlines today of fewer jobs being created. For all the theory, in practice, inequality is being compounded, and the AI age will only it might create a few Keith Teares who get out of their council estates and become wealthy Silicon Valley entrepreneurs. But most people, for one reason or other, won't take advantage of that. It's only gonna compound inequality. So if you reject— hold on. If you reject—

00:16:34 Keith Teare: Just to justify that, Andrew.

00:16:35 Andrew Keen: Well, but if you re— well, if—

00:16:37 Keith Teare: I think that's just blatantly not true.

00:16:40 Andrew Keen: Well, maybe you should get out— maybe get out of Palo Alto. I'm not sure that's necessarily typical of the rest of America. But if you reject Acemoglu's point that there's no need for pro worker AI. Are you suggesting the government should just stand back and just let this revolution happen?

00:17:00 Keith Teare: What even is pro worker AI? This is complete bullshit.

00:17:03 Andrew Keen: But are you suggesting then that no one should do any—

00:17:06 Keith Teare: I don't want you to echo his words. Tell me what that means.

00:17:10 Andrew Keen: What did you— I mean, you need to read—

00:17:13 Keith Teare: Bullshit. There's no such thing.

00:17:16 Andrew Keen: Well, I'm quoting him here. He says, and I'm— pro worker AI could boost productivity across other service industries.

00:17:25 Keith Teare: Pro worker AI.

00:17:27 Andrew Keen: In education, a model trained on— hey. So this is what it is. In education, a model trained on relevant curriculum material might allow teachers to spot patterns in test results and tailor lesson plans for particular groups of students who tend to make the same mistakes. So he's not necessarily completely disagreeing with you. He's simply saying that AI can help teachers, for example, become better teachers. But I think what he wants then, and I'm quoting him here again, a pro worker AI could create good service jobs for workers who aren't college educated and thus begin to uproot the inequality that has been entrenched in the post industrial economy. Do you disagree that— with that, that there is— that inequality has been entrenched in the post industrial economy? I mean, I think that's just a reality. I mean, that's a quantifiable reality.

00:18:22 Keith Teare: I think what it's doing is anthropom— anthropom— how do you say that word? Anthropomo— turning AI into a— a— a light [sic] living being. Anthropomorph— I can't say it. Can you say it? No. You know what I mean? The idea that there's there could be a pro worker AI or an anti worker AI. The very framing of AI as being, you know, we can choose. It's a mistake. I actually agree with his point, And he's really making my point that AI, creates better individuals, better experts. He's more agree with me than Noah Smith who said heroes go away. I think probably every single teacher becomes a hero in the world of AI.

00:19:15 Andrew Keen: I'm gonna try and get him on, I'm talking to some of his publicist. The book, What Happened to Liberal Democracy, is out next week. I mean, he's a Nobel Prize winning economist. So, hopefully, we can get him on. So maybe we can continue this conversation. But coming back to your— so if you don't agree with the— Acemoglu— that the state government's international institution should be creating what he calls pro worker AI. What do you believe— that just that this is just gonna somehow magically work itself out, that everyone's gonna be empowered with this new intelligence and we'll have a Hollywood ending to all this?

00:19:56 Keith Teare: Well, I think as a starting point, you have to start not with the word worker, but the with the word people. I think AI is for people. The class struggle doesn't really play a role.

00:20:10 Andrew Keen: Okay. But that's a sema— it doesn't matter what you call work or people.

00:20:13 Keith Teare: It really isn't semantic, Andrew. It's so, you know, I watch, kids in school using AI in class in a way that previously they couldn't. And by the way, that's even in East Palo Alto.

00:20:29 Andrew Keen: Have you been to East Palo Alto?

00:20:32 Keith Teare: Many, many times. In fact, East Palo Alto—

00:20:34 Andrew Keen: We won't repeat our East Palo Alto stuff.

00:20:38 Keith Teare: East Palo Alto comes to me. Right? My middle son is—

00:20:42 Andrew Keen: Yeah. Most expensive. Your kids drive your fancy cars in East Palo Alto. That's all I can say.

00:20:48 Keith Teare: Not such fancy car. But,—

00:20:51 Andrew Keen: Well, fancy in East Palo Alto.

00:20:53 Keith Teare: You know, you're making the mistake of branding, an individual that's done well as somehow not capable of understanding poor people. And I don't agree with that.

00:21:05 Andrew Keen: I didn't say that. I think you are.

00:21:07 Keith Teare: You're implying that—

00:21:07 Andrew Keen: we're just not answering. You're not answering my question if you reject Acemoglu's argument about creating, quote, unquote, pro worker AI. You don't like his language. Fair enough. And you don't like his policies. Fair enough. What are you suggesting? What needs to happen? We just stand back and let it all happen. Let this upheaval happen and it'll work out well.

00:21:33 Keith Teare: Well, well, use the word it and upheaval. I don't think AI is an, upheaval. I do think it's change. And, there is a distinction between upheaval and change, but they can often appear the same. I think what's happening is that the entire human race is gonna have a starting point in life from the first moment you start to ask questions. You're gonna be able to get answers that would have been impossible, for most people most of the time. And, you know, that is, true. That's gonna be true when I go to South Africa and I see, you know, young Africans with Android phones having Claude on their phone. That is a revelation. It's huge and positive. And so it's an upheaval in one way, but it's a very positive—

00:22:37 Andrew Keen: So you're not answering my question. We had the same—

00:22:39 Keith Teare: No, no, no. [ed.: the raw ASR repeats "No." fifty-one times across this two-second overlap, split across all three speaker labels; collapsed here]

00:22:42 Andrew Keen: No, they did. And it's still free, but I'm not sure it necessarily compounded opportunity. But sorry. Go on.

00:22:59 Keith Teare: Well, I— I think I am answering your question, to be honest. It's just you don't like the answer.

00:23:03 Andrew Keen: But what is the answer?

00:23:05 Keith Teare: The answer is that, yes. We should allow it to happen. And it and as far as the word upheaval is appropriate, it's a good upheaval.

00:23:16 Andrew Keen: Okay. Well, you I know you're not keen, and I'm in some ways, sympathetic with you on this about this reaction against data centers. One of the other pieces in this week's newsletter is a piece about the number of, quote, unquote, regular I don't know what a regular person is, as opposed maybe to an irregular person or a hero. But the number of regular people being arrested for protesting data centers is astounding. I don't know what exactly those numbers are, but I know you feel that people are probably not wise to protest data centers. What's your take on this and this growing movement for better or worse against data centers?

00:23:57 Keith Teare: Look. I think it's a bit of a bandwagon. Erin Brockovich was on Bill Maher's Real Time last week talking about data centers, And, she is trying to bring class action lawsuits, and is kind of portraying data centers as being, morally worthy of being a target in the same way that drug companies were in her previous, previous—

00:24:22 Andrew Keen: She was on my show. She's— a few years ago. She's actually very good. I mean, very entertaining. I'm not sure if she's right on this particular—

00:24:29 Keith Teare: Yeah. So there is a bandwagon. I think it's largely a poli— politicians-led and stirred up by politicians in local areas. And I noticed that Ezra Klein this week has started to include anti big tech language into his narrative around abundance, and is also jumping on that bandwagon about data centers. So, unfortunately, this is the left thinking they have a cause célèbre against the Republicans. And I, I think it's incredibly regressive, but it's quite real. And it, of course, keys into some genuine concerns in local areas about, you know, similar to say windmill power generators kind of uglifying the neighborhood.

00:25:26 Andrew Keen: So, you know, I take your point, and I think, demonstrably, you may be right in some ways. But isn't that— doesn't— if we collectively are able to reshape the conversation, say, on data centers, Don't we need something like Acemoglu's, arguments about pro I know you don't like the term, but pro worker AI or pro people AI. Yeah. Don't we need that more visible so that people can understand why data centers aren't actually against their interests? Because at the moment, most people are just mystified. It's these huge amounts of money, billions of dollars being poured into these data centers. They some of them are benefiting, but most of them don't really see the long term upside for the from their point of view.

00:26:18 Keith Teare: I think that's where Acemoglu and I would agree that the case for AI as something that everybody will benefit from is not being well made. And it's astonishing actually. It you know, when you consider that Sam Altman and Dario Amodei and Demis Hassabis, who's now kind of exited the operational role at Google this week, and not the people making that case. That's kind of astonishing to me. The case is quite simple. And if you could substitute the word workers for the word people, I'd almost entirely agree with Acemoglu. And, and, and, and that's what this week's editorial, by the way, is—

00:27:09 Andrew Keen: But then as it becomes semantic, I mean, worker and people, I mean, it's the same

00:27:13 Keith Teare: No. Really because he focuses in detail on the wage labor capital relationship and wants to preserve it. He really does do that, and that's a mistake.

00:27:25 Andrew Keen: Well, I tell people who rely on their income to pay their rent or feed their kids or clothe themselves. I mean, it's a real issue for people, isn't it?

00:27:36 Keith Teare: Well, I mean, now you're conflating a short term need with long term trends, and he's an intellectual. He's not really talking about short term. He's wanting to make permanent the current social contract.

00:27:52 Andrew Keen: So you're saying— you mentioned, Demis Hassabis, who did run DeepMind up until earlier this week. There's been a shift away from— big shift within Google. Hassabis has been replaced. And, the Financial Times actually ran an interesting piece, yesterday about the and I'm quoting the headline, Google's AI shakeup boosts Brin as DeepMind's Hassabis steps aside. There's a lot of frustration, at least according to the FT, within Google, that Hassabis is not focused enough on sales. And I guess, Sergey Brin, the one of the cofounders of Google has replaced him as the AI czar within the company. Is this because no one's— are you saying that they're not making the Omen [unclear]? I mean, Hassabis won the Nobel Prize for his work in AI and medicine [ed.: the 2024 Nobel was in Chemistry, for AlphaFold's protein-structure prediction], in— at DeepMind. He's highly respected. Why isn't he making the argument? What's he doing wrong, or what's he not saying?

00:28:54 Keith Teare: He's not really a villain in this. He's mainly, I think, a good guy from my point of view, but he's very specific. I mean, I mean, to give him credit, he is a scientist who gets energy from being a scientist. And that is quite different from being, you know, the boss of a large organization with people reporting to you, where operational outcomes are your focus. And he's decided to step back into science and that's not a terrible thing. And his main narrative is the, the power that AI can bring to scientific development, like AlphaFold. Although it is, you know, concerning that the entire AlphaFold team at Google has been broken up and no longer exists as a team. So something strange is going on there. But that yeah. I think they're too busy in their businesses to be to articulate a vision and that, you know, they're too tactical. And week by week, they have to make statements that are highly tactical. So if you're listening, where do you hear the pro AI sentiment? I don't think you do. And I think, Acemoglu is trying to create that by making the case that it should be pro worker. And I don't disagree that it should be pro worker, but it should be pro everybody.

00:30:20 Andrew Keen: Yeah. It's still not convincing me about the semantic. I mean, whether you use pro everybody or pro worker. It's, to me, that's a semantic issue. But there's still the reality that a lot of this AI boom isn't for real. You, you link with an interesting piece by Ed Zitron, another of your favorite authors, who suggests that, according to analysts at least, 70% of Amazon, Microsoft, and Google's AI revenues come from OpenAI and Anthropic. And he goes on to say, there really isn't an AI industry without OpenAI and Anthropic. And he argues that this is a huge problem even if you don't wanna think about it. Is Zitron right, Keith? Is this still, in some ways, an artificial boom where cloud sales of Microsoft, Google, and Amazon are being built on Anthropic and OpenAI who are raising more and more money, and they're still private companies. I actually did an interview this week with someone who argues that one of the problems with Wall Street is, that you shouldn't be able to have so many unicorns. A— a law professor at BC has written a book about taming the unicorns [ed.: Renée Jones of Boston College Law School; the book is Untamed Unicorns: Why Startup Finance Is Broken and How to Fix It]. Maybe we'll talk about that next week. Do you disagree with Zitron? Is this still a real economy?

00:31:42 Keith Teare: I actually agree with about two thirds of what he said. And if you take word length as the measure, it is definitely true that a lot of the revenue from the what they're called, the hyperscalers is coming from Anthropic and OpenAI. It's definitely true that there wouldn't be an AI industry without Anthropic and OpenAI. It isn't a problem. That's the bit I disagree with him. It's it if you ask the question, which he doesn't, where is the money coming from for OpenAI and Anthropic to pay Microsoft and Amazon and Google? The answer is from about 2 billion people in the world, some subset of which pay subscription fees to those two companies and are seeing dramatic growths in revenue. And so, yes, it is certainly highly concentrated. But from the point of view of usage, it's highly distributed. And, the money is coming from a highly distributed number of people. The venture capital money and the debt, certainly exists as well. But that would be unsustainable were— were it not for the revenue, and the revenues are entirely sustainable.

00:33:12 Andrew Keen: Your post of the week is by Porter Stansberry about Leopold Aschenbrenner's huge losses. We mentioned this last week, and I think the news broke of this whiz kid who graduated, I think, from Columbia at 19, had a huge AI fund and lost a lot of money, sold it to a Wall Street shark. What is the reality of this market? Are you not concerned about these kind of losses? To some people, Keith, you know this better than I do. It suggests a very, very fragile economy.

00:33:53 Keith Teare: Well, I— Aschenbrenner is very specific, so you really do have to not generalize. He lost money on publicly listed stocks, specifically two Korean stocks, Samsung and, is it SK Hynix, I think it's—

00:34:10 Andrew Keen: Yeah. I mean, the point of the— and it's a good, it's a good post, although it's more like an essay, that suggests that he just got it wrong. He, Aschenbrenner bet on chip companies rather than the next layer up, and he actually got it the wrong way around, that he bet against, Adobe and those types of companies and bet on the chip companies. And the reality of the market is the reverse.

00:34:38 Keith Teare: Well, yeah. He got it wrong, but, you know, getting it wrong in public markets is mostly about timing. And he got the timing wrong. He probably didn't get the thesis that wrong, to be honest. Now if you, then drill down to what happened, Citadel bought his public positions probably at a discount.

00:35:00 Andrew Keen: Ken Griffin. I mean, not probably for sure. I mean, he wouldn't have done it out of the way.

00:35:03 Keith Teare: He wouldn't have done it. However, Aschenbrenner kept his Anthropic shares, which are his most valuable asset. And so the guy is still a multi billionaire. And I think it's fair to say, nobody really lost money because when Citadel bought those assets, they went back up in price. And, and, so, you know, the pool of money pretty much stayed the same. It was a drama in a teacup.

00:35:42 Andrew Keen: So from what you're saying, it's a real economy. People want AI. So even though Google and Amazon and Microsoft AI revenue is being built on having Anthropic and OpenAI as clients, it's not that important. So the real test, the next big test, Keith, of this economy are the IPOs of Anthropic and OpenAI. SpaceX's stock fell this week. I think there's some questions about that. Is that fair, do you think? I mean, to get a real— I mean, everyone has different positions on this, and it's always easy to have a position before it happens. But is that gonna be the reality check in this economy, the IPOs of Anthropic and OpenAI?

00:36:30 Keith Teare: Well, I think though and I having done a couple of IPOs in my career, I'll tell you, the reason you do an IPO is because you draw the conclusion that raising new capital is easier in the public market than the private. So to be honest, neither of those companies will go public unless they've reached the point that they think that's where the future capital will come from. And that may have already happened. It's possible, in which case they'll list. Well, then you move to the share price and the share price is very different. It's a subjective price derived from the supply and demand for the shares. And, SpaceX, you know, listed at $135 and is now about $108, I think, last time I looked. So it's not that far off. Certainly no catastrophe there. And they did unlock, I think about 20% of the locked up shares were unlocked yesterday without any dramatic impact. In fact, the share price went up yesterday. So, so the public markets really aren't a place where you determine success or failure of the underlying business. I think that's more to do with investment, return on investment, revenue, rates of growth. Those are the key numbers. And in those numbers, both Anthropic and OpenAI are doing extremely well.

00:37:59 Andrew Keen: Well, we shall see. Maybe next week, we can do the taming the unicorns conversation this week. My, interview of the week is with Claire Stapleton, former Google person, who's written a book called, Don't Be Evil, Bad Bosses, Fake Promises, and My Escape from Big Tech. We never escape from Big Tech, do we, Keith? We're on it. We're in the Big Tech prison every week. We will be back next week. This week, humans create intelligence. Next week, they'll be creating more intelligence. Have a good week, Keith, and, I will see you next week.

00:38:38 Keith Teare: Bye, everyone.