It's the most important economic leveler of our lifetime.
People are like, 'Oh, look at these guys.' They were like exotic animals at the zoo. Knowledge is different than expertise and intelligence.
Has America created a new class of oligarchs that are based kind of right around where we are right now?
Chamath, thanks for joining The Axios Show.
It's great to see you, Dan.
So you've been in Silicon Valley for about 25 years, and now it's kind of become AI Valley. What's the feeling, what's your feeling here now compared to say when you came?
Gosh, I think I mentioned this last week, but I think it's a bit of the tale of two cities here. We've always had this cycle of boom and bust. You know, when I first came, it was right before the first .com bust. And so I, like a lot of people at the time, I was a 20-year-old graduate from engineering, and I wanted to be, you know, on Broadway, so to speak. And within months of me getting here, it was the top of the market. And we kind of went through the .com bust and the rebuild. Went through it again in the mobile wave. We had a little burst of it in crypto, but that really didn't amount to much really. And now this is the third big cycle. So in one perspective, sort of what's old is new again, you know, we're going back through this cycle. The thing that's different this time around is in these other two cycles, we were very much these cute, nerdy kids over there in California. And for the most part, we were both left alone, but generally trusted, and generally pretty reasonably well-liked. And when we would always venture outta Silicon Valley, people are like, 'Oh, look at these guys.' They were like exotic animals at the zoo. And unfortunately what's happened I think in this cycle is that there's a good amount of admiration, but there's equal parts, I think what people are frustrated, they don't trust this community anymore. Certain people are viewed as pariahs. And I think that is hugely problematic considering what this technology is capable of, which far exceeds what anything we've done in the past. So this is why, I mean, by we're sort of, it's a tale of two cities.
You do find this technology, generally speaking, AI, to be more important, consequential maybe within the past cycles?
Yeah, I really don't, I wanna be non-hyperbolic about this, but I think it's true what I'm about to say, which is, it's the most important economic leveler of our lifetime. It is the thing that will create the most amount of equality if it's allowed to get to market. It is the thing that will allow anybody to follow their dreams and to build something, to build a life for themselves, to build a life for their family outside of the conventions of what society says has to happen to make a living. Why is that? We are at a point where, you know, kind of this old adage where like you stand on the shoulders of giants, right, or you stand on the shoulders of all these people that have come before you. Think of what has had to happen for AI to happen, and what is AI? Well, you build the internet, then you build the first generation of companies, right? They make computers, they make mostly janky software. It's kind of okay, we kind of figure out how to string them all together. We, you know, Andreessen builds a browser. We now start to graphically navigate around the world. There's a Cambrian explosion of information. So far, all amazing things, but frankly precursors to this moment. And you get to this AI thing, and it says, 'Hey Dan, guess what? I have a pretty cool thing, because of all of this stuff, you can view all of it as just training data.' And you're like, 'Well, what does that mean?' I can create this thing, this magical thing, that will give you entirely all the expertise of all the great experts and thinkers in any domain that have come before you. And I'll put it in your pocket, and now you can use it to do whatever you want with it. But it's not just you. I give it to your daughter, I give it to your mom, I give it to your sister. That's what I mean by—
Why is that different? I mean, you and I probably both have smartphones in our pockets, that, you know, if we were having this conversation five years ago, we would've said something similar, right? We have all the world's knowledge sitting in our pocket.
Knowledge is different than expertise and intelligence. Knowledge is being able to go and search for the design of a PN junction. You have no idea what that means.
You can Google it and you'll show it to somebody else. He also doesn't know what it means. But you're able to get that information. PN junction is the core structural unit on which we build a transistor. Now though what you can say is, design me a transistor, and you have a spec that you can now mail to somebody and they can actually tape that out for you, and you can get back in the mail an actual functional thing. It's this level of expertise that you can shape that was not available before. So you're taking knowledge and you're putting it through this process that makes it actionable. And I think that's the breakthrough that's so profound. This is why it's also so troubling. Like, how did we get in this place where half the people actually think this is a bad thing?
So let, we'll talk about that. Lemme first ask about something you said, which kind of the, not even democratization, but this leveling I think is the way you referred to it. We're also seeing at the same time though, folks who own these companies, or at least lead these companies in their larger shareholders, venture capitalists, which is something you do and have done. There seems to be definitely a capital inequality that gets formed by this. You talk about leveling. Is there also not a concern that just insane amounts of wealth, more than we've seen before, are gonna get concentrated in an even smaller number of people?
I think if we do it right, the answer to that question will actually be way better than what we want it to be. Does anybody wanna live in a world where there are individually five humans that win and 8.7 billion minus five who are not winning as much? Nobody wants that, including those five people, because their lives become absolutely terrible as well, okay? What do we all want? We all want to live in a place where two things are true. I decide to pursue something, nobody gets in my way. Nobody tells me I can't do it. That the amount of capital that is required to do it gets smaller, and smaller, and smaller and approaches zero, right? And then the second is, I live in a society where so many other people are doing those things that are equivalently what that means for me for them so that we all appreciate that we're all trying to make something useful for ourselves. I think everybody wants that.
Do you think that's where we're actually heading right now?
It's not where we are, but it's where we can go. So meaning, I had this famous meeting, I tweeted about this a lot. It was me and Zuck and DeAngelo, Adam DeAngelo, back when we worked at Facebook. And we were raising this round in the middle of all this commercial chaos right before the great financial crisis from Microsoft. And I think we raised like $1 billion, which now today seems like a pittance, but at the time just seemed like all the money in the world, had this crazy valuation. And we had this meeting with Bill Gates and it was something along the lines of, why do you think you're building a platform? And I go through this whole song and dance and he says something to the effect of, you know, 'That's not a platform.' He said it a little bit more colorful, but I'll sanitize it for your viewers.
It's YouTube, you can say what you want to say.
Oh yeah. It's like, are you fuckin' stupid? That's not a platform. A platform is when the economic value that the participants generate greatly exceeds that of the platform maker. That is not what this is telling us, basically that at the time our efforts at Facebook platform were mischaracterized. Forget about Facebook platform for a second 'cause it wasn't successful. But if you take that sentence and you, for example, look at the Google ecosystem, as much money as Google makes, it is undeniably true that the ecosystem that works around and on top of Google, all the advertisers, all the companies selling things, that economic value far exceeds what Google makes.
And that's what you think will happen here whether that be OpenAI or Anthropic. You think what's built on top of them will exceed the values there.
It has the potential if we don't fuck it up. So go back to this. So this is why I think it's so important for everybody to realize, we have the ability to put a genius level co-founder beside every single human, man, woman, child, independent of age, creed, country, sexual orientation, all this equality stuff that we've been blathering on about for years is actually finally possible. What's real equality? It's when you are as smart as the smartest human, you are as capable as the richest human, you're as ingenious as the most connected human. That's possible today. And so we should figure out a way to get out of our own way and give that to people.
When you say get out of our own way, what is one thing you would do that you think is currently in the way?
Well, I think, look, and I've suffered from this, so this is the kettle calling the pot black in some degrees. Part of the reason that Silicon Valley works is that a lot of the people that start here have had some pretty hard scrabbled backgrounds, complicated psychological relationships with their parents. Typical stuff. I don't think it's atypical. I thought I was special to have grown up so fucked up, turned out, no, everybody here is fucked up. Those boundary conditions generally create deep wells of insecurity. For many people, work fills that well, and then they're able to pour themselves into it to such a degree of obsession, they create these incredible things. That's what's driven this place. If you look at the great outliers here, they all come back to this general common denominator. I've struggled with that insecurity in filling it, but then it spilled over beyond that to, there's personal grudges, there's enmity, there's these complications. And specifically in AI, what you find is a very complicated origin story where there are factions, there are hurt feelings, there are people that have been lied to, there are people that have lied, there's misrepresentations, there's hirings, there's firings, there's complicated cap table. It was not as clean as Larry and Sergei met on campus. They liked arguing. They start Google, they get some venture capital, and off to the race as they go. It's, in fact, in many ways it's the exact opposite of that. So what you are seeing, I think, is in some measure, some amount of that personal insecurity and anger and frustration, and personal situations in the origin story of AI that have spilled out.
How does it manifest? I mean, is this the doomer days?
People can't control their own proclivities to want to either advance themselves and/or to hurt others by advancing themselves.
So why do you think with AI, the heads of these AI companies are saying, please regulate us because otherwise we might destroy humanity as we know it?
They're not even saying, please regulate us. I think they're saying we may destroy humanity. They're for sure saying that. And then what's odd to me is that once they've, and I think that part of it, if I had to try to unpack psychologically why, part of it was in a moment where there was such a voracious need for, you know, two companies specifically to get so much money, it was almost like trying to knife the other one while they were getting the money. Because if you look at the cycle, Dan, I think the cycle marries incredibly well, hopefully maybe your team can do an infographic. You would see people talking about the insanely apocalyptic characteristics of AI leading up to a model release, which usually was leading up to a fundraising. That was—
But they're still saying it today.
Lemme, so that was act one. Then in act two they would then basically talk about this doomerism part almost to try to just get all the agitators up in such a tizzy to almost sabotage the other people's processes they were embarking on. Then they would flip the narrative. They would go back to, oh, here's this mythical release that's about to come out that's gonna solve everybody's problems, but could also end the world. Investors would clamor to get in, and then they would start the doomer narrative. So they were going back and forth, back and forth. And to your point, I was looking at this thinking, beyond the game theory and just like the personal enmity driving this, this is a really bad strategy because going back to where we started, you're taking this tool that is really valuable to so many people, and you're gonna cast it in all of your own personal issues. And I find that deeply selfish.
Do you think, do you feel, I know you are not a doomer and you criticize doomerism, is there any part of you, a small part of you that thinks, 'I don't think that's what's gonna happen, but it might be.'
There is no scenario, and I, look, I think that I am pretty risk adjusted. I think probabilistically, I think I think through things pretty reasonably. There is no version of this where we're all just standing around doing nothing. There is no version of that. There is a version of it where we all are doing a lot more, we're all thinking about doing our jobs in a totally different way. And I think that that's very powerful. And then there'll be transitions. We've dealt with these transitions in the past.
Lemme ask it I guess in a different way, and this is the way I think about it. You've got kids, I've got a kid, I think about my daughter who happens to be 15 and I'm 50, and I think what, if she became really good at whatever it was she wanted to do, what job do I think she could have, or what career do I think she could go into that I'm certain when she's 50, she could be employed doing again, assuming she's good at it, and I can't come up with anything for sure.
And nor should you, and why are you like defining her value as a human being in such a basic way?
That's not value as a human being, but employment does matter.
Sure, but you're not saying that she, you, that's, but again, are you saying that you think she's just gonna be unemployed and a ward of the state?
No, what I'm saying is I can't, I think of every career, blue collar, white collar, et cetera, and given where we are with advancements, and not just in terms of AI, but AI and how that's applying to robotics, et cetera, I can't think of a thing necessarily that her or her friends, it's not just about her, or your kids will necessarily be able to do that is gainful employment in 35 years.
Because every job I can currently think of, I could make a case potentially that it could absolutely be replaced.
Plumber, you've got physical models, robot, an R2D2 version that is better at getting under cabinets than you or I would be, comes up to steps and can do that job.
Who owns the plumbing business? A robot?
Which would, a robotics company, which wouldn't necessarily need to have all that many people to run it anymore. You don't need an accounting back office, you don't need HR departments, you don't need all the things.
But there are no people in the world, maybe a couple that pee and poo, like I don't understand, like we can go unpack at all.
So you don't have, so this doesn't. It's an incredible headline, and I think it's great to spark a debate. But if it's then not rooted in some amount of like at least looking at some patterns of the past and trying to critically analyze, will there still be people in the world? Seems like that's growing unabated. Will they continue to live, eat, and breathe? Seems so. Will they continue to need housing? Seems so, clothing, seems so. Will they wanna do more things with their time? So for example, today, let's say that if you went back 1,000 years, how did a person allocate 24 hours a day? Probably a task, right? And if you fast forward 1,000 years, that person now has 35 tasks that they can allocate their time across. Some are work, some are leisure, some are with their family, some are vacationing, some are thinking, there's a panoply of options that technology and progress has given the human individual in order to allocate their time. I suspect if you just trend it, that 35 things now goes to 300 things over the next 1,000 years. There's gonna be more ways in which we allocate time. Maybe what we find is all of this stuff about, you know, a poorly developed prefrontal cortex and all this stuff that we today label ADHD and context switching actually becomes an unlock in the future. Because what you're able to do is context switch rapidly across many of the things where you can allocate your time, and maybe how we've defined concentration changes. My point is there's always to have this argument that aren't as reductive as, it's all over, let's just give up.
On the other piece of doomerism isn't the labor side, but is, you've talked about kind of the end of humanities, the bio weapons piece. The war piece. Does that worry you or does it not?
Yeah. Every single technology that's ever been created is fundamentally deterministic. What does that word mean? It means anything that is possible will be possible and done. With respect to technology, every single time we've invented something, every single time, we've had the good cases and the bad cases, it's come as a basket. We've never been able to cork innovation in a way. All we've been able to do is shape it where it's overwhelmingly good, where it's overwhelmingly positive. So to your point, in the distribution of outcomes, when you create a new tool, like when we figured out how to take atoms and smash 'em together, the overwhelming majority of those benefits are seen in the lights that are here. It's the energy that's relatively abundant, it's super clean. But we also know Hiroshima and Nagasaki, and we know that there are tail risks. It will be the same for AI. There will be nefarious and bad uses of it. I think the question that you're asking me is, could they dominate? And I think that we have to find the right checks and balances. And again, this is why I think if we had that debate, that's a lot more grounded and rational. It's about technology and terms like KYC that are like really boring I guess.
You've been obviously a supporter of President Trump in his most recent campaign. He's been pretty light touch when it comes to AI. Do you think something like that KYC is something he would go for?
It's a great question. I dunno, 'cause I haven't asked him, but I think it's a very reasonable thing because, you know, we have KYC, conceptual KYC as implemented in pharma. We have conceptual KYC as implemented in the Department of Commerce, in treasury, for all different kinds of things. There are reasonable checks and balances to get access to the bleeding edge or the leading edge. This is a reasonable class of product that is important enough that we should do that.
You said you hadn't asked Trump specifically about this. What is it like talking AI with him?
Oh, he's really inspiring. I mean, I have to say here, here's the thing with him, that in my limited interactions with him, but he's really curious. He's really smart. His intuition is off the charts. I'll tell you a very specific example. Had a meeting with him, had a deck, I had a whole, I had like a complicated—
What were you pitching him?
It's not a pitch. I was just like walking him through like some structural issues explaining, in this specific case it was like data centers and electricity rates, and you know, the underlying reasons why electricity prices are going up, which have more to do with the business model of the utilities, have very little to actually do with data centers, water consumption, and then some structural ways to, you know, decrease Americans' electricity, right? It's like a 15-page deck, and he's one of these guys where he's listening, he'll scan it, he's scanning it while he's talking to you. He listens to you, and then he goes right to the one thing. And I have a lot of respect for this because like I've seen some of the most incredible survivors, like the people that have really thrived over decades. They have that one skill. They can like go in and go that thing, that little detail, that thread and pull it. So my experience with him is like, it's quite impressive, when I anchor him to the other people in business that I've interacted with, he's really up there.
What's the smartest or, if not smartest, maybe most interesting question he has asked you about AI?
