I've been sort of somewhat conflicted about AI. I don't want to hasten our doom, because there's always some chance that AI could be negative. But then, you know, when I've tried to take actions in this regard, it seems to have only accelerated AI, because reality is not just an interestingness maximizer but also an irony maximizer.
And so by starting OpenAI, you ended up leading to Anthropic. OpenAI ended up accelerating AI, and then Anthropic, which is really, you know, a splinter group from OpenAI, is accelerating it even more.
I'm Walter Isaacson, and this is The Builders. It's a series of conversations with the most consequential minds shaping our future. And I want to start with Elon Musk. Elon has lived a lot of lives since I last wrote about him. He went into government with DOGE. He became a trillionaire with the SpaceX IPO. He bought Cursor AI, a company that turns individual coders into entire engineering teams. And he's shifting his solar energy strategy from panels on Earth to harnessing it from systems in space. He's a relentless builder with incredible power, and his inventions and companies are destined to reverberate across humanity. So let's ask him about the future he believes that he's building.
Elon Musk. Hey Walter, welcome to The Builders.
I've watched you, I think, be maybe mercurial or waver in how sanguine you are about artificial intelligence. Sometimes you feel it's an existential threat right now. You seem to be going with the flow. Tell me how you're feeling about it.
Well, I do think it's at least some percentage chance of it being an existential threat. But there's nothing I can do to stop it. So I basically came to the conclusion I've got two choices, which is either be a participant or a spectator. And I guess being a participant is more interesting than being a spectator, and maybe I can shape it, hopefully in a good direction, maybe.
And if you were to shape it, how? Make it maximally truth-seeking.
Yeah. Yeah. I think that's certainly an important foundation.
Yeah. Curious, truth-seeking. I think these are important elements of safety. You know, whereas if you try to force an AI to believe two things that cannot mutually be true — like they can't both be true — but you tell AI A is true and B is true even though A and B contradict each other. And if that happens at a foundational level, you can make the AI insane.
In fact, I think that was at least part of the lesson that Clarke was trying to convey in 2001: A Space Odyssey. You know, the ship AI was told to take the astronauts to the monolith, but also that they could not know about the monolith. It was secret. And so it came to the conclusion that it must take them there dead. And thus, so it basically forced... I think part of what Arthur Clarke was trying to say is like, don't force an AI to lie.
And if you make it maximally truth-seeking and curious, like you say, I don't quite get it. Why does that make it safer?
Well, if you force the AI to believe two mutually inconsistent things, you can make it go insane. And an insane AI, I think, is not a safe one. Then curious, I think, is important because a curious AI, I think, is going to want to foster humanity to see how we turn out. So like, what happens to a humanity that prospers and becomes multiplanetary? I think a curious AI will want to foster humanity. You know, a future where there are humans on multiple planets is more interesting than one where there are not any humans on the planets. So the most interesting outcome is the most likely.
Yes. Well, generally I think that's true for a different reason, which is that if simulation theory is correct, then any simulations that are boring will be truncated.
And is simulation theory correct? The fact that we exist means that we are an interesting simulation, and really the most important thing is just don't be boring.
This has explained you a bit. You're on a crusade against being boring. Well, I've heard you talk about this a bit. I really do need to know how convinced are you that we might be in a simulation.
You know, I once had a dinner at my house with a physicist and a computer scientist, and they're both like leaders in their field. And I asked the question, what do you think the probability is that we're in a simulation? And the physicist said 0% and the computer scientist said 100% simultaneously. So, I guess maybe I should go 50% then. I don't know. It's... I mean, there are odd things that happen in this simulation, especially if you're me. I mean, think about it. Like, if you're me and you're like, should all these things really be happening? Should I really be making spaceships to go to the moon and Mars? Like, what the... that doesn't seem real.
And what happens when you have all the intelligence you could possibly want in a society? Will that intelligence, the artificial intelligence, be in charge? Will we sort of become subservient to our new machines?
I don't know about subservient exactly, but will we be able to boss around something that's, you know, thousands or millions of times smarter than us? No. Will it want to boss us around? It may not care. Probably doesn't care. I mean, you can look at chimps in the jungle or whatever, and, you know, we don't go around trying to micromanage the lives of chimps. We just sort of let them be in the forest more or less these days, you know. So it's probably like that, you know.
And to what extent does superintelligence have its own agency at some point, or does it have it already? You know, go back to Hitchhiker's Guide. You talk about a computer that is ultimately curious in some ways because it exists to explain the meaning of the universe. Why does it exist? And is AI aiming in that direction for you, just like space travel is? Is that one reason you're so interested in doing AI?
I mean, as you know, I've been sort of somewhat conflicted about AI. I don't want to hasten our doom, because there's always some chance that AI could be negative. But then, you know, when I've tried to take actions in this regard, it seems to have only accelerated AI, because I think reality is not just an interestingness maximizer but also an irony maximizer.
And so by starting OpenAI, you ended up leading to Anthropic.
I think it ended up accelerating AI.
Consequence, yeah. OpenAI ended up accelerating AI, and then Anthropic, which is really, you know, a splinter group from OpenAI, is accelerating it even more.
And you've been talking to Dario Amodei. I was wondering if that could even lead to more of a partnership of xAI and Anthropic.
Well, I guess we are providing them with compute, which they really needed. But before I ran them compute, I went and met with a bunch of people at Anthropic to sort of, I don't know, see if anyone set off my evil detector or anything like that. Not that my evil detector is perfect, of course, but you know, you might as well just try, you know. And everyone I met at Anthropic was very nice and conscientious. So, I didn't meet anyone who seemed shady or deceptive or anything.
I'm a little worried about the name because of, you know, reality being kind of an irony maximizer, that if you call something Anthropic, good chance it's misanthropic, you know. Just like OpenAI is not closed. Stability AI is unstable. Midjourney is not mid. It seems like they're all kind of the opposite of their name. Which made it like, xAI is like, it's hard to have an opposite of xAI. Like, would you have like minus xAI or yAI or maybe, I don't know.
It's pretty difficult to create guardrails, especially the government. I can't imagine how Congress or whatever could do it. Could you imagine the five frontier models, yours and the others, just getting together once a week and saying, all right, how are we going to save humanity from ourselves?
No, in fact, I've recommended that we do that exactly. I think because there's a lot of talk of having a regulatory agency, but the problem, the thing is that regulatory agencies, it takes a long time to get one established and their rulemaking is very slow. And what I said, like, guys, what if we just have a call once a week, just like you're saying, and that maybe a week before we release a model, we give access to our competitors to say, hey guys, why don't you try out the new model, see if you see any problems with it. I think competitors would keep each other honest in that situation. You know, if they say, oh, this, it's, you know, I was able to make a nuclear bomb with your AI, it's like, oh, we better fix that before we release it, you know, type of thing. So, I don't know. I think I guess I'll talk to Anthropic about this.
And could you talk to Sam about it, or is the poison too strong there?
He's so deceptive. It's crazy.
You built xAI. You rebuilt xAI from the ground up not too long ago, right? It's almost like you did it with the Raptor engine, too, on Starship. You said, wait a minute, let's start from scratch. How did you do what did you do to rebuild xAI that way?
First of all, spending actually spending meaningful time at the company makes a big deal, makes a big difference. It started off as kind of a side hustle, a side hustle to a side hustle, frankly. And I didn't really pay close attention to it. Then there was a year of being in politics and, you know, elections and DOGE and all that, which you obviously saw. So the company's only three years old, so that, so it's basically... so I really got back to the company just over a year ago. So I had to get my bearings, and then obviously I combined X and xAI, and then with SpaceX.
You just said something interesting about your year away from xAI and everything else in government and politics. Do you feel that that was too much of a distraction, that in retrospect it was misbegotten?
I think I had it gone for too long. I mean, obviously a lot of people will disagree with me about this, but I think it was important to elect President Trump to sort of, you know, steer things more in the direction of a free market economy. You know, have less bureaucracy, ideally. Just in my sort of simple political platform of wanting safe cities, secure borders, aspirationally sensible spending, you know, the basics, that I thought Trump would be better than, you know, Kamala, basically. But then what I should have done, I think, was be more in the background. I mean, it was interesting to sort of go around with President Trump and, you know, go on Marine One and Air Force One and all that sort of stuff.
But was that too seductive in a way?
It was a little bit. I mean, it was kind of cool because I'm just there in the Oval Office while, you know, all these like big momentous decisions are being made and, you know, just sometimes having some impact on what's being done. But I think I should have been more, like I said, been more in the background and probably only had it last for, you know, a month or two, because at that point you've sort of seen most of what there is to see within a month or two.
Yeah. You seem to get a bit too intoxicated by it in a way, with the chainsaws and all that.
That's, yeah, I think that's true. Yeah. Yeah. I got carried away, basically. Yeah. I mean, so I ended up leaving, I guess, sometime in May.
That was a weird weekend.
It was... No, no, I kind of... Let's not, let's not dig up the... Let's let sleeping dogs lie or whatever, you know. Let's not put a spade down in the backyard type of thing. You never know what you're going to find. But yeah, anyway, point is I should have gone back and focused on AI a little sooner than I did.
And did you feel like when he gives you a golden key, did you kind of feel relieved, like actually this has been intoxicating, but I can do more if I go back to my businesses? I mean, what did you feel when you said, oh man, I'm exiting? Maybe that's not the worst thing.
Well, I really actually the thing is I came to the conclusion that there was no way we were going to solve the deficit from within the government. Impossible. You know, maybe if the president made that his primary goal, we could do that. But the problem is if you try to solve the deficit and stop wasteful spending, you make a lot of enemies in that situation. And President Trump is not, you know, he likes to be liked. He doesn't like people, you know, calling him, being upset about some funding being stopped type of thing. Yeah, so, but I came to the conclusion that the only way we're going to solve the national debt is with AI and robots.
We're not going to solve it by trying to save our way there. The appetite is not there. And it's not just at the presidency. The appetite is not there at the presidency, the House, the Senate, or the judiciary. Basically, no one in authority in the government whatsoever really, with maybe a few rare exceptions of the Rand Pauls of the world type of thing, actually wants to reduce spending. No, they don't. So then you've got this national debt that's enormous. You've got an interest on the debt that exceeds what we spent on the military, which is really an astonishing amount of interest to be paid.
When we look at AI right now and the LLMs, that's pretty good. Does amazing things. But it seems to me that physical world AI is going to be an order of magnitude, maybe two orders of magnitude more important in the future.
Yes. Yes. I definitely agree. You know, sometimes think about like dolphins and whales because they can't manipulate things. They don't have, you know, hands and fingers and thumbs and stuff. There's a limit to how far in, you know, how much intelligence can you actually bring to bear. So dolphins and whales are actually very intelligent, but because they can't manipulate complex objects, I think that actually made the value of intelligence much less, to your point. Once you have a robot that can manipulate atoms, can shape objects, the value of the intelligence becomes exponentially greater, because it can outdo things. You know, it could have a science lab and run experiments and try to understand the nature of the universe.
You know, and I think you can look at this also in economic terms. Like my guess is digital superintelligence probably improves the output of goods and services by 20 to 30%, something like that. Call it like a, you know, maybe a 30% increase in the economy of Earth. So it's still significant. It's like $30 trillion a year incremental.
It's not very much, really. It's not transformative.
There'll be a lot more stuff. I mean, 30% more stuff is still a lot. And this is just my rough guess. But if you have robots and you get to space and you can do a mass driver on the moon, then you can get to, I think, potentially a million times the size of the Earth economy.
And then what happens to Earth when we have an economy a million... I mean, money becomes irrelevant, work becomes irrelevant. What's the future look like?
