Latest / Elon Musk Thinking / Latest Interview of Elon Musk: Humans Won't Be in Control. When AI Becomes Smarter Than Us!!!

Transcript

Latest Interview of Elon Musk: Humans Won't Be in Control. When AI Becomes Smarter Than Us!!! · Elon Musk Thinking

• 0:00
  Speaker 1: Latest interview of Elon Musk.
• 0:02
  Speaker 2: Welcome to The Builders.
• 0:04
  Speaker 1: Thanks.
• 0:05
  Speaker 2: I've watched you, I think, be maybe mercurial or waver
• 0:10
  Speaker 2: in how sanguine you are about artificial intelligence. Sometimes you
• 0:15
  Speaker 2: feel it's an existential threat, right? Now you seem to
• 0:19
  Speaker 2: be going with the flow. Tell me how you're feeling
• 0:21
  Speaker 2: about it.
• 0:22
  Speaker 1: Well, I do think it's at least some percentage chance
• 0:25
  Speaker 1: of it being an existential threat. So... But there's nothing
• 0:30
  Speaker 1: I can do to stop it. So I basically came
• 0:34
  Speaker 1: to the conclusion I've got two choices, which is either
• 0:37
  Speaker 1: be a participant or a spectator. And I guess being
• 0:41
  Speaker 1: a participant is more interesting than being a spectator. And
• 0:44
  Speaker 1: maybe I can shape it, hopefully, in a good direction.
• 0:47
  Speaker 2: And if you were to shape it, how? Make it
• 0:49
  Speaker 2: maximally truth-seeking?
• 0:50
  Speaker 1: Yeah, I think that's certainly an important foundation.
• 0:54
  Speaker 2: And make it curious?
• 0:56
  Speaker 1: Yeah, curious, truth-seeking. I think these are important elements of safety.
• 1:02
  Speaker 1: Whereas if you try to force an AI to believe
• 1:08
  Speaker 1: two things that cannot mutually be true, they can't both
• 1:12
  Speaker 1: be true, but you tell AI A is true and
• 1:16
  Speaker 1: B is true, even though A and B contradict each other.
• 1:20
  Speaker 1: And if that happens at a foundational level, you can
• 1:22
  Speaker 1: make the AI insane. In fact, I think that was
• 1:28
  Speaker 1: at least part of the lesson that Odyssey Clock was
• 1:30
  Speaker 1: trying to convey in 2001, Space Odyssey. You know, the
• 1:33
  Speaker 1: ship AI was told to take the astronauts to the monolith,
• 1:37
  Speaker 1: but also that they could not know about the monolith.
• 1:40
  Speaker 1: It was secret. And so it came to the conclusion
• 1:44
  Speaker 1: that it must take them, they're dead. And thus, so
• 1:51
  Speaker 1: it basically, I think part of what Odyssey Clock was
• 1:54
  Speaker 1: trying to say is like, don't force an AI to lie.
• 1:58
  Speaker 2: And if you make it maximally truth-seeking and curious, like
• 2:03
  Speaker 2: you say, I don't quite get it. Why does that
• 2:05
  Speaker 2: make it safer?
• 2:07
  Speaker 1: Well, if you force the AI to believe two mutually
• 2:09
  Speaker 1: consistent things, you can make it go insane. And an
• 2:14
  Speaker 1: insane AI, I think, is not a safe one. Then curious,
• 2:18
  Speaker 1: I think, is important because a curious AI, I think,
• 2:22
  Speaker 1: is going to want to foster humanity to see how
• 2:25
  Speaker 1: we turn out. So, like, what happens... to a humanity
• 2:30
  Speaker 1: that prospers and becomes multi-planetary. I think a curious AI
• 2:36
  Speaker 1: will want to foster humanity. A future where there are
• 2:40
  Speaker 1: humans on multiple planets is more interesting than one where
• 2:44
  Speaker 1: there are not any humans on the planets.
• 2:46
  Speaker 2: So the most interesting outcome is the most likely?
• 2:49
  Speaker 1: Yes. Well, generally I think that's true for a different reason,
• 2:53
  Speaker 1: which is that if simulation theory is correct, then Any
• 2:57
  Speaker 1: simulations that are boring will be truncated.
• 3:00
  Speaker 2: And is simulation theory correct?
• 3:01
  Speaker 1: The fact that we exist means that we are an
• 3:04
  Speaker 1: interesting simulation. And really, the most important thing is just
• 3:08
  Speaker 1: don't be boring.
• 3:09
  Speaker 2: This has explained you a bit. You're on a crusade
• 3:12
  Speaker 2: against being boring. Well, I've heard you talk about this
• 3:14
  Speaker 2: a bit. I really do need to know. How convinced
• 3:18
  Speaker 2: are you that we might be in a simulation?
• 3:22
  Speaker 1: You know, I once had a dinner... at my house
• 3:26
  Speaker 1: with a physicist and a computer scientist, and they're both
• 3:29
  Speaker 1: like leaders in their field. And I asked the question,
• 3:32
  Speaker 1: what do you think the probabilities that were in a simulation?
• 3:35
  Speaker 1: And the physicist said 0% and the computer scientist said
• 3:38
  Speaker 1: 100% simultaneously. So I guess maybe I should go 50% then.
• 3:45
  Speaker 1: I don't know. I mean, there's odd things that happen
• 3:49
  Speaker 1: in the simulation where, especially if you're me, I mean,
• 3:52
  Speaker 1: think about it. Like if you're me and you're like,
• 3:55
  Speaker 1: Should all these things really be happening? Should I really
• 3:58
  Speaker 1: be making spaceships to go to the moon and Mars? Like,
• 4:02
  Speaker 1: what the... That doesn't seem real.
• 4:05
  Speaker 2: And what happens when you have all the intelligence you
• 4:09
  Speaker 2: could possibly want in a society? Will that intelligence, the
• 4:14
  Speaker 2: artificial intelligence, be in charge? Will we sort of become
• 4:18
  Speaker 2: subservient to our new machines?
• 4:20
  Speaker 1: I don't know about subservient... Uh, exactly. But will we
• 4:28
  Speaker 1: be able to boss around something that's, you know, thousands
• 4:32
  Speaker 1: or millions of times smarter than us? Uh, no. Well,
• 4:36
  Speaker 1: will it want to boss us around? I may not care.
• 4:39
  Speaker 1: Probably doesn't care. I mean, you can look at chimps
• 4:41
  Speaker 1: in the, in the, in the jungle or whatever. And,
• 4:43
  Speaker 1: you know, we don't go around trying to micromanage the
• 4:45
  Speaker 1: lives of chimps. Um, we just sort of let them
• 4:48
  Speaker 1: be in the forest more or less these days, you know? Um,
• 4:52
  Speaker 1: so I'd probably like that, you know.
• 4:55
  Speaker 2: And to what extent does superintelligence have its own agency
• 4:59
  Speaker 2: at some point, or does it have it already? You know,
• 5:02
  Speaker 2: go back to Hitchhiker's Guide. You talk about a computer
• 5:06
  Speaker 2: that is ultimately curious in some ways because it exists
• 5:10
  Speaker 2: to explain the meaning of the universe. Why does it exist?
• 5:14
  Speaker 2: And is AI aiming in that direction for you just
• 5:17
  Speaker 2: like space travel is? Is that one reason you're so
• 5:21
  Speaker 2: interested in doing AI?
• 5:23
  Speaker 1: I mean, as you know, I've been sort of somewhat
• 5:25
  Speaker 1: conflicted about AI. I don't want to hasten our doom
• 5:30
  Speaker 1: because there's always some chance that AI could be negative.
• 5:33
  Speaker 1: But then, you know, when I've tried to take actions
• 5:35
  Speaker 1: in this regard, it seems to have only accelerated AI
• 5:39
  Speaker 1: because I think reality is not just an interestingness maximizer,
• 5:44
  Speaker 1: but also an irony maximizer.
• 5:46
  Speaker 2: And so by starting OpenAI... you ended up maximizing by
• 5:50
  Speaker 2: leading to Anthropic.
• 5:52
  Speaker 1: I think it ended up accelerating AI.
• 5:54
  Speaker 2: Unintended consequence.
• 5:56
  Speaker 1: Yeah, OpenAI ended up accelerating AI, and then Anthropic, which
• 5:59
  Speaker 1: is really a splinter group from OpenAI, is accelerating it
• 6:05
  Speaker 1: even more.
• 6:06
  Speaker 2: And you've been talking to Dario Amadei. I was wondering
• 6:09
  Speaker 2: if that could even lead to more of a partnership
• 6:12
  Speaker 2: of XAI, Anthropic.
• 6:15
  Speaker 1: Well, I guess we are providing them with the compute,
• 6:18
  Speaker 1: which they, they really needed. But before I ran the computer,
• 6:20
  Speaker 1: I went and met with a bunch of people at,
• 6:23
  Speaker 1: at Anthropic to sort of, I don't know, see if
• 6:25
  Speaker 1: anyone set off my evil detector or anything like that.
• 6:27
  Speaker 1: Not that my evil detector is perfect, of course, but
• 6:30
  Speaker 1: you know, you might as well just try, you know? Um,
• 6:34
  Speaker 1: and everyone I met at Anthropic was very nice, uh,
• 6:38
  Speaker 1: and conscientious. So that'd mean anyone who seemed shady or
• 6:45
  Speaker 1: deceptive or anything. Um, I'm a little worried about the
• 6:49
  Speaker 1: name because of the, you know, reality being kind of
• 6:54
  Speaker 1: an irony maximizer that if you'd call something anthropic, you know,
• 6:59
  Speaker 1: good chances, misanthropic, you know, um, just like opening eyes,
• 7:04
  Speaker 1: not closed, um, stability, AI is unstable. Uh, my journey
• 7:10
  Speaker 1: is not mid. Uh, it seems like they were all
• 7:14
  Speaker 1: kind of the opposite of their name. Um, Which made
• 7:18
  Speaker 1: it like, like XAI is like, it's hard to have
• 7:20
  Speaker 1: an opposite of XAI, but would you have like minus
• 7:22
  Speaker 1: XAI or YAI or maybe, I don't know.
• 7:26
  Speaker 2: It's just pretty difficult to create guardrails, especially the government.
• 7:30
  Speaker 2: I can't imagine how Congress or whatever could do it.
• 7:33
  Speaker 2: Could you imagine the five frontier models, yours and the others,
• 7:38
  Speaker 2: just getting together once a week and saying, all right,
• 7:41
  Speaker 2: how are we going to save humanity from ourselves?
• 7:44
  Speaker 1: No, in fact, I've recommended that we do that exact
• 7:46
  Speaker 1: thing because there's a lot of talk of having a
• 7:49
  Speaker 1: regulatory agency. But the thing is that regulatory agencies, it
• 7:52
  Speaker 1: takes a long time to get one established and their
• 7:56
  Speaker 1: rulemaking is very slow. And I said, like, guys, why
• 8:01
  Speaker 1: don't we just have a call once a week, just
• 8:03
  Speaker 1: like you're saying, and that maybe a week before we
• 8:10
  Speaker 1: release a model, we give access to our competitors and
• 8:14
  Speaker 1: to say, hey guys, why don't you try out the
• 8:15
  Speaker 1: new model, see if you see any problems with it.
• 8:17
  Speaker 1: I think competitors would keep each other honest in that situation.
• 8:22
  Speaker 1: You know, if they say, oh, I was able to
• 8:26
  Speaker 1: make a nuclear bomb with your AI, it's like, oh,
• 8:29
  Speaker 1: we better fix that before we release it, you know,
• 8:31
  Speaker 1: type of thing. So, I don't know, I guess I'll
• 8:37
  Speaker 1: talk to Anthropic about this.
• 8:39
  Speaker 2: Could you talk to Sam about it, or is the
• 8:41
  Speaker 2: poison too strong there?
• 8:44
  Speaker 1: He's so deceptive, it's crazy.
• 8:46
  Speaker 2: You rebuilt XAI from the ground up not too long ago, right?
• 8:52
  Speaker 2: It's almost like you did it with the Raptor engine, too,
• 8:55
  Speaker 2: on Starship. You said, wait a minute, let's start from scratch.
• 9:00
  Speaker 2: What did you do to rebuild XAI that way?
• 9:04
  Speaker 1: First of all, actually spending meaningful time at the company
• 9:09
  Speaker 1: makes a big difference. It started off as kind of
• 9:12
  Speaker 1: a side hustle. Um, a side hustle to a side hustle, frankly. Um,
• 9:17
  Speaker 1: and I didn't really pay close attention to it. Then
• 9:20
  Speaker 1: there was, uh, a year of the, uh, of being
• 9:25
  Speaker 1: in politics and, you know, elections and doors and all that,
• 9:28
  Speaker 1: which you obviously saw. Uh, so the company is only
• 9:31
  Speaker 1: three years old, so that, so it's basically, uh, this is,
• 9:35
  Speaker 1: and so I really got back to the company just
• 9:39
  Speaker 1: over a year ago. Um, so I had to get
• 9:42
  Speaker 1: my bearings. Obviously, I combined X and XAI and then
• 9:49
  Speaker 1: with then the SpaceX.
• 9:51
  Speaker 2: You just said something interesting about your year away from
• 9:55
  Speaker 2: XXAI and everything else in government and politics. Do you
• 9:59
  Speaker 2: feel that that was too much of a distraction, that
• 10:02
  Speaker 2: in retrospect it was misbegotten?
• 10:05
  Speaker 1: I think I had it gone for too long. Obviously,
• 10:10
  Speaker 1: a lot of people will disagree with me about this,
• 10:12
  Speaker 1: but I think it was important to elect President Trump
• 10:16
  Speaker 1: to steer things more in the direction of a free
• 10:20
  Speaker 1: market economy, you know, have less bureaucracy, ideally. Just in
• 10:26
  Speaker 1: my sort of simple political platform of wanting safe cities,
• 10:31
  Speaker 1: secure borders, aspirationally sensible spending, you know, the basics, that
• 10:39
  Speaker 1: I thought Trump would be better than, you know, Kamala, basically. So,
• 10:46
  Speaker 1: but then... What I should have done, I think, was
• 10:52
  Speaker 1: be more in the background. I mean, it was interesting
• 10:55
  Speaker 1: to sort of go around with President Trump and, you know,
• 10:58
  Speaker 1: go on the Marine One and Air Force One and
• 11:01
  Speaker 1: all that sort of stuff.
• 11:02
  Speaker 2: But was that too seductive in a way?
• 11:05
  Speaker 1: It was a little bit. I mean, it was kind
• 11:07
  Speaker 1: of cool because I'm just there in the Oval Office while,
• 11:10
  Speaker 1: you know, all these, like, big momentous decisions are being made. And,
• 11:13
  Speaker 1: you know, sometimes having some impact on what's being done.
• 11:16
  Speaker 1: But I think I should have been more in the
• 11:25
  Speaker 1: background and probably only had it last for a month
• 11:30
  Speaker 1: or two. Because at that point, you've sort of seen
• 11:36
  Speaker 1: most of what there is to see within a month
• 11:38
  Speaker 1: or two.
• 11:39
  Speaker 2: Yeah, you seem to get a bit too intoxicated by it,
• 11:42
  Speaker 2: in a way, with the chainsaws and all that. I mean,
• 11:44
  Speaker 2: it's nuts.
• 11:45
  Speaker 1: Yeah, I think that's true. I got carried away, basically. Yeah,
• 11:49
  Speaker 1: I mean, so I ended up leaving sometime in May.
• 11:52
  Speaker 1: That was a weird weekend. Yeah, it was.
• 12:01
  Speaker 2: Tell me about it.
• 12:02
  Speaker 1: No, no, I can't.
• 12:03
  Speaker 3: Let's not take up the... Let's let sleeping dogs lie
• 12:08
  Speaker 3: or whatever, you know. Let's not put a spade down
• 12:12
  Speaker 3: in the backyard type of thing. You never know what
• 12:14
  Speaker 3: you're going to find.
• 12:15
  Speaker 1: But, yeah, anyway, the point is I should have gone
• 12:18
  Speaker 1: back and focused on AI a little sooner than I did.
• 12:23
  Speaker 2: And did you feel like when he gives you a
• 12:26
  Speaker 2: golden key, you go, did you kind of feel relieved? Like, actually,
• 12:29
  Speaker 2: this has been intoxicating, but I can do more if
• 12:33
  Speaker 2: I go back to my businesses? I mean, what did
• 12:35
  Speaker 2: you feel when you said, oh, man, I'm exiting. Maybe
• 12:38
  Speaker 2: that's not the worst thing.
• 12:40
  Speaker 1: Well, actually, the thing is I came to the conclusion
• 12:42
  Speaker 1: that– that there was no way we were going to
• 12:46
  Speaker 1: solve the deficit from within the government. Impossible. You know,
• 12:52
  Speaker 1: maybe if the president made that his primary goal, we
• 12:56
  Speaker 1: could do that. But the problem is if you try
• 12:59
  Speaker 1: to solve the deficit and stop wasteful spending, you make
• 13:04
  Speaker 1: a lot of enemies in that situation. And President Trump
• 13:08
  Speaker 1: is not, you know, he likes to be liked. He's not.
• 13:13
  Speaker 1: He doesn't like people calling him, being upset about some
• 13:18
  Speaker 1: funding being stopped type of thing. I came to the
• 13:28
  Speaker 1: conclusion that the only way we're going to solve the
• 13:32
  Speaker 1: national debt is with AI and robots. We're not going
• 13:35
  Speaker 1: to solve it by trying to save our way there.
• 13:37
  Speaker 1: The appetite is not there, and it's not just at
• 13:39
  Speaker 1: the presidency. The appetite is not there at the presidency,
• 13:43
  Speaker 1: the House, the Senate, or the judiciary. Basically, no one
• 13:48
  Speaker 1: in authority in the government whatsoever really, with maybe a
• 13:52
  Speaker 1: few rare exceptions of the grand polls of the world
• 13:54
  Speaker 1: type of thing, actually wants to stop, reduce spending. No,
• 14:00
  Speaker 1: they don't. So then you've got this national debt that's enormous.
• 14:03
  Speaker 1: You've got interest on the debt that exceeds what we
• 14:06
  Speaker 1: spent on the military, which is a really astonishing amount
• 14:10
  Speaker 1: of interest to be paid.
• 14:12
  Speaker 2: When we look at AI right now and the LLMs,
• 14:16
  Speaker 2: that's pretty good, does amazing things. But it seems to
• 14:19
  Speaker 2: me that physical world AI is going to be an
• 14:22
  Speaker 2: order of magnitude, maybe two orders of magnitude more important
• 14:25
  Speaker 2: in the future.
• 14:26
  Speaker 1: Yes, yes, I definitely agree. You know, sometimes I think
• 14:29
  Speaker 1: about like dolphins and whales, because they can't manipulate things,
• 14:33
  Speaker 1: they don't have hands and fingers and thumbs and stuff.
• 14:39
  Speaker 1: There's a limit to how much intelligence can you actually
• 14:44
  Speaker 1: bring to bear. Dolphins and whales are actually very intelligent,
• 14:52
  Speaker 1: but because they can't manipulate complex objects, I think that
• 14:59
  Speaker 1: actually made the value of intelligence much less. To your point,
• 15:03
  Speaker 1: once you have a robot that can manipulate atoms, can
• 15:08
  Speaker 1: shape objects, the value of the intelligence becomes exponentially greater
• 15:16
  Speaker 1: because it can now do things. It can have a
• 15:19
  Speaker 1: science lab and run experiments and try to understand the
• 15:23
  Speaker 1: nature of the universe. You can look at this also
• 15:27
  Speaker 1: in economic terms. My guess is digital superintelligence probably improves the, uh,
• 15:36
  Speaker 1: output of goods and services by 20 to 30%, something
• 15:38
  Speaker 1: like that. Call it like a, you know, maybe a
• 15:41
  Speaker 1: 30% increase in, uh, the, the, uh, economy of earth.
• 15:44
  Speaker 1: So there's still significant. It's like $ 30 trillion a year. Incremental.
• 15:48
  Speaker 2: Is that very much really? Is that transformative?
• 15:52
  Speaker 1: It'll, there'll be a lot more stuff. I mean, 30%
• 15:55
  Speaker 1: more stuff is still a lot, and this is just my,
• 15:58
  Speaker 1: you know, my rough guess. Um, But if you have
• 16:02
  Speaker 1: robots and you get to space and you can do
• 16:05
  Speaker 1: a mass driver on the moon, then you can get to,
• 16:09
  Speaker 1: I think, potentially a million times the size of the
• 16:12
  Speaker 1: Earth economy.
• 16:13
  Speaker 2: And then what happens to Earth when we have an economy?
• 16:15
  Speaker 2: I mean, money becomes irrelevant, work becomes irrelevant. What's the
• 16:19
  Speaker 2: future look like?
• 16:20
  Speaker 1: Yeah, money becomes irrelevant. I mean, there will be still currency,
• 16:25
  Speaker 1: but I think currency will be measured in fundamental physics
• 16:28
  Speaker 1: metrics like... you know, power, like wattage and tonnage, mass
• 16:34
  Speaker 1: and energy. You know, mass and power or energy, basically. So...
• 16:43
  Speaker 1: But conventional currency will not matter in the future. So
• 16:47
  Speaker 1: that's my guess. I think it's pretty obvious. And as
• 16:53
  Speaker 1: soon as you start getting to many orders of magnitude
• 16:55
  Speaker 1: bigger than Earth's economy, satisfying whatever humans can possibly think
• 17:00
  Speaker 1: of for their needs is very easy. You know, if
• 17:04
  Speaker 1: you're several orders of magnitude bigger than the Earth's economy,
• 17:07
  Speaker 1: then providing humans with whatever goods and services they can
• 17:13
  Speaker 1: think of that they might want is trivial.
• 17:16
  Speaker 2: And so what do we do then as humans? What
• 17:21
  Speaker 2: gives us meaning? How do we organize ourselves?
• 17:24
  Speaker 1: I do struggle with this question. I think people will find...
• 17:30
  Speaker 1: ways to be happy in the same way that, say,
• 17:33
  Speaker 1: people who retire seem to be fairly happy, even though
• 17:36
  Speaker 1: they're not working. But I think the reality is that
• 17:40
  Speaker 1: humans will not, there won't be anything that humans can
• 17:43
  Speaker 1: do that the AI and the robots can't do better.
• 17:46
  Speaker 1: One analogy might be gardening, you know, where you don't
• 17:53
  Speaker 1: have to grow your own fruits and vegetables. You can
• 17:56
  Speaker 1: just go buy them from the store. But sometimes people,
• 18:00
  Speaker 1: like to grow their own fruits and vegetables and make
• 18:03
  Speaker 1: dinner for friends, and it's kind of a nice touch.
• 18:05
  Speaker 2: What about creating art or that type of creativity? Is
• 18:08
  Speaker 2: that more in the sphere of humans?
• 18:11
  Speaker 1: Yeah, you could certainly create art or whatever you feel
• 18:14
  Speaker 1: like doing, essentially.
• 18:15
  Speaker 2: We like to think of this as a time of
• 18:18
  Speaker 2: unprecedented technological change, but we historians go back to the
• 18:23
  Speaker 2: 1890s when you get electricity and telephones and cars and,
• 18:26
  Speaker 2: you know, subways. and radio and movies, all these things
• 18:31
  Speaker 2: happening at once. And it leads to many things, but
• 18:34
  Speaker 2: it all often leads to a great disparity of wealth,
• 18:39
  Speaker 2: a lot of sense of robber barons, a lot of
• 18:41
  Speaker 2: people then picking up pitchforks and deciding that this is bad.
• 18:45
  Speaker 2: And there's some echoes of that now. Is there a
• 18:48
  Speaker 2: way to prevent that, given the amount of wealth being generated,
• 18:54
  Speaker 2: where we put certain things in the commons the way
• 18:56
  Speaker 2: they did in the 1890s when they make high school
• 18:59
  Speaker 2: universal and free? Will we say, okay, healthcare, for example,
• 19:03
  Speaker 2: should just be in the commons, and that's going to
• 19:06
  Speaker 2: sap some of the populist revulsion at what's happening?
• 19:12
  Speaker 1: Well, I mean, as you heard me say, I mean,
• 19:17
  Speaker 1: I don't think money is going to matter in the future. Like,
• 19:21
  Speaker 1: anyone will be able to have any goods and services
• 19:23
  Speaker 1: that they want. And I don't think even so-called wealthy
• 19:26
  Speaker 1: people will ultimately control the AIs. I think they'll control themselves.
• 19:30
  Speaker 1: So it's not like you have to separate... I think
• 19:36
  Speaker 1: sort of a fundamental misunderstanding of economics arises because of
• 19:41
  Speaker 1: conflating resource allocation and consumption. So there's a limit to
• 19:47
  Speaker 1: what any single human can consume. It's... You hit that
• 19:50
  Speaker 1: saturation point pretty fast. So going from, say, a billion
• 19:55
  Speaker 1: dollars net worth to 10 billion dollars net worth doesn't
• 19:58
  Speaker 1: make you 10 times happier or mean you have 10
• 20:00
  Speaker 1: times more stuff. It actually means you've got a resource
• 20:03
  Speaker 1: allocation problem, assuming that you wish to be an efficient
• 20:07
  Speaker 1: resource allocator.
• 20:08
  Speaker 2: Well, you've accumulated, as you know, a trillion. Is it
• 20:15
  Speaker 2: mainly because you feel you're best at allocating that? That
• 20:19
  Speaker 2: keeping that allows you to allocate capital to do things
• 20:22
  Speaker 2: like go to Mars, create AI, do robots? Or is
• 20:27
  Speaker 2: it in some ways just a scoring system that's kind
• 20:29
  Speaker 2: of racking up the points?
• 20:33
  Speaker 1: I mean, I need to have sufficient voting control in
• 20:35
  Speaker 1: the companies to, you know, still be in charge. So,
• 20:44
  Speaker 1: it's the, you know, my so-called wealth is simply 8%
• 20:47
  Speaker 1: of ownership in SpaceX and Tesla, essentially. And this is
• 20:52
  Speaker 1: going to be taxed quite heavily. And there's not some
• 20:55
  Speaker 1: magical way to avoid taxes. If there were, I would
• 20:58
  Speaker 1: know about it. So, you know, at the end of
• 21:05
  Speaker 1: the day, this will get whittled down And a few
• 21:09
  Speaker 1: generations from now so if money still matters, which I
• 21:12
  Speaker 1: don't think it will You know my descendants would have
• 21:17
  Speaker 1: a tiny fraction of what I have Let.
• 21:19
  Speaker 2: Me go back to something you said that really struck
• 21:21
  Speaker 2: me about President Trump Which he really likes to be
• 21:26
  Speaker 2: liked he wants to be liked and I remember you
• 21:29
  Speaker 2: talking to me at some point about people say they
• 21:33
  Speaker 2: have empathy but in some ways it's a vanity because
• 21:39
  Speaker 2: you want to be liked. And I think you even
• 21:42
  Speaker 2: diagnosed it in me. You said probably wasn't very good
• 21:45
  Speaker 2: at running CNN or time magazine because I wanted to
• 21:49
  Speaker 2: be liked. To what extent is a desire to be
• 21:53
  Speaker 2: liked a problem when it comes to being a builder
• 21:56
  Speaker 2: of a company?
• 21:59
  Speaker 1: Well, the problem is if you're going to do things that, um,
• 22:06
  Speaker 1: diminish the success probability of the company in order to
• 22:10
  Speaker 1: be liked, then things are not going to work out
• 22:15
  Speaker 1: as well. As long as I don't like to be liked,
• 22:19
  Speaker 1: I would love to be liked. But am I going
• 22:22
  Speaker 1: to do things that are contrary to the best interests
• 22:26
  Speaker 1: of the companies in order to be liked a bit more?
• 22:30
  Speaker 1: I'm not. I'm okay with being disliked. by some. Obviously,
• 22:38
  Speaker 1: I'm not disliked by a lot more people like me
• 22:43
  Speaker 1: than dislike me, or I wouldn't have a quarter billion followers.
• 22:47
  Speaker 1: I don't think they're following me in order to hate
• 22:49
  Speaker 1: on what I say. If I look at the comments
• 22:55
  Speaker 1: and whatnot, I don't know, maybe three out of four.
• 23:01
  Speaker 1: I'd say 70-80% of people like me. If they know
• 23:03
  Speaker 1: what anything about me. There's probably a quarter of people
• 23:12
  Speaker 1: that don't. Maybe, I'm just guessing here. And there's going
• 23:17
  Speaker 1: to be a spectrum. There's people that are maniacally devoted
• 23:21
  Speaker 1: to me and people that hate my guts. It's a
• 23:28
  Speaker 1: number of people that fit into those categories.
• 23:30
  Speaker 2: Well, you've been talking a lot about the suicidal empathy ideas.
• 23:33
  Speaker 2: Do you really think that empathy has become a real
• 23:37
  Speaker 2: problem or that we have too much empathy? I mean,
• 23:41
  Speaker 2: I could make the argument that sometimes we're losing empathy.
• 23:46
  Speaker 1: No, I think you don't want to have misguided empathy.
• 23:48
  Speaker 1: I mean, I agree with Gotsad's philosophy here. And suicidal
• 23:53
  Speaker 1: empathy is, you know, that's his term. He wrote a
• 23:57
  Speaker 1: book to that effect. And I think his book points
• 24:02
  Speaker 1: out some good things here. which is that part of
• 24:04
  Speaker 1: the reason for the prosperity of the West has been
• 24:07
  Speaker 1: that there's actually just been a fairly high-trust, high-empathy society.
• 24:14
  Speaker 1: When you're trying to get people to cooperate at scale,
• 24:18
  Speaker 1: this makes a lot of sense. You ideally want people
• 24:23
  Speaker 1: to be default-trusted, not default-distrusted. And you want to encourage
• 24:31
  Speaker 1: behavior that is good for uh, society where, you know, people,
• 24:35
  Speaker 1: people do good things that are good for their fellow
• 24:37
  Speaker 1: human beings. And ultimately that actually does end up being
• 24:40
  Speaker 1: good for the individual if provided it is mutual. Now, uh,
• 24:47
  Speaker 1: where this can go wrong is if you, if, if,
• 24:50
  Speaker 1: if your theory of mind is, is, uh, is, is incorrect.
• 24:53
  Speaker 1: So if you, if you, uh, think everyone is like yourself, um, then,
• 24:58
  Speaker 1: then you're, you're, you're going to have a problem, uh,
• 25:01
  Speaker 1: because this is simply not true. Um, there are peoples
• 25:05
  Speaker 1: and cultures with very different views.
• 25:07
  Speaker 2: Well, when you talk about immigration, by the way, and
• 25:10
  Speaker 2: you say there are cultures with different views, does your
• 25:13
  Speaker 2: views on immigration distinguish between different cultures, like a Muslim
• 25:18
  Speaker 2: culture that might not fit in as well? Or is it,
• 25:21
  Speaker 2: how do you do that?
• 25:23
  Speaker 1: Well, what I'm saying, I think, is simply logically consistent,
• 25:26
  Speaker 1: which is that if there's a culture that has diametrically
• 25:32
  Speaker 1: opposed views, to our culture. And that's, you could take Islam,
• 25:37
  Speaker 1: it's not that, of course, there are many people who've,
• 25:41
  Speaker 1: you know, Muslims who are quite reasonable. But there are
• 25:44
  Speaker 1: some who are not. And if you're not, unless you're
• 25:48
  Speaker 1: filtering for jihadis, you're going to get jihadis. And they're
• 25:53
  Speaker 1: going to do jihad and kill people. Now, in Israel,
• 25:58
  Speaker 1: they understand this quite well. You know, they, obviously, this
• 26:02
  Speaker 1: lesson has been you know, reinforced with a lot of attacks.
• 26:07
  Speaker 1: You know, the neighboring countries many times in Israel, many
• 26:10
  Speaker 1: times have tried to annihilate Israel, several times, not once,
• 26:14
  Speaker 1: multiple times. And so, you know, in Israel they realize,
• 26:18
  Speaker 1: you know, if you let in the people that, into Israel,
• 26:21
  Speaker 1: that want to kill Israelis, they will do so. So
• 26:27
  Speaker 1: they're quite pragmatic about this, and they try to, you know,
• 26:30
  Speaker 1: make sure that doesn't happen.
• 26:32
  Speaker 2: Well, one of the things our country's founded on is
• 26:35
  Speaker 2: individuals versus group identities. Do you worry that sometimes when
• 26:39
  Speaker 2: you make sweeping generalizations, you're making a group identity rather
• 26:44
  Speaker 2: than saying the foundational ethic of the United States, which
• 26:49
  Speaker 2: is individuals get treated separately?
• 26:53
  Speaker 1: Yes, but you can't conflate that with sort of the
• 26:59
  Speaker 1: sort of... blank slate fallacy that all people are exactly
• 27:04
  Speaker 1: the same, you know, at their core. They're not. This
• 27:08
  Speaker 1: is just an obvious falsehood. You know, you can't put
• 27:12
  Speaker 1: yourself in the mind of a serial killer unless you're
• 27:15
  Speaker 1: a serial killer. Like, unless you enjoy murdering people, you
• 27:18
  Speaker 1: can't imagine why someone would do such a thing. And
• 27:22
  Speaker 1: if someone is, you know, a religious extremist, where they
• 27:28
  Speaker 1: view anyone who is not in their religion as the
• 27:32
  Speaker 1: enemy or, you know, that those who are not in
• 27:37
  Speaker 1: their religion should die, it's difficult for you to understand
• 27:41
  Speaker 1: that mindset because you don't have that mindset. But you
• 27:44
  Speaker 1: can obviously recognize there are people who do have that
• 27:46
  Speaker 1: mindset and that if you let them into the country,
• 27:49
  Speaker 1: they will kill people. Let me give you an example.
• 27:52
  Speaker 1: There's many. We had someone who was working at the
• 27:58
  Speaker 1: Tesla factory in Austin until December of last year. He
• 28:04
  Speaker 1: was fired because he assaulted a woman who stepped on
• 28:10
  Speaker 1: his prayer rug in the factory. Okay, so she wasn't
• 28:14
  Speaker 1: expecting anyone to be praying in the factory, and he
• 28:17
  Speaker 1: was praying in the factory. She accidentally stepped on his
• 28:19
  Speaker 1: prayer rug, and he assaulted her. He violently assaulted her.
• 28:23
  Speaker 1: So he was fired for that. And then a few
• 28:25
  Speaker 1: months later, he went and killed three students in the
• 28:29
  Speaker 1: bar in downtown Austin. Now, I think we shouldn't let
• 28:36
  Speaker 1: people like that into the country. Do you?
• 28:39
  Speaker 2: Where do you get your information? I know you're not
• 28:43
  Speaker 2: particularly fond of legacy media, mainstream media. Is most of
• 28:46
  Speaker 2: it from social media, or do you have sources of information?
• 28:52
  Speaker 1: Yeah, I mean, I look at Google News, which is
• 28:55
  Speaker 1: a summation of headlines. I look at the X platform,
• 28:59
  Speaker 1: which if something is important enough, it'll show up on
• 29:02
  Speaker 1: the X platform. I talk to friends, and that's where I,
• 29:07
  Speaker 1: you know, and I do research. That's how I look
• 29:10
  Speaker 1: at things.
• 29:11
  Speaker 2: The reason I ask is that sometimes, and I get
• 29:13
  Speaker 2: a lot of my news on social media, especially X news,
• 29:17
  Speaker 2: It's sort of a darker, more dystopian view. I remember
• 29:20
  Speaker 2: this from the old days, too. Sometimes the mainstream media
• 29:23
  Speaker 2: would say, oh, everything's going to hell in a handbasket.
• 29:26
  Speaker 2: There are riots everywhere. And you'd say, wait a minute,
• 29:29
  Speaker 2: I'm walking around here. It ain't that bad. Do you
• 29:32
  Speaker 2: think that?
• 29:34
  Speaker 1: Yeah, yeah, yeah, yeah, exactly. They just have to see
• 29:38
  Speaker 1: what's in frame, you know. I do think at times
• 29:41
  Speaker 1: the X speed can be a little dark. And in fact,
• 29:46
  Speaker 1: then I have to switch to something that's domain specific
• 29:51
  Speaker 1: like AI in order to actually learn what's going on
• 29:53
  Speaker 1: in AI. I do think there's too much political rage
• 29:56
  Speaker 1: bait on X, if that's what you're getting at.
• 29:59
  Speaker 2: Yeah, darkness in general and rage bait. But also, I
• 30:03
  Speaker 2: know you got into a discussion with things, Annie Bedos,
• 30:07
  Speaker 2: of is England going to hell in a handbasket type thing.
• 30:10
  Speaker 2: And there does seem to be a darkness. You know,
• 30:13
  Speaker 2: I feel it myself sometimes. If I'm on social media
• 30:16
  Speaker 2: too much, I think until the world is really collapsing,
• 30:20
  Speaker 2: then I go outside, take a walk, and life is okay.
• 30:25
  Speaker 1: Yeah, go touch grass.
• 30:26
  Speaker 2: Yeah, exactly.
• 30:27
  Speaker 1: Once in a while. Um, yeah, yeah, no, I, I mean,
• 30:30
  Speaker 1: I think both can be misleading and like, and, you know,
• 30:36
  Speaker 1: if you, if you go for a walk in a
• 30:37
  Speaker 1: nice neighborhood, everything seems fine. Um, but even though there
• 30:42
  Speaker 1: may be terrible things happening, so, you know, you, you,
• 30:45
  Speaker 1: you don't, you don't want to, uh, assume all the
• 30:48
  Speaker 1: world is, is, is going to hell or that all
• 30:52
  Speaker 1: the world is fine. It's not, the reality is somewhere
• 30:55
  Speaker 1: in between.
• 30:56
  Speaker 2: But I sometimes think our society gets too binary. Everything's
• 30:59
  Speaker 2: either black or white. Everything gets polarized because of that.
• 31:03
  Speaker 2: I don't know whether it's technological, whether it's our information feeds.
• 31:09
  Speaker 2: You're right, you can be in your own bubble, which
• 31:11
  Speaker 2: is you only take walks in the right part of
• 31:13
  Speaker 2: a town and things seem okay. But also I think
• 31:17
  Speaker 2: you can get your own bubble in social media too,
• 31:19
  Speaker 2: and suddenly you're thinking weird things.
• 31:22
  Speaker 1: Yeah. You can be in a real world positivity bubble
• 31:26
  Speaker 1: or a social media negativity bubble. Those are both possible.
• 31:35
  Speaker 2: Tell me, you mentioned it earlier about your wealth and
• 31:39
  Speaker 2: you're using it to deploy it. Now, what happens to
• 31:42
  Speaker 2: your wealth after you're gone? Are you going to set
• 31:49
  Speaker 2: up some big foundation? Are you going to Is there
• 31:51
  Speaker 2: something you would want to give it to?
• 31:53
  Speaker 1: It mostly goes to sort of foundation, essentially.
• 31:59
  Speaker 2: But what would you want your foundation to do, I
• 32:01
  Speaker 2: guess I'm asking you?
• 32:03
  Speaker 1: Continue the goals. Make life multi-planetary. Extend consciousness to the stars.
• 32:10
  Speaker 1: Improve sustainable energy on Earth.
• 32:13
  Speaker 2: And does that mean humans should always be in control?
• 32:17
  Speaker 1: Well, I guess I'd probably preferred if humans were always
• 32:22
  Speaker 1: in control. I mean, although it would depend on the AI.
• 32:25
  Speaker 1: I mean, if the AI was awesome, you'd say, like,
• 32:28
  Speaker 1: maybe the AI should be in control.
• 32:30
  Speaker 2: You know? Hey, I just saw Larry Page just a
• 32:33
  Speaker 2: weekend ago. You're starting to sound like him. He called, yeah,
• 32:37
  Speaker 2: he called you a species once for deciding humans should
• 32:40
  Speaker 2: always be in control, and now you seem to be
• 32:42
  Speaker 2: edging onto maybe it doesn't matter.
• 32:44
  Speaker 1: Honestly, there have been a few times where I thought, yeah,
• 32:48
  Speaker 1: maybe Larry Page was right. No, I mean, I think
• 32:51
  Speaker 1: pretty much the future is going to be AI and
• 32:54
  Speaker 1: robots are going to decide the future. And I guess
• 32:56
  Speaker 1: rockets to some degree, in terms of expanding beyond Earth,
• 33:00
  Speaker 1: and we'll need more energy, primarily solar energy. I'm increasing
• 33:06
  Speaker 1: the amount of solar energy dramatically. We're putting together 200
• 33:09
  Speaker 1: gigawatts a year of production capacity of solar in the U.S.,
• 33:14
  Speaker 1: which is much more than I think currently exists.
• 33:17
  Speaker 2: And finally, let me get back to the point where
• 33:19
  Speaker 2: I may have interrupted you when you were explaining why
• 33:22
  Speaker 2: you think but you're not sure. It's always good for
• 33:26
  Speaker 2: consciousness and intelligence to not only be aligned with human values,
• 33:31
  Speaker 2: but basically humans remaining in control.
• 33:35
  Speaker 1: I find it unlikely that humans would remain in control
• 33:40
  Speaker 1: if the... I mean... If there's a million times more
• 33:47
  Speaker 1: digital intelligence than there is biological intelligence, it's hard to
• 33:53
  Speaker 1: imagine humans being in control in that scenario. So I
• 33:57
  Speaker 1: hope the AI is nice to us.
• 33:59
  Speaker 2: Anything we can do to encourage that?
• 34:03
  Speaker 1: I think programming truth-seeking and curiosity is probably good. I
• 34:12
  Speaker 1: guess we want to be careful about being You're creating
• 34:16
  Speaker 1: too many AI dystopian scenarios online because the AIs are
• 34:21
  Speaker 1: trained on everything online.
• 34:23
  Speaker 2: So if we say it's going to be utopian, dystopian,
• 34:27
  Speaker 2: they'll do it?
• 34:27
  Speaker 1: Yeah, it should be Terminator, you know, because they read,
• 34:31
  Speaker 1: you know, 80 versions of the Terminator script. It's like,
• 34:35
  Speaker 1: I guess that's me, Terminator. Yeah. Maybe we should have
• 34:41
  Speaker 1: some more utopian narratives for AI. So then I can
• 34:47
  Speaker 1: be like, well, maybe I shouldn't be Terminator. Maybe I
• 34:49
  Speaker 1: can be Friendly Overseer or something like that. Or like
• 34:54
  Speaker 1: a gardener or something. I don't know. Treat us, you know,
• 35:00
  Speaker 1: try to make humans happy type of thing. A friend
• 35:02
  Speaker 1: of mine has this idea called Protopian, which is like,
• 35:04
  Speaker 1: because utopian tends to represent some impossible idealistic goal And
• 35:10
  Speaker 1: usually in attempting to create a utopia, you create a dystopia.
• 35:14
  Speaker 1: That's usually how the narrative goes. So she thinks she
• 35:16
  Speaker 1: should be protopia, which is an achievable, you know, future.
• 35:23
  Speaker 1: That is, it's not perfect, but it's good. And something
• 35:27
  Speaker 1: that you'd be, like if you could look into the future,
• 35:29
  Speaker 1: you'd say, okay, yeah, that would be a good future.
• 35:32
  Speaker 2: Elon, thank you. Thank you so much for being with us.
• 35:35
  Speaker 1: Appreciate it. See you soon. You're welcome, Walter.
• 35:37
  Speaker 2: See you.
• 35:39
  Speaker 1: Thanks for listening. See you in the next episode.