Jack Clark (3:02)

Yeah, there's a certain set, particularly within language tasks. And maybe we can get over to coding, which is sort of a whole different beast. But within language tasks, it seems like there is a limit still to how good it can be. And so if, for example, even if you're trying to either pull out the most interesting segment from a podcast or caption it and make clips that are postable, you know, those clips need to be perfect. And like a human still, you know, you're going to trust a person more to do that. Picking what's actually interesting might be kind.

Lex Fridman (3:28)

Of hard, but even there. So offline, we were having a conversation about how to write a tweet for something for your podcast. Right. And then we were discussing, oh well, one strategy might be like write it for a group chat. Maybe you can add that to like you would think this is exactly what LLMs are for, right? It's like write the tweet. This is like language in, language out. Here's a system prompt. But like, why is it, I assume you aren't getting that much use out of just like having the AI write your tweets for you. And why is that? Not because, by the way, I'm focusing on this. Not because writing tweets is like the most important thing in the economy. This is just like, this is the first thing they should be able to do, right? Why are we not delegating this to them yet? If you posted something you like, notice it doesn't do well. You have this, you can Think about what went wrong. You also have this experience from what do your users want or what do your followers want that these AIs can't pick up?

Jack Clark (4:20)

Well, it's actually a reasonably high. So think about you posting something on Twitter or substack or whatever. That bar for you is actually going to be pretty high. There's a lower bar of thing where language models are really useful, like customer support tickets, for example. It works really well there because you don't need the language to be perfect. You don't need insane nuance. And it's actually okay for a certain class of thing if it's right 97% of the time and if it knows how to, you know, say, hey, I'm wrong here. So I do think that what we're describing here still does drive a lot of GDP growth in certain areas at least, because, you know, you don't need the high bar everywhere.

Lex Fridman (4:57)

Yeah. I mean, I'd be curious to see what the actual numbers are on customer service, employment. I think from what I understand, they're not like, down that much.

Jack Clark (5:05)

Yeah.

Lex Fridman (5:05)

So it's interesting. Even in those areas, you're not seeing these transformative impacts.

Jack Clark (5:09)

Yeah. I mean, the venture view against that would be that it's like the beginning of this massively exponential curve. And these things are like in the first half of the first inning, but all the curves are like this. And so three years from now, it'll show up.

Lex Fridman (5:21)

I agree with that. Maybe not just in three years, but I agree with that over the course of a decade. But that's just because I think this is such a big bottleneck to getting these models to be valuable that it has to be like, people will want to solve it. Right now, OpenAI or Anthropics, revenue is on the order of 10 billion ARR. And that's a lot, obviously. But like, McDonald's and Kohl's make more money and those companies aren't AGI. Right. So, like, if you have real AGI, like trillions of dollars a year, that's what like, humans around the world earn in wages.

Jack Clark (5:51)

Yeah.

Lex Fridman (5:52)

Given that that is sort of the addressable market.

Jack Clark (5:54)

Yeah.

Lex Fridman (5:54)

And given that this is one of the key bottlenecks to getting there, I just think a ton of effort we put into this problem. We've had a lot of progress in the past. So I'm not, I'm not like one of these people who's like, AGI is not coming.

Jack Clark (6:03)

Yeah, of course.

Lex Fridman (6:04)

I'm just saying, like this. This definitely needs to be solved before.

Jack Clark (6:06)

We get there, I guess maybe then to the broader thing about, like, being, you know, AGI pilled in general. Like, you obviously have spent a lot of time talking to researchers. You're very close to that. Do you still feel as confident as ever in that, even if your timelines are different than, you know, maybe what you've gotten from some of your guests?

Lex Fridman (6:21)

There is one thing really interesting to observe, which is that, like, they have cracked reasoning. It just so happens that reasoning ended up being much easier than something we take for granted, which is just that, day in, day out, you're going to be picking up information in your workplace. But you go back to Aristotle and his big take was like, look, the thing that makes humans special from other animals is that we can reason and the other animals can't. And it's sort of funny that these models just aren't that useful yet. They can't do almost anything, except for the one thing they can do is reason. Given the fact that these sort of ambiguous capabilities do come online, that makes me think that continual learning might also be something that in the 10 years. It's also useful to remember deep learning is not that old. It's 13, 14 years old, or at least, sorry, Alexnet. Is that old when we started training these models? Yeah. It's very possible to me that in a decade or two we find a solution to this problem.

Jack Clark (7:17)

It's very interesting to hear you continue using language that's like, these models aren't that useful yet. Which part of me very much agrees with. And part of me, it's like putting language to an experience that I have when I use them myself. The other side of it would be they are an amazing, amazing replacement for search. In a lot of cases, they do things like code generation in a very effective way. It seems like doctors are able to use them to handle a lot of scribe. So there are these things that are, I would say, clearly working. Do you see it that way too, where you're like, some things are highly useful?

Lex Fridman (7:51)

Yeah, no, 100%. I'm putting it more in the context of the real potential for AGI, just like a genuine replacement for human labor. I expect that to cause a 10x increase in the level of growth. And so if it's on the scale of the Internet, I'm like, oh, wow, this is so much shorter of what AGI could be. So where this is clearly not that useful yet, and a sort of more tangible reason to expect this kind of change, but one is just that the amount of labor supply just dramatically increases So I think people often, especially in tech, focus on how it will make a specific industry more productive. This narrow productivity improvements, where it's just like, no, imagine a trillion people in the world who are each specializing and each discovering new knowledge, or we just get all these gains from comparative advantage as a result. But another is that because these minds are digital, they have advantages even if they're the same amount of intelligence. Specifically advantages in collaboration that humans just can't have because of the way our minds work. One example of this is, okay, suppose this problem is solved where you can actually learn on the job. Now a human can on the job, learn from their work, and then so over the course of 20 years, they become master of their craft, they're incredibly valuable. You're one such person, right? You've picked up all this context in the tech industry from running companies, from investing in companies. If we get models that have this human like capability, not only could they learn on a single job, copies of the model are deployed all through the economy. They can amalgamate the learnings from basically doing every single job in the economy at the same time. And at that point, even if you don't have further algorithmic innovations, you would still have something that's functionally becoming a superintelligence. You have this broadly deployed intelligence explosion. They're just one of the many ways in which the fact that they are.

Jack Clark (9:39)

Digital just gives them emerging intelligence ants or something.

Lex Fridman (9:42)

Yeah, exactly.

Jack Clark (9:43)

You're kind of describing a world where the way that the impact happens is through a trillion new white collar workers. There's another version of it where it never actually does exactly that, but does this other thing, which is just a higher level of intelligence than anything we've ever encountered. And it creates new paths and identifies new ways to do things that humans could still then do sometimes. The way I think about this is the 400 IQ AI, even if it can't do all the things that a person can do, it could help or fully identify new drugs, help us get to space more effectively, all sorts of innovations like that.

Lex Fridman (10:21)

I think a good way to think about this is maybe China. So, okay, why has China been so successful in not only catching up in science and technology, but in fact in many ways surpassing America in a lot of key domains. And obviously China is full of lots of brilliant people. And so that does lend credence to your argument that, look, they have to have. I think the intelligence is a big part of it. But I think more fundamentally, just once you've hit a benchmark of intelligence the scale is what makes China so successful. Right. You have just within manufacturing, there's 100 million people who have built up, who are working in manufacturing in China, who have built up all this process knowledge in whatever subdomain is relevant to whatever is being built. So that scale I think is like if China is graduating, I think like tens of millions of STEM graduates every single year. It's not that any one of them is super brilliant, it's just that each of them can specialize in whatever radar technology that BYD needs or whatever production technology is needed.

Jack Clark (11:23)

I'm thinking about this out loud, but it opens up the question of would more impact happen from a trillion more super connected, super collaborative human level intelligence people, or from one just like demigod level intelligence who could like figure stuff out and tell us all what to do?

Lex Fridman (11:41)

Do we have evidence in Silicon Valley history that it's more of the latter? It seems to me that, well, it depends.

Jack Clark (11:46)

You know, there's the great man theory thing where it's like there are these special people who, you know, that's how the big leaps happen. Whether it's, you know, like a Steve Jobs or Elon or whatever, where like there is some just outlier person who directs the resources and that one person can pull greatness out.

Lex Fridman (12:01)

Yes. I haven't seen them up close as you have, but it seems to me, I agree that they've had a huge impact. People like Elon and Steve Jobs, it seems to me their impact has been more so a product of just like, you will do this, otherwise I will throw a tantrum, which is good, right? You should throw a tantrum and I'm going to sleep in the office for years on end and just get people in the right place at the right time. But it's less so. Only Elon can come up with how the fins on the SpaceX rocket should be designed. And therefore his uber intelligence allows him to design across five different hardware verticals.

Jack Clark (12:33)

It's probably, it's definitely something that's more to do with the leading of people and the clarity of vision, I would say, than the probably like engineering genius or something like that.

Lex Fridman (12:43)

Yes, but I mean, this is actually another way to illustrate why digital minds are such an upgrade. Elon has obviously been super successful across so many different areas of technology. But of course he's just like one person. To the extent that you think there's something unique about him or about the small teams he had assembled, like early SpaceX, early Tesla. If they were digital, you could just replicate them early SpaceX team, replicate them 1000 times, throw them at a thousand different harder verticals and see what happens. Because you can't scale that with humans.

Jack Clark (13:14)

The other thing I'm thinking about now as we're talking is could a digital intelligence be like a leader of 10,000 people? And I do think that coordinating large groups seems to be incredibly important to getting big societal things done. So now I'm wondering, could you have a digital mind at the top or does it need to be a human?

Lex Fridman (13:32)

I think you'd have a much easier time if it was digital. This is another one of the key advantages being digital. Because right now Elon has the same 10 to the 15 flops in his head that every single other human has. Now this could be negative as well, right? Like Xi Jinping also only has 10 to the 15 flops in his head that every other human has. The ratio of compute or deliberation happening at the top versus through the hierarchy is just so lopsided. And that requires a lot of delegation. Now that's good in a sort of like free society sense. You don't want the president to have that much control over your life. But if you. Within these specific domains where we want, companies have this outer loop where if they fail, they can just go down, unlike a country. So there it might make sense to have more of a. Have the company be the product of a single coherent vision. This is a founder mode idea. But obviously that's limited by the fact that if a company goes to a certain size, it's just hard for a single person to monitor everything. If it was an AI, and especially if you could inference scale Mega Elon, the thing that's running on a huge data center that's dedicated to just him. Mega Elon can read every pull request every comms, input output to the company. He can micromanage every single employee down to the technician at the dealership. Especially, I don't know, Elon of 10 years ago, it just would be like incredible.

Jack Clark (14:49)

So I guess what I'm wondering is that the AI COO under the human who is giving them superpowers, or is that person, is that AGI rather actually leading?

Lex Fridman (14:57)

I think in the near term it will be like the latter. But over the long term, obviously like.

Jack Clark (15:03)

If AI, you think it'll be AI CEOs basically in the long term.

Lex Fridman (15:06)

And so it does seem like the thing they're lacking at least now is like this sort of more taste oriented stuff, right? They can do the research for you. O3 can go do the research for you. But then you're not going to let it make the investment decision for you. That just comes down to your own unique insight into the field or your unique taste. And maybe running a company will work like this for a while where they can just curate a lot of information for you and then you still have to make the final call. It's just like the Steve Jobs model of he makes the final call, the designers come to him with a bunch of different ideas, eventually that will also go, but maybe in the near term that's what it looks like.

Jack Clark (15:41)

So maybe back on AGI, all the time that you spent with researchers and companies and you really inspecting this and you're very truth seeking, you still do believe that AGI is going to just like wake up like you. You think it's all gonna happen still? Or has anything as you've learned over the last year or so changed your mind and you're like, well, it's a really cool idea, but it's not for sure it's gonna happen to you.

Lex Fridman (16:01)

Yeah, I think the biggest consideration there is the progress for the last 10 years in AI. I mean more than 10 years actually. But even going back to the 60s, the progress in AI period has been driven by increases in compute. And especially in the deep learning era, it's been stupendous increases in compute dedicated to frontier systems. If you just look at public announcements of how big the compute runs are, it looks like the trend Since I think 2012, 2016 has been 4x per year. So over four years is 160x the biggest system trained from the one before in terms of the compute used that physically cannot continue past this decade from how much energy is used to what fraction of advanced chips at TM Cincy you need to procure to even raw fraction of gdp. So then you would just need progress from other. The progress would just have to come from algorithmic progress or I mean it would just have to be that. And because this key input into AI progress would stop after the next five years. There is this dynamic that the yearly probability of AGI is quite high now. Not that in the sense that it will happen, but there's a decent chance every year until 2030. And then it just sort of craters because then you're just like, okay, we just had to think hard about what's missing. We can't just throw more compute at the problem.

Jack Clark (17:16)

Yep, yep, that makes sense. Do you feel right now like AI is making us smarter or is it like increasing brain rot? And do you think like over Time that that goes in a particular direction.

Lex Fridman (17:27)

Did you see the meter uplift study from the other day? Oh, it was super interesting. Meter is this organization that does evals on basically how AI is progressing. They had a very interesting result the other day. So they had open source developers who are working in repositories that have tens of thousands of stars. They did a randomized control trial with these people where they would issue them a random pull request that was open in these repositories and they'd work on it in one case just by themselves. In another case with the help of like cursor and Claude 7 or Claude 3.7. And then they measured in the case where they were working with the AI, one, how much do you think you were sped up? And then two, how much were they actually sped up? And the developers thought that they were 20%. They would be like 20. They were 20% more productive as a result of AI. They were actually 19% less productive as a result of AI.

Jack Clark (18:23)

Wow.

Lex Fridman (18:24)

It was really interesting to read the threads of the developers who participated trying to explain. So people who have looked at this who are even more bullish on AI are like, they think their experimental design was extremely robust.

Jack Clark (18:33)

Even like senior engineers misread themselves, especially as senior engineers. Wow.

Lex Fridman (18:37)

Senior engineers had, from what I remember, the biggest decreases in productivity. So these are people who are experiencing this rotor. They've been working on it for decades. And there's a couple explanations. One is just that I think in a lot of domains there is this common failure mode in intellectual work where you default to procrastinating by doing a thing which seems productive but is not moving the ball forward that much. So the classic example of this is in college, instead of rereading the textbook, you should just do the practice problems and maybe using these AI tools and then going on social media for 30 minutes while you're waiting for the completion to complete is another example of this. Why did this happen? So at least for now, to go back to your original question, it's not obvious to me that it's making us smarter.

Jack Clark (19:20)

It's interesting. The version for everybody is, I think a large number of people are having ChatGPT, sort of like give them guidance in life at this point. I don't mean that in some huge philosophical way, although maybe in some cases, but even just like day to day people, like, here's my whole setup. This is everything about me, like what should I do? And that is a big influence. And so it's sort of we kind of, to the Extent that these models have quietly gripped a lot of people, either through their decisions, through their engineering work, through whatever else. It's like pretty important that these get really good.

Lex Fridman (19:57)

But have you found it useful? That kind of stuff just like. I mean, it is sort of like, okay, plan out a cute date for me or something.

Jack Clark (20:02)

I use it some. I think I use it less than a lot of my friends. I think more. I'm commenting that I think broadly. A lot of people do use it for that. Even if you or I happen to be in the group that uses it a little bit less. I think a lot do.

Lex Fridman (20:16)

Right. Yeah.

Jack Clark (20:17)

You know, and even probably taking personal advice and things like that.

Lex Fridman (20:20)

Yeah.

Jack Clark (20:21)

Which is probably good in a lot of cases. Like, I think it is very smart and people know how to use it and things like that. But it just speaks to. Even with engineering and then also on, you know, people using it in sort of like search and question asking for personal use. That's got a lot of influence.

Lex Fridman (20:35)

It feels like it's made me smarter. Especially when I interview people in domains where there's not a lot of just written down, like biology. This is a classic example. I was preparing to interview George Church, who's this famous pioneer in synthetic biology, and there most of my prep time was just dominated. Talking to these models and then just telling them, teach this to me as if you're a Socratic tutor. Don't move on in the explanation until you're satisfied that I have completely understood. I guess I'd be curious. Yeah. If people in other domains have.

Jack Clark (21:04)

Do you feel actually on biology in particular, have you spent time with Patrick at ARC by chance?

Lex Fridman (21:08)

A little bit, yeah.

Jack Clark (21:09)

I feel like the bullishness that a lot of people in biology have for AI's ability to do drug discovery and things like that seems very promising to me. I don't know. What you found is you've spent time in biology.

Lex Fridman (21:19)

One interesting question I had for these people is, in biology we can either employ models which think in thought space. So just like humans, they can come up with hypotheses and so forth, or models which think in protein space or DNA space or capsid space. You know, like humans just aren't. We can't. Like, I think G sounds really good here. Then let's do T next. And so I'm curious which one they think is a more promising or more useful complement to the current progress in biology. Is it having better models, I can think in the alphafold type stuff, or is it just have O3 come up with Hypotheses and just write them out in English. At least George Sturg seemed to think it was the biospace, thinking in proteins or DNA or so forth. Because this is while we have millions of life science PhDs who can come up with ideas, being able to prune through them in simulation.

Jack Clark (22:10)

Yeah, exactly. Like a digital cell kind of thing.

Lex Fridman (22:12)

Yeah, exactly. Yeah. Is like the more useful compliment there.

Jack Clark (22:15)

Yeah, I mean, that seems like that would be like a very unequivocally positive output for humans if we can sort of just wildly change biology and pharmacology and things like that.

Lex Fridman (22:25)

Yeah. I think in the long run, I sort of worry about the fact that we know ways in which things can just go horribly wrong. We have the equivalent of nuclear weapons, but in different domains. So mirror life and biology. Apparently, if you have life with the opposite chirality, there's just no defense. Plausibly, it could render many life forms unviable on Earth. And so George Church was one of these people who wrote this letter saying, look, this thing exists. Let's not work on it. But I don't know, over the course of 100 years, what's the equilibrium here in physics? I interviewed this physicist. One of the things he's worked on is thinking about this thing called vacuum decay. And the TLDR is basically like, it might be plausible to just literally destroy the universe.

Jack Clark (23:14)

Well, what's the idea there?

Lex Fridman (23:16)

Look, I'm a podcaster, so you're asking a content creator about physics, and so I'll do my best. Apparently, in quantum field theory, we're in a sort of what's called a metastable state, where if it's sort of having a huge valley and then a little bit of a hill, and then we're in this little rump here, it's possible to throw such a huge amount of energy, and what would happen is that this bubble would expand at the speed of light, which would just be like total destruction. It sounds like some wild sci fi thing.

Jack Clark (23:47)

No, but it actually takes me to the next thing I wanted to ask you about, which is you have this wide range of guests and interests, so maybe first, outside of AI, what domains are you most interested in right now? Because I feel like I've seen you talk about, you know, politics and Russia and math and science and longevity. Like, what are you interested in most?

Lex Fridman (24:05)

Right now, I'm interested in what the year 2050 looks like. Obviously, you need to understand AI in order to understand what happens in 2050, but throughout history, there's never been a case where there's just been a single technology which explains why, say, the Industrial Revolution happened. Right. It wasn't just that we made better textile machines. You have improvements in sector after sector which are enabled by key innovations in specific sectors. But for example, yeah, it's important to learn about what's happening in bio and robotics, et cetera. So I want to get into those fields also. I've just been interested in the fact that we are finally getting to a pace of change that we actually have seen before in history, but not for a long time. I most recently interviewed this biographer of Stalin, Stephen Cochin, and I think Stalin is born in the 1870s and you just think about his life from the 1870s onwards. Railway planes, airplanes, steamships, radio, telegraph, light bulbs, combustion. I mean, World War I is a crazy example where you start off the war. I think there were. The Wright brothers had flown, but there weren't like many. They were like on the order of hundreds of planes in the world. And there were no tanks. Like, tanks was not a thing. And World War I ends with, it's a tank war, it's a plane war.

Jack Clark (25:31)

In not that many years.

Lex Fridman (25:32)

Yes. You go from almost no trucks to tens of thousands of trucks over the course of four years.

Jack Clark (25:38)

And planes.

Lex Fridman (25:39)

Yes. So I think even planes were at extremely low level before the war and.

Jack Clark (25:43)

Then military use during, obviously.

Lex Fridman (25:45)

So, I mean, the reason Germany thought it was going to win is because it had this railway network and there was this plan how you could do this two front war and knock out both France and Russia at the same time, just by. You'd be amazing at railway logistics. And I think von Molke, or whoever the leader of the German command was, said at the end of the war, we lost because of trucks. Right. Like, we didn't anticipate that there was another way to ship enemy combatants to the front. Yeah, we're just going to see this level of change across so many different sectors.

Jack Clark (26:18)

There was somebody posted on Twitter, Bucko posted something that I thought was an interesting point about you, whether or not it's true. But I'm curious how you react to it, which was. It seems like you went through this evolution where you were learning a ton about AI and. And then it seems like you believed that AGI was coming. And then your interest started expanding out into all these other things, like, you know, geopolitics, biology, all these other areas. I'm guessing the, you know, which was sort of like saying, you know, the technology is what, you know, creates, you know, moment for change. But then this backdrop of the world is what really influences it.

Lex Fridman (26:52)

Yeah.

Jack Clark (26:52)

I'm guessing that like the chronology of your interests were a little bit different than the framing. But I'm curious just like how you think about that.

Lex Fridman (27:00)

Given what I do, I'm actually quite pessimistic about being able to learn from other fields. I just know people who will read some philosopher in the 19th century and they think, oh, this explains how Silicon Valley works, or this explains AI. And I'm like, no, I think you just have to read the papers about AI. I just think it's very hard to generalize.

Jack Clark (27:17)

You're saying to understand the technology you have to read the papers.

Lex Fridman (27:21)

Yeah, I think just people have this idea that I'll come up with a grand theory of history.

Jack Clark (27:25)

Yeah, you're like, it's not philosophy, it's science.

Lex Fridman (27:27)

Yes, but just like in any domain, it's just very hard to have this. I know a couple of people who can do this and I find it really impressive. But what I noticed about them, they're just not hand wavy at all. So there are people who, like, for example, if you're trying to model how AI will impact economic growth, one way is just to read the sort of firsthand accounts of people going through in the 1500s and oh, let's read the biography of the Medici and so forth. And then there's other people who are like, okay, let's look at the growth rates going back 10,000 years. What is the long run secular trend? What actually explains what changed? Well, there's the endogenous growth theory, where the key change is that population growth and more people come up with more ideas. Okay, well AI is more people. They'll come up with more ideas, there'll be more specialization. So there's this very different mode of learning from other fields, which is empirical and empirical is maybe the wrong word, but just like very falsifiable and grounded versus I'm just going to read, I'm going to go to the library and just like read a bunch of random books.

Jack Clark (28:28)

Totally. So how do your interests connect to each other then? Like, are you following any particular threads or, you know, does one connect to another in a certain way? Like what's, what would you say is driving what creates various interests over time?

Lex Fridman (28:40)

Honestly, it's just a super bespoke whatever I happen to be interested in that week. If I'm reading an interesting book, how.

Jack Clark (28:48)

Are you spending your time? Like, are you reading a lot? Like, are you, Is it, do you learn mostly through reading, through talking to people? Are there other Methods.

Lex Fridman (28:55)

Reading, Reading. I think there's a couple of people you learn a lot from talking to in general. I've just sort of been disappointed about like, I mean, given the fact that.

Jack Clark (29:02)

You'Re talking to like some of the smartest people.

Lex Fridman (29:04)

Though it's also really interesting, especially so in some domains people can be super like generative outside their domains. I've sort of been disappointed about the fact that, look, you might hope that you could talk to some historian about World War I or about the history of oil or something and then they'd have insights about how this applies to AI but really those connections will likely come from you and not from them. When I was interviewing Daniel Jurgen, who's the author of the Prize, it's this book about the 200 year history of oil. One thing I thought was really interesting is that Drake discovers the first oil well in Pennsylvania, I think in the 1850s. And then the Model T car, I think it was like 1905 or something, that you finally have cars with internal combustion engines that people are using to transport all around the world. And this is a sort of industrial case use of oil. Before that most of oil was just wasted. It was only the kerosene component. So all of Rockefeller, all of that history that you sort of think about as a gilded age, like oil baron stuff that's happening when a small fraction of oil is being used just for lighting. Before the electrical light bulb was invented. And in fact when the light bulb was invented, I remember getting my days wrong on the Model T and the light bulb. But it's like around that area when the light bulb was invented, people are like, oh, Standard oil is going to go bust because what's the other use case of oil? And so I do find it interesting that it took more than 50 years to go from we have discovered limitless energy in the earth to here's a way to use billions of gallons of this stuff. And I think it has maybe interesting implications for AI where look, we have. Basically it's sort of shocking how cheap AI is. I think that's why I always find it confusing that people are optimizing on cost. Do I want 2 cents per million tokens or 0.2 cents per million tokens? So we have this commodity that we could potentially use at an industrial scale and we don't know how to. Part of it is just technological. We don't know how to get those tokens to be more valuable. And part of it is just like, yeah, what do we do? What's the internal combustion engine? Equivalent for AI, but.

Jack Clark (31:10)

So you feel like most of your learning comes from what you read rather than, you know, through your conversations with people.

Lex Fridman (31:16)

Yeah, I'm very lucky that there's maybe six to 12 people who I've known for five years, almost all of them I've had on the podcast, but who I'm in just extremely regular touch with. I've genuinely learned a lot of what I know from like this handful, this, like this group chat. In one sense it makes you feel really another sense are weird that I've like known them for five years and now they're also like super successful. But like, we were all college students at some point.

Jack Clark (31:40)

Yeah, there's a lot to be said for that. You know, it's that five closest friends thing. It's a big deal.

Lex Fridman (31:44)

Right. How do you feel about this? Do you. Do you learn more from talking to people?

Jack Clark (31:47)

I think I learn more from talking to people. I also think that, like, there's different sorts of things that you can be seeking. You know, like seeking the truth versus seeking a good decision are very related, but not exactly the same thing. There's, for example, if you're trying to learn from people about how do you spot great talent and what can you pick up from somebody, you're not going to get to the truth. You're just going to get to techniques and things that have been useful for other people that you try to apply to yourself. So for that kind of thing, I wouldn't know how to read about it.

Lex Fridman (32:16)

Anyway, but on object level stuff. So if you want to learn about what's happening in robotics, I don't know, I've been sort of underwhelmed by how.

Jack Clark (32:27)

Yeah, I mean, the schools of thought to me would be either you can try to go learn it for real yourself, or if that hill is too high to climb. Which, you know, for me, getting into robotics in like, the white papers way, I'm like, I would be, you know, kidding myself to think I could catch up and then get to the edge of anything and on any time scale that mattered. And so for me, there's the other move of is there a way, if you're going to do it, to shortcut the decision somehow to somebody and so who can you most trust to give you good information or something like that?

Lex Fridman (32:57)

I think we're also in a lucky position where we have enough public output that we can sort of reach out to people and they'll say, yes, this is a tougher position to be in if you're like 19 and I want to learn about biology. I feel very lucky because a lot of people do great work in many different domains. Mine just happens to be public facing by default. So I get more ability to reach out to people than somebody doing great work in private. Exactly. Which does create this flywheel where if I do make good content, smart people will be willing to talk to me. That helps me make better content. I think that's actually more relevant to why the podcast grows than just audience tells people.

Jack Clark (33:41)

Some natural flywheel, you're saying?

Lex Fridman (33:42)

Yeah.

Jack Clark (33:43)

Before sort of moving on to a new topic I wanted to ask you about that's related to this. You mentioned a bunch of really interesting ideas there around oil and history and Stalin and biology. And I'm curious if there's any other just ideas recently that have just really stuck out to you that you can't stop thinking about that have really gripped you. Just because I love hearing about these.

Lex Fridman (34:02)

The one that's been on my mind for a long time is I interviewed this geneticist of ancient DNA, David Reich. And what his lab and his research has revealed is that human history, first of all, we just didn't realize how much we didn't know about human evolution. Just like the story you learned in high school, all of it is at least somewhat false about how, when, where, who.

Jack Clark (34:21)

What do you mean?

Lex Fridman (34:22)

Like, did it happen in Africa?

Jack Clark (34:24)

Did it?

Lex Fridman (34:24)

A big chunk of it didn't. Like there was a group that went out 400,000 years ago and then they mixed in back with the group that left out of Africa 70,000 years ago. A lot of the evolution that led to this branch of humanity maybe just didn't even happen in Africa. When did it happen? We're learning more stuff about it and then the key thing we're learning is how did it happen? And it seems like we just see this pattern again and again in history, which is very disturbing, but super recurring, is that some small group will figure something out. And it's not clear from the genetic record what it is. Right. 70,000 years ago, there's this group of 1 to 10,000 people in the Near East. So where the Middle East, North Africa are right now, they figure something out and they wipe out every single other species of humans across all of Eurasia. Like there was half a dozen different species of humans. There were the hobbits, obviously the Neanderthals.

Jack Clark (35:25)

The Hobbits.

Lex Fridman (35:26)

I forget what their like real biological name is, but like, I think they're called the Hobbits.

Jack Clark (35:31)

That can't be it. Yeah, yeah, that's good.

Lex Fridman (35:34)

The Denisovans, they're all wiped out by this one group. Like it starts out with like 1 to 10,000 people, expand all through 10,000 years ago, Anatolian farmers, also from the modern day Middle east, they expand out, kill off like 90% of the hunter gatherers in Europe through Asia. This also happens again, by the way, with a group that goes through the land bridge to America. They also keep doing this. There were multiple waves and one of the waves killed off the remaining ones. The only people who survived, by the way, interestingly, that we have genetic evidence of is this group in the Amazon, where, because the Amazon is so dense to get through the genocide wasn't completed. And so it was more of an intermixing. Then 5,000 years ago, the Yamnaya, which is this group of steppe nomads, they sweep through all of Eurasia again. And we're talking about 90% death rates of the domestic population and not just multiple continents. All of Europe, the people who build the Stonehenge are killed off by these people.

Jack Clark (36:39)

This is insane. And they're pretty sure this is right?

Lex Fridman (36:42)

Yes.

Jack Clark (36:42)

Wow.

Lex Fridman (36:43)

By the way, the way you learn why it's a genocide or why it was violent is you look at the fraction of maternal versus paternal DNA that comes from the native population versus the invading population, and what they'll find is that the maternal DNA comes exclusively from the native population.

Jack Clark (37:00)

Wow.

Lex Fridman (37:01)

All the paternal DNA comes from the invading population, which means that the guys were killed off. Wow. It also explains a bunch of. So like India today is a mix of the original Indus valley civilization from three to 5,000 years ago, plus the Yamnaya. And so all of India is just like this gradient, basically. Like north is more of this Yamnaya ancestry, south is this Indo, Indus Valley civilization ancestry.

Jack Clark (37:28)

So basically we just had a lot of this wrong.

Lex Fridman (37:30)

Yes. But you know what else is really interesting here? For hundreds of years, anthropologists, archaeologists have been doing this Indiana Jones thing. We're going to read the squirrels and we're going to think hard about what they're trying to say here and read the literature or whatever. And it was just so useless in comparison to one mathematician who went into this field and he's like, okay, let's just look at the haplotypes and see how they compare. And just totally redefined our understanding of basically all of history going back millions of years, even to things that happened like 500 years ago. All of that is totally. We can re understand it. All these mysteries about why did the Mycenaean civilization Fall. And who exactly were the Mycenaeans in Greece? Just all around the world we have.

Jack Clark (38:19)

Stuff right up to, you know, a certain amount of history though, right? Okay, yeah. At least there's something we can hold on to.

Lex Fridman (38:25)

But I just think it's very interesting, actually. It's like that. No, it was just like all this sort of like esoteric reading and understanding.

Jack Clark (38:33)

What was happening in the year 500, that we could have that super wrong.

Lex Fridman (38:36)

For example. Oh, I mean, there's something really interesting we learned. Speaking of the year 500, is the Roman Empire, I think is like, was it 540 or so? Is that there was basically the Black Death kills off like half, close to half of big fractions of the Roman Empire. And this is around when the Roman Empire falls. There were also previous plagues like the Antonine plague in the 2nd century or 3rd century. I interviewed somebody about this. I didn't realize the extent to which Rome actually had something as bad as a Black Death and how that contributed to their collapse. Because another bias we have in history is just to think about, oh, we had the four good emperors and then they did a really good job, but then the next guy fucked it up. No, just like there was a climate optimum during, quote, unquote, four good emperors where the breadbasket was really.

Jack Clark (39:25)

Civilizations happen to fall after a certain amount of time. It's like, no, a thing happened.

Lex Fridman (39:29)

Yeah, exactly.

Jack Clark (39:30)

I often wonder because over the last few hundred years we keep learning a new thing. Something was completely wrong, whether it was the Earth's round or gravity works this way or whatever else it is. And I'm like, we must still have big fundamental things wrong, like human history and that kind of thing would be like a good example. My 5 year old recently, which was a funny question, was like, do you believe in Jupiter? And I was like, I do, but that's a great question. And you could ask me about something else in the universe and I might think we have it super wrong, but we must still have big basics wrong.

Lex Fridman (40:03)

Yeah. I feel like this way, especially when I talk to physicists, we're just like very basic questions of like, is the universe infinite?

Jack Clark (40:10)

Yeah.

Lex Fridman (40:10)

Or not.

Jack Clark (40:11)

We have no idea.

Lex Fridman (40:12)

And there's like, there's many different kinds of infinities. It could be.

Jack Clark (40:14)

Yeah.

Lex Fridman (40:15)

But just that very like, it just seems like so consequential.

Jack Clark (40:18)

Or like the fact that like time warps when you go different speed, I'm like, we just like can't possibly have it all. Exactly. Right.

Lex Fridman (40:24)

Yeah.

Jack Clark (40:25)

It's just too crazy.

Lex Fridman (40:26)

Right.

Jack Clark (40:26)

Like it just seem. Or like dark matter. Like there's just like enough big stuff that I'm like, I bet we don't have it quite right. Yeah, but that's like, that might sound like some, you know, PhD physicist is listening and being like, what? What is this idiot Jack talking about right now?

Lex Fridman (40:39)

Not that we have the answer, but yeah, I mean, many of them would agree that like, yeah, a lot of stuff has to be.

Jack Clark (40:43)

Yeah, okay, this kind of flows into the next thing I wanted to ask you about, which is your broad sort of perceptions of the way learning is happening. And you know, you've obviously been sort of like learning in public. You do a lot of self directed sort of education and things like that. You know, you. But you're also, you've got one foot sort of very tied to people in academia and at the top of research fields and things like that. And I'm curious, sort of your opinion on sort of this transition that's happening where a lot of learning and the way people think that the stuff should get consumed is self directed and not part of the big institutions and things.

Lex Fridman (41:16)

The standards people have for like, is this thing true? Do I really believe it have just really degraded, especially in sort of podcast land to criticize my own tribe? People will just like, people just fucking say shit. Whatever you say about academia, there is this idea like, okay, does this make sense? Have you actually made a clear argument? Do you even have a clear end statement? Or is it just the thing that people are saying? On the other hand, look, I mean, is it net good for the world? If you read history and you read about the worst things that have ever happened, the Cultural Revolution in China, the Great Terror in the Soviet Union, you can complain that Twitter has low average iq, that the conversation is very dumbed down, but you just don't need that many IQ points to realize the Cultural Revolution is bad. You just need some mechanism where you could have gone on and been like, why is Mao having us kill all the sparrows? Isn't that actually really bad for pest control and just making fun of this deification? And I think that actually has worked. Woke was a thing for a second and then I think social media contributed to that being less. And I think also making fun of Trump is like a thing that people do and has worked and so on net. I just think like getting rid of the worst accesses is much more important for making history go well than making sure that we can have these giga brain genius level takes all the time. And I think social Media does it, like, a reasonably good job of helping.

Jack Clark (42:35)

Us correct the worst accesses on this sort of truth point. I think part of the issue is that, like, the legacy kind of media corporations, in my view, have, like, lost quite a lot of trust from people like myself included in many cases where they seem, you know, like they've got agendas and it seems like they are, like, you know, for profit and maximizing eyeballs. And so I'm like, you know, citizen journalism on Twitter, I'm not going to trust that completely. But I also don't trust, you know, like, the institutions completely. So I don't know that one's like, that much better than the other at this point.

Lex Fridman (43:05)

I disagree with this. My attempt to do this thing has actually given me more respect for the media in a couple of ways. One, I think they genuinely are better at holding power to account than sort of independent creators. Like, talking to somebody, extremely powerful politician or business leader and then asking tough questions is a thing that the media will do and often won't happen if they could just get to go on the podcast of their choice. And it's like, it's harder than it looks Y. And they're willing to uphold these standards when they do these interviews? No, I mean, I agree that it can often be sanctimonious when they do this.

Jack Clark (43:47)

Is that to do with the person or the institution? Like, as an example, like, you know, Tucker Carlson goes Fox News to Indy. Same guy, theoretically. Like, does that change his truthiness?

Lex Fridman (43:58)

I mean, I think this is another example of, like, his show and many others. This is not coming from a place where I'm making an object of a political point. I'm a libertarian. I'm sort of close enough in embedding space to these people, but the standards of discourse in these new places are just abysmal. If you had a conversation, you could just save one of a thousand things. And if you just paused on one of these. What exactly do you mean here? Why do you think this? And obviously, oh, I heard a thing in a group chat or whatever, whatever you say by the New York Times, they just genuinely have fact checkers who will go through content. I know that there's many cases where they failed by their own standards, especially in tech journalism. And I don't like their bias in these kinds of things, but just the standards of an order of magnitude difference.

Jack Clark (44:46)

I mean, my view is we actually probably at first, with AI, it might have looked like that was going to be a big problem for the New York Times. And maybe in Some ways it is, but I would actually, as time's gone on, I think we probably need these institutions more than ever in a certain way, which is that AI also now brings a whole new layer of FUD to what's true with deepfakes and random content generation by bots everywhere. And so you kind of, at some point do go back to needing somebody to really hold the standards of truth as much as possible. So I would think that should make these institutions more necessary.

Lex Fridman (45:20)

Yeah. And I think we always had to compare AI against the counterfactual. I mean, this goes back to the. Are they making us smarter or dumber? Yes. Do they hallucinate? But they are probably. I don't know. When I talk to an AI, I feel like this goes back. Do you remember, like, 10 years ago, people were like, oh, we can't trust Wikipedia because anybody can edit it. I feel like that was always a psy off. I think it was reasonably trustworthy. Always.

Jack Clark (45:40)

Well, there's one version.

Lex Fridman (45:41)

We have this attitude towards AI. We're like, oh, hallucinate.

Jack Clark (45:43)

But there's one version of AI where you're asking it directly. There's another version where somebody who has propaganda intentions uses it to their advantage. We don't have a lot of time left. And I wanted to get to the last topic. So up until you, all of my guests have been, you know, VCs or founders. And, you know, I've been sort of using this as an excuse to kind of publicly learn from them. One of the things I wanted to sort of learn from you is about the way you've done podcasting, because you've done it as well as anybody I've seen. You were like the. The, you know, the one person I reached out to when I was getting started for advice. You gave me really good advice, which was to just, you know, basically all centered back, to just like, be authentic, follow your interests, like, don't post stuff you're embarrassed about, like, all that stuff. And that was basically like the one North Star that I had. But I'm curious, because you've had so much success with it, if you can kind of reflect at all about what's made it work or why has it played out the way that it has so far?

Lex Fridman (46:36)

It's sort of very hard to say from the inside. I feel extremely lucky that my job is I get to sit down this morning and decide what do I want to learn about. Over the next few weeks, I'll interview the person who's the best in the world at that. I Get to pepper them with questions for a few hours, and then I get to repeat that. Week after week, I try to ask the questions that I generally want the answers to, including the questions I want the answers to. After having done two weeks of prep in that field and hopefully having had learned about it over the previous years, and so much content is very much like, give us the intro chapter of your book again. Explain this very basic thing in your field. I think people just really appreciate the feeling of being a fly on the wall. Like, one of the reasons it's valuable to be staying in San Francisco is that you go to dinners or events where you will miss a ton of context. Like, people will know a bunch of things and you won't know what they're talking about. But it sort of raises the bar. And immersion learning works. I try to provide that kind of environment in whatever field I'm trying to learn about. And I think people appreciate, like, not being talked down to having a sense that, like, the host is actually interested in the questions they're asking. These are like, if they were having a private dinner. I think the dynamic to replicate is if you were, like, at a private. A dinner party. Well, with. You wouldn't, like, just be deferential at a dinner party. You'd be like, you'd hassle them if you disagree with them about something. There'd be a fun vibe, and you wouldn't, like, can you explain this concept for everybody else here? You wouldn't have that dynamic.

Jack Clark (48:00)

Somebody that I spoke to recently who worked closely with Steve Jobs, who I'm actually going to have on the podcast at some point soon, said something that I really liked, which was that one of the things that made Steve Jobs special was that the fundamentals of just operating day to day, the way you talk to people, the way you give feedback, the way you ask questions, was just really good at those fundamentals. Before we started, I asked you about, what have you learned about conversations? Because I see that as a big fundamental thing that everybody does that you're very practiced at. But you made a point, which is that actually what it is for you is preparation, and that that's the centerpiece, and that's your fundamental. I still think that's highly applicable to everybody because we're all, you know, going to interviews as, you know, either somebody on the candidate side or, you know, the employer. We're all, like, you know, meeting people for, you know, all sorts of things, but everybody's preparing for stuff. So what is your preparation like? Like, what does that mean? For you when you're saying, I'm preparing really hard for this and it's like, you know, the center of your YouTube.

Lex Fridman (49:00)

In some sense, it's the very obvious stuff. It's if you're interviewing a researcher in a field, read the key papers. A couple years ago, before, back when I was just started getting into AI and I was about to interview Ilya, I'm like, okay, I'm going to program the transformer. This is how I'll learn about this and then try to talk to as many researchers as I could. If I'm interviewing a scholar in field, I interviewed this person who actually wrote a rebuttal to the Power Broker, which is this book about how Moses changed in New York City. That itself, I think is a 1500 word book or 1500 page book. I read that and I read his rebuttal of that book and I read review articles or whatever, different things about New York construction history. I try to do this for all my guess, but in some sense it's like very obvious. Just like read the things that could potentially be relevant and then write down questions. Obviously, the thing I've changed over the last year is I've started using Spaced Repetition. Spaced Repetition is this tool where you basically write flashcards for yourself and this software serves them to you every couple of months.

Jack Clark (49:57)

Like you have to be preparing for something well ahead of time.

Lex Fridman (50:00)

No, this is for consolidating knowledge across interviews. So if I do an interview, I'm actually going to retain what I've learned because a lot of concepts connect. I mean, especially with AI. With AI, you're just trying to predict what a future civilization of different kinds of beings will look like. And there's no domain of knowledge which is not relevant to this question. Right. So obviously technical AI stuff is relevant, but history, anthropology, even primatology, what happened between primates and humans, everything is relevant. It'll come up in the interview. And so just having it cached through tools like this is extremely valuable.

Jack Clark (50:32)

So it's like you're retaining a curriculum of all your work.

Lex Fridman (50:34)

Basically, it's the kind of thing where if you're reading a book, I think either you just shouldn't read, at least if you think you're doing it for learning, or you should have a very intensive practice around. I'm going to make the cards, I'm going to write the whatever thing is the equivalent of doing practice problems for the domain you're trying to study. Because it's shocking to me how often I will make a flashcard for A topic where I'm like, okay, this is so basic. I don't need to write this down, but I just have to do something right now. And then a week later, I'm like, I was on the verge of forgetting it. And you just think about, how many books have you read in your life? Hundreds, Right? How much have you taken away from them? Or conversations, whatever other medium? The lack of efficiency here is really striking. And so I've been thinking a lot about how to make this a process where over the coming years I can be like, I'm really getting better over time. I'm not just, like, doing the next thing.

Jack Clark (51:24)

It's sort of the spiritual opposite of the idea of an AI that's always listening and remembering for you so that you don't have to remember any conversation you ever had. And it's sort of just like some thing on your person that is constantly listening and you can always go back to it, but everything's captured for you. You're like, it needs to get in my brain so that I can learn the next thing.

Lex Fridman (51:41)

This is exactly full circle from where we started, because people will say in response to the continual learning on the job training stuff that, oh, we'll just have an external memory system, that it'll be just like a document of things the model has learned. ChatGPT already has this, and I think a lot of cognition is just memory. Like, it has to be on board. It has to be cached the whole time.

Jack Clark (52:02)

This is a great place to end. Thank you so much for doing this. I hope you didn't mind being a guest and keep doing your amazing work. I love to watch it.

Lex Fridman (52:09)

It was super fun. Thanks for having me on.