The Exchanges

Q So what do you, what do you recommend to people who might not be technical, but who are listening to you now and be like, you know what? This guy's crazy, but he might be onto something. Um, so I'm willing to, to dedicate some, some of my time in 20, 23 to learning to be more technical and shipping one product or two products.

A I mean, I think the, you know, what I did was I really wanted to build something and The only way for me to build that thing was to learn how to code because I had no money to hire anybody else, which is true for most people. Engineers, by the way, are very expensive. So if you can't learn a code, let's just say you're, you're, you're going to spend 250 grand a year on one person, right? So good luck building a successful company, uh, uh, with that. That's just almost impossible. Um, so I would say find something that you really want to build. It doesn't matter how simple or easy, uh, the, actually the simpler, the easier, the better probably. And build that thing. Like the first thing I think I ever built was a GPA calculator, where in just an Excel, you, I put in my math classes and then I put it or, you know, my classes or whatever. And then I put in like my, you know, A, B, C, D, whatever. And then I just had a, you know, column that said like, if A, four, if B, three, whatever. And then, you know, like some divide by number of classes. And that was like a calculator. I built like little GPA calculator in Excel. And then build that with Python. Like, build the exact same thing, but just build it with Python. Like, you know, use ChatGPT, probably write 99% of that code for you, and then just write a Python script that's just local on your computer. But you need to solve, lik…

AI assessment note: “find something that you really want to build... And build that thing.”

Q So I think, you know, this is a great place to end, which is if people Want to get the life that they imagine for themselves? Like, how do you recommend them to reach that life?

A Honestly, I think everyone has it. That's my current belief is that everybody basically has the life that they want, because the truth is like, people do have the freedom to change if they wanted to, and they don't. Um, so most people like don't want to leave their family, and that's totally fine. Uh, I did. You know, most people don't want to learn how to code, and that's totally fine. I did. I think what's most important is that people should want, stop wanting things that they don't actually want. I think the biggest concern I have right now is that people look at social media and they think they want all those things. Um, and maybe they even do want those things, but they wouldn't have if they never saw it. They wouldn't want a six pack if they never saw someone with a six pack, uh, right? Like these sorts of things. And so I think most people should just really focus on what's really important to them. And make sure that they're getting those sorts of things. And I bet that they are, uh, and that they maybe just feel guilty because other people are saying, oh no, you should be doing this and you should be doing that. And you should be doing that. No, like, I don't think most people should be learning how to code. I never said that. All I said is if you want to like live my lifestyle, you're going to need it. Cause that's the only way you can automate stuff is with code.

AI assessment note: “stop wanting things that they don't actually want”

Q where it's going to go. So yeah, could I imagine that like, instead of a dog, it's a human and instead of a human, maybe they're playing soccer. And instead of like being trained by a human behind it, maybe it's trained, you know, with AI, et cetera. It, it makes, that's the arrow of progress or, or, or, yeah, I think that's, that's certainly the arrow of technology, right?

A Like if you wanted to build the best soccer player on earth, 100%. It would be a robot. It would not be a human being, right? Uh, the question, I think, is also, like, what does that mean, right? Like, for example, will people replace dogs with robot dogs because they no longer have to take them for walks if they are functionally as similar as a dog, right? Um, and I think a lot of people will not. I think a lot of people will say, will basically reject, uh, this stuff, just like, They're, I still paint with oil paints, right, for example, um, but will you be able to make money? No, because what it does is it shows you that this is basically valueless. I think these things have value in this weird cultural sort of bubble and zeitgeist, but the minute that a human or a robot can kind of do it just as well as a human, it sort of reveals the illusion that this thing is actually useful and interesting. And people will just move on to something that only humans will do. I think that's really what the pattern will be is that people will play soccer for fun, but not to make money. I think there will be new sports that robots cannot yet do like surfing. I think my, my, my, my feeling is that surfing is much harder, um, than soccer, uh, for a robot to do. And so, you know, uh, you'll, you'll kind of move the, the goalpost essentially. Right? But just for example, like how many people ma…

AI assessment note: “if you wanted to build the best soccer player on earth, 100%. It would be a robot.”

Q impact and a hundred times return on investment in the next two to three years. Personally, I think there's a bull case and a bear case. The bull case is way more conservative than what the median person on my timeline considers as completely self-evident. And the actual outcome we'll see is statistically likely to lie in between somewhat closer to the bear case. What do you think about that?

A I mean, I have, I don't care at all about the, the financial investment piece of it. Right. So, uh, is it going to be a hundred X? Is there any way to make money on AI? Like I have no idea. Um, but will it dramatically transform society? Yeah. A hundred percent. I mean, I, I don't, I don't think this is anything new. I think this is what data does. It automates rote input, and I don't see why, I don't see how, I mean, definitely I can see why people are saying, oh, this is just like Web three in the sense that there's lots of people loud and proud on Twitter. Um, but I think you could apply the same thing to tons and tons of other stuff, right? Um, that people, it just, we, we coalesce mimetically about certain topics and then Share those things. And the truth is people on Twitter, uh, are attention whores. And so we are, I mean, that's why I love the, uh, the, the view count, frankly, because it exposes us for the attention whores that we are. That's why we tweet. We want views. Um, why do we want attention? So we can make money at some point in the future based on that attention. And so, yeah, of course you're going to have like a similar sort of zeitgeist build up around AI. Um, 99% of it is, is BS. Um, however, like I spent most of the web three stuff reading papers, and I spent most of my AI time reading papers, uh, and so I can compare the papers from AI and compare the p…

AI assessment note: “will it dramatically transform society? Yeah. A hundred percent.”

Q What about, like, do you need to be writing code, or could you just understand how to read code? Like, you know, I get the argument where it's like, okay, um, You need to understand the fundamentals.

A So for example, if you want to be an artist, you have to learn perspective. You have to learn anatomy. And if you are not able to do perspective and anatomy, it doesn't matter how good AI is. It will not help you currently at least do anatomy, perspective, design, composition, uh, colors, all of that stuff you have to learn. Uh, and by the way, if you look at artists who are not that good, it's often because they don't actually have the fundamentals in place. Like they know how to draw dragons, but they don't know perspective. So the dragons look all stupid and wonky, right? Or they draw eyes that look like almonds. No, you have to learn the anatomy of the eye. And guess what? Is that fun?

AI assessment note: “if you want to be an artist, you have to learn perspective.”

Q I'm right. Yeah. Let's talk about that for 1:02. What would, If Naval was designing, you know, a high school or something, what would, what would it be like?

A I've talked to him about this. Actually, I asked him this exact question. I was like, how do you, you know, he has kids. Like, how do you think about, how do you just think about raising kids and teaching them? And he said, he, he said, not too different. I think it's like focus on math, basically. Like if I think, I think math is the underlying. You know, now I'm saying my own stuff, but like, basically, math is the underlying substrate of reality. It's the language that we use, and so if you can teach someone math, Then they can learn whatever the hell else they want, given the internet. Like, you know, math is the sort of, like, for, like, I've been thinking about this actually with AI. It's like, what is, what is missing? What is missing currently from AI right now? Like, what is next, right? Like, how do you, like, can it get better? What, what, what it does it take to get better? And I think the number one thing that's missing is basically discrimination, where right now you basically get all this data and you just, You know, you use that data to, but you, like, currently the AI doesn't get a choice of, like, did I learn the, is this actually right or not, right? Like, it just says, oh, cool, yeah, like, these strings lead to this strings, I can do that, right? Um, and I think what's really important, what a lot of people, I think, miss is, like, when you get new data, ri…

AI assessment note: “I think it's like focus on math, basically. [...] I've been thinking about this actually with AI.”