Sam Lessin on Artificial Intelligence

Essays

From the Podcast

If you just step back and say, "Look, if you wanna understand our era, looking back 200 years, we are entertaining ourselves to death." This is the most entertaining business cycle in the fucking history of the world. It's so entertaining, right? And if you're just like, "Okay, you know what the algorithm's really good at? You know what really AI's really good at? Entertainment." And so the entire model of the world right now is let's make all the most entertaining stuff happen, and we're gonna love it Right? That's kinda where we're at.

— Sam Lessin, More or Less Ep 167

People love this shit. This is the most entertaining business in the history of the world.

— Sam Lessin, More or Less Ep 167

Longevity age scores are like the doomerism of health apps. Which is funny, 'cause, like, why don't all these apps say you're gonna die because of AI? Like, I'll get upset when they're like, "Sure, you're 20, but you're gonna die next year." The AIs aren't talking to themselves. Yeah, it's all just collective delusion, whatever. Kind of fun, though.

— Sam Lessin, More or Less Ep 167

No, no, no, it won't tell me who's Jewish. It won't help me my distribute my Jewish apps, but it'll help me build them, right? I need Grok to distribute them. I basically... Look, it's, it's the high holy days. I think that a lot of stuff is interesting, so I literally, like, I did a few things, but I created a great app which is just, like, combination Bible, Talmud, and then history kind of woven together with an agent I can talk to and ask questions of that's quite good and interesting. And then I just started having it send me, like, daily, like, "What's the Torah portion? You know me. Like, what am I gonna be most interested in? What are the big debates in it? How does it relate to, like, religion disconnected from land?" Which is one of the big innovations in Judaism. And it's like, it's so good, and I've said it to people who know a lot more about religion than I do, and they're like, "This is insanely good. Like, what are the sources on this?" So, like, I don't know. There's enough stuff online to make great Bible study, I guess, for interesting Old Testament Talmud stories, plus Talmud stories.

— Sam Lessin, More or Less Ep 167

Oh, I just like... It's... Like, Claude is so good at Judaism. It's wild. Like, I just, I-

— Sam Lessin, More or Less Ep 167

Well, and what's really interesting, right, is you have this thing where, like, take Bending Spoons. Bending Spoons is making really good business deals offering, like, a tenth of the private price because there's no liquidity. They're the liquidity provider and, right, and for this type of business. And then you have a lot of people who are like, "There's so much money with asymmetry on the minds." And if it... Drying up is more of a macro event than it is a tech event, right, is the interesting thing from my perspective. I don't know. I don't know how the ballgame ends. I mean, I know that we're so disconnected from any semblance of normal capitalism or reality, and I know it's gonna drive an unbelievable amount of, like, fear, loathing, anger, and, like, resentment in all sorts of ways. Like, this makes the dot com bubble look quaint. But who knows? Like, I'm open to the fact that this could go on for years. Or it just could be we've hit a too big to point fail, and this is the greatest stories ever told, right? The greatest stories ever sold.

— Sam Lessin, More or Less Ep 167

Look, Openwater was... It was good. It was a good, good in, good out. But just remember, I mean, I just go back to, like, don't you remember the quaint old days when Instagram was shockingly worth a billion dollars when it was bought? I mean, it's, like, hilarious.

— Sam Lessin, More or Less Ep 167

I'm investing, but you know what, Dave? You and I have not made gajillions of dollars because we're not trading on AI. Like, we're doing fine.

— Sam Lessin, More or Less Ep 167

Well, this is going back to the whole, the business thing. We have trillion-dollar stories being told, right? And you're like, if you're using many of them, and you can hot swap them, and it's a constellation of friends, it's very hard for any one of them to be worth infinity, right? Like, there's not a compounding advantage, and, like, I think that's all she wrote. Now, again, I wanna be really clear, that doesn't mean I wouldn't love to have a bunch of Anthropic stock at zero and trade out. There's money to be made, and I think everyone kinda gets that, but any thinking person, it's just so hard. The only bet I've heard people making, which I'm like, I rationally get it, is you're like, this shit, the world gets so weird, and, like, everything kind of falls apart, and it's just an asymmetry story. Who the fuck knows? And that's, like, pretty loose analysis.

— Sam Lessin, More or Less Ep 167

It's like the future is instead of having 10,000 nuclear weapons, you're gonna have 10,000 super intelligences in silos, and, like, when you need them, you call them, and the nice part is, is once your country is subverted by an evil AI, there's no price you won't pay to fix it. That's how you get, like, a great Austin Powers movie, by the way.

— Sam Lessin, More or Less Ep 167

I don't know. It's an interesting question of how it plays out. I mean, like, look, this does speak to my other big hot take on this, which is the only way, in my mind, that Anthropic and OpenAI become really, really valuable companies long term, not trading, not short term, whatever, five to 10 years, longer, uh, is effectively as defense companies, right? If you said to me, "Hey, what they're actually doing all day long with these super fancy computers and all this computing," you can't compete in the token marketplace You just can't, like, make any money. What you can do at the edge with, uh, 10,000 PhDs is develop a bunch of, like, zero days against other LLMs and not release them. Like, make Mythos squared-

— Sam Lessin, More or Less Ep 167

That, that's like the biggest companies in the world, they kind of unofficially do that. And the Sublime's pitch early on was like, look, that's fine for the top 10 companies, but, like, there's no collaboration in security between anyone else. Like, you're just kind of on your own, right? And so, like, how do you build, like, a collaborative platform for everyone to do security as a, as in, like a, in, like, a place that, like, is compounding? And so you could see some of that type of stuff start to evolve for this stuff. I mean, it's not gonna be humans because again, like, this stuff happens too fast, right? You're gonna have to have kind of, like, a framework for it. Like, how this all works is TBD and, like, who pays for it and whatever. I mean, like, I don't know. It's gonna, like, kinda... It's a nice idea. I will see how it actually plays out in a practical way. You know, for all we know, I would kind of expect there already are working groups on this between all the big companies, just as there are in other security domains. But who knows? I mean, I think it's, it's an interesting moment.

— Sam Lessin, More or Less Ep 167

... but it's just, it's fascinating 'cause, like, in our era, I would say that the policy department, it wasn't that they didn't exist, and there were some very great people who worked for it, but they weren't, like, that important on a rel- Like, it was products, right? Now I think we're in a world where, like, if you think about, like, who gets paid the most, who matters the most, it's who's representing these guys in DC and the spin mastering of it is just, like, very valuable and correct in a way that, like, I, I, like, I wonder if policy people are making 10 times what they used to. I guess they kind of are.

— Sam Lessin, More or Less Ep 167

Yeah. Like, they've done better, right? And so that's good. And then I think honestly, what we now see is, like, I'll call it third wave tech policy, which is the classic answer, which is embrace and extend, right? Like, you're not gonna change the narrative. The narrative is, this is scary. So like, great, let's take that narrative, let's double down on it, and then twist it and use it in a way that's advantage to us. So for OpenAI, it can be like, "Well, we're not gonna IPO," right? Or like, you know, for Anthropic or it's like, it's like we need regulatory capture. Like-

— Sam Lessin, More or Less Ep 167

No, it's just obvious, but it's just a game. I posted about this the other day, which is like, look, there have been three eras of how tech deals with, like, messaging and marketing. Era one, I think we'll call it the Bill Gates era, where, like, you get called before the stand and you're super indignant about it, you're super pissed off, and you look terrible. It ends up being a disaster. Like, it really hurt Microsoft, right? It, like, kind of broke its spirit when the world said, "You're doing something bad." They said-- And, like, it was just the, the indignation. Era two was, like, the Google and Facebook era, where everyone looked at what Bill Gates did and they're like, "Well, that was a mistake, so we're gonna fight this." The fight back has actually worked better than at least what Bill Gates did, right? In terms of being like, "Okay-

— Sam Lessin, More or Less Ep 167

Tweets