The Exchanges

Q So speaking, speaking of that, let's move on to AI, because this is obviously tied into this. You said recently, Alex, that some tech leaders are calling for an AI pause because they don't have a product ready. How much of AI right now is innovation in theater? What's real and what's hype? What should we be paying attention to?

A Well, the, the big thing that we have to avoid in, in, in our country, but, but there's less of a problem for us than in other countries, is there's a, just an attempt to dampen AI and its utilization by people who don't have a product, and that is like, this is actually a big deal inside and outside of government. Now, luckily for us in this country, it's a much smaller deal than say in Europe, where it's like, you know, there's all these discussions Some of which I very much support, and some of which we can power. How, where does the data go? How does it used? Where does it flow? What is in the algorithm? Is the algorithm discriminatory? But some of it is, you'd say in German, which means kind of just put out there in as a theatrical thing, because there aren't really that many. There are some, uh, companies providing AI in the form of advanced machine learning or large language models. Um, the, We, we have a slight bias here at Palantir. We build a software that will allow you to process large language models, rebuild the output of large language models, turn it into what we call an agent, which is a safe algorithm you can run across your enterprise, and we can interact back and forth with large language models or, or other forms of AI, so that the, the algorithm of large, large model understands your enterprise, but you don't outsource the knowledge of your enterprise to t…

AI assessment note: “put out there in as a theatrical thing, because there aren't really that many”

Q And you ended up studying under Habermas. Was that, was he somebody you'd studied previously, you were excited to work with?

A The longer version is, I went over the study with him, I was in his colloquium, and I, and I studied with him, and then we had a falling out, and then I finished my PhD with someone else. Um, but not on bad terms, it's just, I, I, I discovered, while writing my PhD, um, Uh, that the stakes were very low, and the personalities were very difficult, and I also discovered partly by in proximity to him, but in proximity to other people who are kind of world-class, one meeting with Luhmann, uh, watching debates between Habermas and Apo, uh, people came to the colloquium in Habermas that I had certain abilities they didn't have to, like, be a builder, and while I was Good enough to be in the room with them and argue with them on where I was a technical expert. Dedicating myself to reading and writing all day was not what I wanted. And they did not seem to understand. I, I wouldn't have put it in these terms because I hadn't built anything, but they didn't have the right personality for building.

AI assessment note: “I went over the study with him, I was in his colloquium”

Q What are a couple things they could do to fix it and be better? Obviously they've changed, they've improved.

A What do you want to see them do? They changed, they've improved. There's just, you know, in business, it's really important to go back to, like, things that are kind of obvious, but one of the obvious things that's not part of every structure of the U.S. government is almost all enterprise software that works in America, or in the world, has two characteristics. It's built in America, and it's built as a product. Yeah. And it has a third characteristic, quite frankly, that it's sold commercially. So, if you just say, the rebuttable presumption is, if a product is world class, it is a, built in America, again, rebuttable doesn't always mean, but very likely, it is, you could drill down on that, it's built by this, da da da, but just, it is a product, and somebody commercially has bought that product.

AI assessment note: “the rebuttable presumption is, if a product is world class, it is a product”

Q At a high level, what's Palantir able to do that other systems are not able to do?

A Palantir is useful For the Ukraine, some of it's the part that's public is obviously documenting war crimes, but war crimes are hard to document because of data protection issues, to take it to court, where the evidence come from, who's touched it, chain of custody issues, so that's kind of, uh, we're doing, um, understanding who's doing what on the other side is a classic use case for PG, so, uh, how is the adversary Into how, how is a matter of theory, you can just go through the products. So there's a kind of the civilian prosecution. There's in the PG foundry constellation. There's an ability to understand what the adversary is doing kind of with their operatives, who's doing what house doing is doing it with segmented, uh, uh, uh, data access, which is crazy important in the warm up. It's funny is that people don't understand you need segmented or high data protection because The, the knowledge that these people have is so valuable, meaning people will die if it gets out, that you have to control who sees what. Um, Gaia, meaning who is it, will allow you on the battlefield to, uh, be able to see where in real time, what, plan out your battle, attack, retreat, understand the adversary, and then the most powerful use of our product currently, and again, of one of our products is Being able to identify an adversary over a large lying mass using AI, and that obviously it's bee…

AI assessment note: “Gaia... will allow you on the battlefield to... see where in real time”

Q Let's go with it. I want to hear a little bit more about your background for everyone, Alex. Where did you grow up, and how did you become an entrepreneur?

A I think my road to being an entrepreneur broadly defined, so building things. Um, now I would define my role as building things that are, that transform key aspects of society. That, that's what we built at Palantir. It's a trans, it's software as a transforming engine. Really began, ah, with, like many things, something I didn't want, and was not overly, ah, forward about in public, which was my dyslexia, which I hid. So I was very good at certain things, and I would say still am that other people are, Not as good at. I legitimately saw the world differently. My brain is, was structured differently than other people, and I was able to perform at a professional level in certain areas, even as a young kid, but then really was Hindered in other areas, and we had this weird family structure where my mom was in grad school as an artist, uh, and everybody was an artist, so they were outside the norm. It was a heavily Jewish environment. My mom is black. We were pro-Israel, uh, and very, super erudite, heavily Jewish environment. I, everything about my life was further outside the norm than I realized. It's like extreme outsider for, for, there's no, there's no insider bias with that. I would have loved to be an insider somewhere, and the minute I could be an insider, I'm like, shit, I'm dyslexic. So it's like, I used to wonder, as a little kid, how could this get worse? You know, I …

AI assessment note: “I think my road to being an entrepreneur broadly defined... Really began... with my dyslexia”