⭐ A Conversation with Alex Karp, CEO of Palantir #WEF22 #Davos

Well, first of all, I'm very happy to be here. It's high stakes since it's an audience composed of people I know or know of, and so that always makes it harder because I'm going to get negative ratings for all my answers. There's always been this thing about Palantir, partly understandable: shadowy company, works with intelligence services and lots of industry. There's an easy answer to what we do: we're a data integration platform that provides software that's actually operational. But the real answer involves which product you're buying. If you're buying the Intel product, which — kept in all modesty — Europe and many other Western countries and some non-Western countries are largely safe from terrorism with the help of some people maybe eventually in this room, by providing a platform where you can find terrorists and organized crime while protecting civil liberties. That's one thing we do, that's what we became famous for, especially used in highly regulated Europe. We have a number of products we built in the actual shadow world of the military that we mostly have not talked about: products for special operators so they can map their operations, share their data, get data from space, so they come home alive and well (and occasionally other people don't). We built a commercial product that first failed; we built a second commercial product. If you're vaccinated in an English-speaking country, America and England most notably, you got your vaccine because it was distributed in Foundry. It also powers Airbus, a lot of activities, credit to Swiss heavy industry. Then we have a bunch of products that are kind of Germanic: you have to be highly technical to understand why you need them, like Apollo, which allows you to work on-prem as if you were working off-prem and provide primitive maintenance to your software. They're very hard to understand but actually power your life. All that comes under the rubric of a company that's not particularly communicative, so we get this reputation of being a little bit in the shadow of the night. Some of that sometimes justified, sometimes not.

We can't talk about anything we're involved in in that region. I would say, like in the case of anti-terror, the most important actors in Ukraine are the Ukrainian citizens. We have some Ukrainian citizens here and some representatives. I think all of us — certainly I — look at them and not only reflect on their heroism but hope and wonder if we would have the same degree of heroism if we were in the same situation. Something none of us hopefully will ever have to find out. What I can say is there's a turning point in the art of war. If you look at US spending on militaries, about 800 billion dollars; Russian spending is about 65 billion; China's is roughly 300 billion. People look at the American spend and wonder, could they do the same thing for much less money? That has led people to underestimate the value of what will actually happen if you have heroes, the most important component, low relatively simple kinetic things that blow things up, and an ability to organize where those things go. That combination has shown — and the other shift is people tend to think offense-defense, but it's really unclear whether this is offense as defense or defense as offense. It's very clear that the courageous kinetic plus software side of this has been much more powerful than anyone would have realized. That's going to shape the way people — it also means small countries can now perform against big countries, considerations of countries wanting to invade other countries will be very different in the future. It's a hybrid war in many ways.

There's a technical software aspect to this: how do you track a molecule from coming out of the ground, its transformation through the supply chain into products, and how do you calculate that? Non-software people might view it as an easy software problem, but it's actually quite hard. For a very similar reason, it's very hard to find terrorists without violating civil liberties. You have to be able to take apart the data and re-concatenate it across systems that are not built to communicate, and when they are, maybe communicate too much. There's huge heavy lifting involving software. One of our products, Foundry, is ideally built for that. It's something we all care about, and I think you'll see us increasingly involved in that space.

Well, there are some banalities: we actually make an effort to say what we think in public. That may not sound too courageous sitting on a stage in Dallas, but to a journalist it sounds terrific. Being open and relatively honest about what one thinks is an act of courage nowadays. We've been doing this since inception. For example, when we built PG (post-9/11, really no one cared about data protection), we spent an extra two years building data protection at the core of that product and all our other civilian products. We've made hard choices: we haven't worked in certain countries ever. We didn't have to pull out of Russia; we made decisions to not work in certain countries where we could work. You could say that's easy, but tell your investor you've just cut the addressable market from their perspective in half, while telling them you're not planning to go public, and then try to survive with the job. We have walked away from deployments. We were the first to speak out against a Muslim database in America, which presumably people somehow thought we would build. We were the first people, I believe, to explain to Silicon Valley publicly (not over a cocktail party) that we were not going to support the parasitic nature of building companies. We were the first to speak with our feet: we left Silicon Valley, we did a DPO, which basically means I took a huge risk of becoming broke. If you do an IPO, they give you a sum ahead of time, but we did a DPO. I was on record warning my fellow progressives that Trump was going to win, also on record saying that he was not going to be the best president, in graphic terms. It just goes on and on. Also internally, these are among the most talented people I know, also the most difficult. There's a corner over there; I'm talking about you guys. Throuples, I'm sure that's also difficult. One of the more courageous things you can do is stand up in front of Palantirians and explain to them why they're wrong. See if you survive that — it'll definitely sharpen your intellectual weapons.

First of all, not everyone may know ICE — Immigration and Customs Enforcement — the institution well-known for customs and border control, although they do a lot of other things Palantir users use to stop human trafficking, prevent drugs like fentanyl from coming into America. Almost all human trafficking in America is actually stopped by ICE, but they are responsible for deportation of undocumented people. Our product is primarily not used for that, although we avoid white-tooth Disneyland marketing. It is used primarily for deporting undocumented felons. But people were correct to point out that while deporting undocumented felons, you may notice people who are not felons also get deported — that's a real issue in America. Importantly, in Europe, Germany as an example quite happily and according to the law does deport you if you're in the country without a visa. I lived in Germany with a visa for many years. That was a very difficult issue, but quite frankly more difficult because they were using our software for the same thing under President Obama and currently under President Biden. The real issue was people didn't trust Trump to do the wrong thing. There's a real question: if you're in conformity with the law but the actor is someone you don't trust, is the law still the law? That was the core issue. Lots of Palantirians were either unhappy that we didn't back down or unhappy that they were being protested every day. Nothing like the warm welcome of a protest to wake you up in the morning for a year to clarify your views. That clarified our views. It was a hard decision; I'm glad we didn't back away. Part of the reason we didn't back away is because Silicon Valley has every excuse for never backing forward. There's always a reason why Silicon Valley cannot help the West, cannot help America, can't help your governments, but can export the highest quality revenue to their company and get a 50 multiple and capture a lot of personal value. I obviously understood that America, like any large institution, needs to be critiqued. But in general, the West in general, America as well, play a crucial role in peace and well-being. Current events show that bias to be largely correct.

I paid my bills for a number of years writing papers on Freud because I was trying to be an academic in Germany. I always wonder: when you ask a person what kind of person they are, even if they're not trying to lie, do you get an accurate answer? I don't know the answer to that question. I'm definitely much more of an artist than a normal human being, certainly than a normal corporate leader. On the positive side, I have very strong opinions about people and I'm willing to fight for them. On the negative side, it's sometimes hard to convince me, and a lot of the ideas I have look very crazy. I will sometimes avoid engaging in dialogue where I think my crazy is right, and I'll try to convince the person that the crazy thing will work. My parents were very high IQ and also kind of like Palantirians. It's a Sisyphean thing, where you go back to the problem you had as a child: how do you get your parents to be successful? So I have a very deep relationship with Palantirians, but it is probably more about getting us to be successful despite our talent. That's a winding answer, but I don't actually know. That's what you get when you ask people what kind of leader they are.

I was super unhappy at Stanford. I thought, if I'm going to finish this sentence, I'm going to do exactly what I want for the rest of my life. What I wanted to do was become an academic. The only sources that really interested me were Germanic writing, including French people like Durkheim. I thought I'd just spent a three-year bid, I had to do something I really wanted to do. That's why I went. I stayed because I felt culturally very at home. America was never really comfortable — exaggerating the nature of my friendship, I didn't have 50 close friends. I like to be alone. I don't know how to tell someone I love them after three dates; I don't even know what a date is. I felt at home because I thought that's how everybody is. I didn't know there was a whole world where I could just be me and not have to explain it. Honestly, if the Sigmund Freud Institute had money to keep me, I would be watching you on TV with some other weird-looking dude. Maybe I can write about Freud. And then this idea came along.

It depends on whether you look at the accomplishment positively or negatively. It takes five years for us to build one of our core products. It took five years to build PG, took us all these products I mentioned. Then it takes a long time with big enterprises. Build product is five years; maybe you begin selling it year four; full productization is year five. We have a number of products. It's very complex business. I think the duration on building enterprise is very different. I do happen to believe that because the duration is long, the longevity is even longer. These are very complicated enterprise software products we're building. Even if you're the best in the world, you're going to take three years to get them so you can ship them, another year and a half before they're fully productized. We have five products.

It's not particularly with one or the other. It's a cultural critique. Silicon Valley 1.0 built dual-use technologies primarily for the military and then shipped to consumers, which made us healthier, wealthier, more peaceful, and as a logical consequence led to a radical decline in activist views like racism, sexism, homophobia. Silicon Valley 2.0 figured out you could throw out that framework and jump right to being parasitic. The parasitic nature is that the host decides it loves the taste of its own destruction or decline and will pay with its own product — data, enterprise data — to the tune of high-value revenues. It's a Matrix-like thing that works so well for Silicon Valley and investors, and we're all paying the price of the game. You see this all over society. I don't view this critique as particularly insightful now, but it surprises me that no one was the first to really talk about it openly. I hope Silicon Valley moves back to a hybrid where products deliver some value not just to Silicon Valley participants.

There are a lot of reasons to move to Colorado, and one of them is it's nice when your neighbors like you. It's surprising to show them in Colorado, they're like, 'Oh, you're Palantir? Great!' So yes, it's very enjoyable.

Absolutely. I believe I'm certainly in the intellectual tradition of the left. The intellectual tradition of the left is a tradition that believes in quantifying things, believes we should decrease violence, that workers should participate in the means of production, meaning people should have some participation in the economy and ability to participate in all areas of society and be able to prove themselves and do well like everyone else. If you define leftists as something which arguably is not leftist, you could argue I'm not a leftist. Historically, left of center has meant intellectually, I'm pretty squarely on the left. In America, I'm definitely a leftist. In Germany or Switzerland, I'd probably be center-center right. Center-center right people there see me as... well, they think poor people should be well educated, older people taken care of, everyone should have teeth in the front of their mouth. We still count as communists in Texas.

We've been fighting about politics since we met at Stanford. I think these things are a problem only if it's a surprise or if you're committed to only having friends who agree with you. We always disagreed politically, still disagree on a lot of issues. Politically, in my opinion, we seem to have a somewhat similar view of what things are and how things work — though not always the same, and that's quite helpful. What it means to build a business, what a successful business will look like, these are perhaps more ephemeral but more relevant. The fact we don't agree politically is refreshing. You come to the same conclusions, right? Turning your back on Silicon Valley: Peter went to LA, you went to Colorado. Not always on politics, but you can have similar assumptions and different interpretations. We don't always have similar assumptions, but in the Venn diagram there's an overlap of assumptions, often not overlap of conclusions in the political realm, and a pretty healthy discourse on the business side. Steph, Peter, and I have made a formidable team in the last couple years. Both of them are good friends. They're actual friends, so to speak.

It depends on what. I like institutions that are competent playing a bigger role. The size of the role should be proportional to competence. Part of why I think I'm on the left is I'm a modernist, not a postmodernist — I believe in reforming institutions as they are, under the assumption they will stay in our life, and reforming them is key to ameliorating social ills. Like Warren Buffett or Bill Gates, would you like to pay higher taxes? I think the real issue for Western societies is how functional institutions are. It's very easy if you're not highly tax-optimizing. My hobbies are cross-country skiing, tai chi. How many pairs of cross-country skis can I buy? You can make it expensive, but no matter how creative you are, I couldn't afford that. The key thing is that software is crucial. Part of the instability of the West is a legitimation crisis — the average person doesn't know what they're getting from institutions, and they're not happy. You change that by finding terrorists with data protection, fighting wars so you're so strong no one fights you, getting citizens who are happier, more peaceful, more tolerant. That's a better way for all to live.

Take Germany: German companies and political officials come to America, see a company, and think, 'We're smarter than them.' True or not, they go home and build a science project. Since they're smarter and better at science, their science products will be better. But this isn't primarily science; it's primarily art. Get the people who've built the art — many are German, studied in Germany, have life in Germany — get them to build companies here. I tell people this, warned by our PR department that's always unhappy with me: stop spending money. The money will follow you. Every time, countries I care a lot about say we'll spend 10 billion euros after a consortium of consultants who've never built a company tell you how to build one. How about hand out some passports, tell people to come back, call them personally. Germany has one of the best tech scenes in the world, just in Canada and Silicon Valley, not in Germany. There's no sense of urgency. Instead of getting creative and interesting people, you get too many consultants, too much money, and copies of American tech companies. What does it say about an incredible culture that creative, tolerant, interesting people are building the best companies in Canada and America? You get third-rate copies and billions wasted. This region could be the second best in the world, but there's no urgency — and urgency means focusing on what actually works: people who look batshit crazy moving to America. Get those people to move back.

Thank you.