Alex Karp on Why AI Will Hurt the Middle Class: You Can’t Have Billionaires Becoming 100x Wealthier
- Karp argues Silicon Valley has broken the bargain connecting private wealth to public value. A Louisiana carpenter may see little material improvement over 30 years while technology beneficiaries become 100 or 200 times wealthier. Karp says AI lab leaders compound that legitimacy problem when they describe job displacement and dependence rather than shared progress: “building a dream that you don’t belong to.”
- Palantir’s enterprise AI pitch combines deployment with safeguards for proprietary data. Karp says companies pay LLM providers to train on proprietary information and fear the resulting models could compete with them; he cites particular anger at Anthropic. He presents Palantir’s ontology as both an “alpha creator” and a “protective layer,” making model output valuable while preserving institutional control.
- Karp defends legally authorized surveillance while advocating legislation against domestic use of battlefield AI. He distinguishes surveillance under a legal mandate from surveillance without one, including whether data is purged when warrants expire; he claims Palantir has prevented hundreds of terrorist attacks. Isaacson’s unresolved concern is whether that rational defense fully addresses public fear of the surveillance state.
- Karp expects AI to create more or higher-quality vocational jobs while widening wealth inequality. Even salaries rising at twice inflation are unlikely to match technology beneficiaries becoming 10, 20 or 50 times wealthier. He supports higher taxation conditional on identifying half a unit of government waste per additional unit raised, and rejects wealth taxes because he says they risk damaging American tech, breaking consensus and expanding surveillance.
- Palantir’s founding wager was roughly three years without revenue to build security and civil-liberties protections together. The episode’s opening describes Palantir as a $400 billion colossus; Karp credits an outsider’s refusal to follow conventional software economics. Dyslexia, in his account, helps leave him less “colonized by the broader culture.”
- Karp sees national purpose and recognizable everyday benefits as linked—and missing from Silicon Valley’s island culture. He recalls the Valley’s university-government-capital origins and argues military strength is necessary because adversaries will develop AI regardless. His endpoint is human worth beyond intelligence: rights must protect what remains “artistic, individual, particular, special” when machines surpass human cognition.
1. Palantir’s founding model required rejecting the software playbook
- Karp traces his outsider outlook partly to family instability and dyslexia, which he concealed until roughly six years ago. He defines dyslexia as a non-one-to-one relationship with the written taxonomy of language. He resists turning it into a tidy advantage—“of course, you’d prefer not to be dyslexic”—but argues that reading’s translation process leaves him less “colonized by the broader culture” and more able to engage with problems as they are.
- The founding application: Peter Thiel wanted to bring PayPal’s fraud-detection methodology into intelligence. Karp’s unconventional bet was roughly three years of development without revenue, serving two constituencies: “stop the terror attacks and protect civil liberties.” He says Palantir’s anti-terror product stopped hundreds of attacks, changed Europe’s history and later became fundamental to Ukraine’s defense. He criticizes conventional software for prioritizing simple, addictive products that become difficult to remove without necessarily creating much value.
- Halfway through his dissertation, Karp concluded academia was “not for me”: “I need to go and build something.” Builders such as Jobs and Musk, he argues, combine outsider status with aesthetics, humanities and business judgment. America’s ability to attract people who choose to migrate helps produce that mix.
- Karp sees the First and Second Amendments as radical but insufficient without Fourth Amendment privacy: he argues that privacy protects the ability to articulate and develop thought. He also describes the right to bear arms as a check on government, not only a matter of self-defense. “Our Constitution is built for builders.” Still an “unhappy Democrat,” he criticizes Heideggerian politics on both left and right, including movements that call themselves progressive while pursuing what he considers regressive, communitarian policies.
2. Surveillance safeguards have to work inside the software
- Isaacson’s pushback is concrete: what happens when a system records him walking through New Orleans? Karp distinguishes surveillance with a legal mandate from surveillance without one. His technical test is warrant expiry: what happens to the data afterward? He says Palantir’s products can keep data only while a warrant is active and purge it when authorization ends. He also claims Palantir’s products are the only way to use AI while protecting these rights.
- Karp cites European work against organized criminals and terrorist networks, and claims Palantir has stopped “hundreds of terror attacks,” many in Europe. His counterfactual is categorical: had those attacks happened, Europe would have some form of far-left or far-right fascism. He estimates 90% of society supports the relevant law-enforcement work; refusing it, he argues, risks a fascist backlash under which “you are tracked all the time.”
- On commercial privacy, Isaacson observes that students accept Google knowing everything while fearing government knowledge. Karp agrees with their concern about government access but says they underestimate how platforms shape supposedly independent opinions. He credits Europe for focusing on privacy in both commercial and government contexts; in his view, companies controlling data can “heavily influence what you think.”
- AI is a “raw material” usable for good or harm. Karp concedes that data collection and AI “can, and it will” undermine autonomy without resistance. He proposes legislation preventing battlefield AI from being used domestically, while insisting military development must continue because adversaries are building it. Weakening the military to achieve peace is, in his view, a fantasy.
3. AI redistribution must improve workers’ lives
- Karp expects AI to create more or higher-quality jobs, especially for vocational workers, with retraining. Yet wages rising at twice inflation are unlikely to keep pace with beneficiaries becoming 10, 20 or 50 times wealthier. Isaacson presses the 1890s analogy: another technology boom concentrated fortunes, so why not put more into the commons? Karp agrees action is necessary.
- Karp first floats a 50% death tax, then gives a 25%-now example conditioned on government waste identification; the transcript does not make clear how those figures fit together. His general rule is clearer: identify half a unit of waste for each additional unit of taxation. He separates the unfairness of billionaires becoming 100 times wealthier while others gain 5% from the separate task of improving poor people’s lives. Policy must address both.
- When Isaacson brackets wealth assessment as impractical, Karp pushes back: “They’re not impractical with AI.” AI could infer possessions and their values; that is precisely his surveillance objection. He argues a wealth tax would not necessarily help the poor, could break social consensus, and would damage tech culture because wealth is illiquid. He supports people like himself paying more, provided redistribution preserves American tech, avoids expanding surveillance and materially benefits working- and middle-class people.
4. Proprietary data is the enterprise AI sovereignty dispute
- Karp says enterprises are paying heavily for LLM products while providers train models on proprietary data. He says companies fear that data could help produce products competing against them. He reports particular anger at Anthropic, while defending Dario against customers’ caricatures: “He’s not Lord Vader.” Isaacson compares the arrangement to management consulting; Karp’s response is “times 5 million.”
- Palantir’s answer, he says, combines partner alignment, guidance on contracts and open versus closed models, and an ontology: a contextual framework that makes raw inputs useful. It is an “alpha creator” and a “protective layer.” Sovereignty also applies to military customers: the Department of War, Ukraine and Israel have datasets and battlefield methods that, he argues, nobody else should understand.
5. Silicon Valley lost the connection between wealth and national purpose
- Isaacson challenges technology companies’ reluctance to serve defense, noting Silicon Valley’s military origins. Karp agrees, describing the original Stanford-government-venture-capital hybrid. He calls the U.S. military the only institution a supermajority considers indisputably noble, and credits Elon, Palantir and Palantir graduates with helping restore government’s appeal in defense-related work.
- Karp’s broken bargain: wealth was tolerated because it accompanied public value. Consumer and advertising businesses create value, he concedes, but a Louisiana carpenter may see little improvement over 30 years while others become 100 or 200 times wealthier. He also argues that venture returns of 100 or 1,000 times help finance misfits such as Palantir.
- AI means “dumping gasoline on every part of it.” Karp criticizes lab leaders’ messaging that workers will lose jobs and depend on wealthy founders. Isaacson identifies two lost promises: better everyday lives and national ambition. Karp links them: common purpose requires recognizable benefits today and tomorrow, rather than “building a dream that you don’t belong to.”
6. Human worth must survive machines surpassing human intelligence
- Karp grants legitimate reasons for government mistrust: technology can undermine rights, COVID was mismanaged in some form, national purpose is weak and successful elites seem detached. His correction to the conventional right is explicit: capitalism is “better than the alternatives,” not a solution to every problem. He says its negative externalities include wealth dislocations that distort political power and even make human life’s dollar value depend on net worth, as in differing legal damages.
- When Isaacson asks what remains special once machines surpass human cognition, Karp invokes rights grounded beyond cognitive performance. He argues America’s constitutional tradition protects a human spirit that is artistic, individual and particular. He calls that arrangement “unique to this country” and connects preserving it to skepticism about open migration.
- Isaacson’s closing judgment preserves the disagreement. He finds Karp sincere about putting technology in service of the republic, yet remains unsure whether someone so rational and comfortable with difficult trade-offs fully appreciates public fear of the surveillance state.
Full transcript
The Builders is a production of Kaleidoscope and iHeart, presented by Ramp. Ramp is a financial operations platform that helps ambitious companies manage their finances so their teams can spend less time chasing receipts and approvals and more time bringing extraordinary ideas to life. I'm Walter Isaacson, and welcome to The Builders, my series of conversations with the most consequential minds shaping our future.
In the years after September 11th, Peter Thiel began thinking about a new problem. At PayPal, he and his colleagues had developed software that could identify fraud by finding patterns hidden across enormous amounts of information. Could a similar system help intelligence agencies discover threats before another attack? The idea became Palantir.
Palantir is a leader in analyzing data, a skill that’s at the heart of the economy and surveillance.
Demonstrators are accusing the data analysis firm Palantir of working with ICE.
To lead the company, Thiel turned to an old friend from Stanford Law School, Alex Karp.
Palantir’s leader, Alex Karp, is an intellectual known for his big hair and bigger personality.
Having Karp, a self-described socialist, running this company gave people some assurance that the company could be trusted.
Karp was an unusual choice to run a technology company. While others were building code, he spent his 20s studying neoclassical social theory and philosophy in Frankfurt.
Karp had no background in computer science or business when he joined Palantir in the early 2000s, and yet, 20-plus years later, he has turned this company into a $400 billion colossus.
Alex Karp occupies an unusual intersection: philosophy and warfare, Silicon Valley innovation and national security. So I wanted to ask him: Who should be trusted with systems this powerful? Can artificial intelligence make governments and militaries more effective without making them less accountable? How do we defend an open society without building the machinery of a surveillance state? Here’s my conversation with Alex Karp.
Alex Karp, welcome to The Builders.
Very nice to be here. A huge honor.
1. Karp’s Family Origins
You know, one of my theories in biography is that it’s all about Dad. So tell me first about your father.
Well, my dad is a very smart person, very shaped by the combination of his German-Jewish mother passing away very early. I think my parents were very much in love. But it was not a coincidence that my dad married a jarringly, excruciatingly talented, somewhat difficult Black woman. I think he loved the fact that it poked his stepmother, who was more status-conscious, in the eye.
My mom is enormously captivating in her level of talent—artistic talent. The thing about my parents—they hate it when I say this—is that both my parents should be more famous. It’s kind of weird that I’m the most famous one. Had they been more of what I think they could have been, they would not have birthed me.
I had this discomfort about the fact that this family, which exploded in a messy divorce and had complicated parents, created an artifact that is ideally morphed for building, essentially, Palantir. If you take someone like me and put me in this chaotic environment, where you had the feeling you were on your own and were subconsciously or partially consciously misaligned with your parents, you can get a deeper trust in your own ability to survive, but maybe a latent mistrust of structures.
2. Dyslexia as an Advantage
Tell me about dyslexia. How did being dyslexic give you wings to fly?
You always have to walk back from where you are to see where it got you. I was in the closet about being dyslexic until about 6 years ago. Then I got tired. Honestly, I got frustrated. I told a buddy of mine, and from there I just started talking about it.
It’s interesting for me to watch how people derided everything we did, derided everything I said, and now basically copy it verbatim. Why does that happen? It happens because they’re looking for a playbook. If you’re dyslexic, the way I define it—because it’s a broad term—is having a non-one-to-one relationship with the taxonomy of a language, the written taxonomy.
Even now, if you read and I read, there’s more of a translation function, even though I did a PhD and still do a lot of the writing at Palantir. Why is that important? Of course, you’d prefer not to be dyslexic, but the truth is, if you are dyslexic, you’re not colonized by the broader culture. When I’m interacting with the broader culture, a tech problem, or any problem, I’m interacting with the problem as it is.
That’s how we were able to build the anti-terror product. That product changed the history of Europe. We’ve stopped hundreds of terror attacks, some of which would have happened in other places, but many of which would have happened in Europe. Think of the Europe you would have if that had happened. Later, our product became fundamental to the defense of Ukraine.
Well, wait. Explain to me how that comes from being not colonized.
First of all, if you went to school to learn how to build a software company, you would have learned to build a product that isn’t very valuable to the client, but that they’re addicted to. Then you hire lots of salespeople and get lots of people to buy it against their will. That’s how you build software.
The derivative of that is a simple but addicting product. It doesn’t really create value, but it’s hard to get rid of. It’s like a venereal disease: easy to get, maybe hard to get rid of. What you had to avoid at all costs was anything that looked like labor. We spent 3 years building a product that was valuable.
Okay.
Then you get to the question: What is value? We said value is both stopping the terror attacks and protecting civil liberties.
Okay.
Already, you had 2 factions: one that wanted to stop terror attacks and one that wanted to protect civil liberties. So, to build that product, you had to violate every single norm of how you build software. The biggest norm we violated was that it took roughly 3 years without revenue. Three years, no revenue.
You would have to be completely insane to do that. I do think it took a neurodivergent person at the helm to fight for that. People often make these comments to me in private—business executives will ask, “Can you teach me how to be unpopular?” We’re oddly kind of popular right now. But I don’t think it’s possible, because what they’re really asking me to teach them is to be dyslexic: to look at the material, digest it, but it’s not inside me.
3. The Outsider’s Political Lens
You clearly grew up feeling a bit of an outsider. Another theory I have about biography is that being an outsider can be a superpower, in some ways, in the sense that you aren’t trying to fit in.
Or maybe you’ve tried to fit in, but it’s failed.
Did you try?
No, but I think human beings as a species want to fit in. And I think most of us who are not trying subconsciously realize it would have failed to try.
To what extent does that help you, as opposed to somebody like me who feels, “Oh, I’ve always fit in. Good childhood. Want to get along”?
Yeah, Jewish guy in New Orleans. I’m not sure that’s true, because you would not be able to write the things you’ve written if you were purely an insider. In any case, it’s not me interviewing you. If I were interviewing you, I’d push back on that statement.
No, I think being an outsider is essential to understanding Palantir or my life story. It’s very unlikely you’re going to get it as an insider. I think the inside track is overworn and not working. And if you link it to what we’ve built at Palantir—the culture, the aesthetics, the products—there’s not a single thing we’ve done that an insider would ever do. Most outsiders are pretending to be outsiders. Since there’s so much value in being an outsider, you have lots of people who are fake outsiders.
But that’s not you.
Well, I think people like me would love to go on an insider vacation. Give me the 8 weeks a year where I can be an insider. Again, I know it would be totally stupid, but I want to be reborn as a—my reborn fantasy is: Catholic, Bavarian, veterinarian.
Well, getting to Bavaria and to a related concept, which is a collaborationist instinct in the human species. You wrote about the Nazis when you were in graduate school, was it?
I wrote a critique of Heidegger. For better or worse, it would be hard to say Heidegger was not one of the 2 or 3 most important philosophers in the last couple hundred years, along with other people, maybe Wittgenstein. Heidegger developed a form of phenomenology, which is a fancy way of explaining how things in the world could impute value onto the world. There are many different forms.
His advisor was Husserl, who was Jewish. Heidegger became an ardent supporter of Hitler, even though his mistress, Hannah Arendt, was Jewish, and his advisor was Jewish. I think Heideggerian philosophy is the most important political philosophy in the West now, for worse, because most political movements that I would call regressive far-right—meaning they want to regress to something that obviously has never worked—on the far right and on the far left, pretend to be Marxist, certainly on the left, but are actually Heideggerian.
That’s why I’ve been very early and, for years, have been saying, “I’m still an unhappy Democrat.” I’ve been warning the Democrats that a faction of their party is actually not progressive.
They call themselves progressive, but then they want a form of communitarian economy that looks futile. It’s not even actually socialist. And they pretend to be isolationist when they’re actually just focused on the groups they’re uncomfortable with. That’s a right-wing party.
In any case, I wrote about Heidegger. The truth is, I also wanted to get out of jail. Halfway through my dissertation, I realized this wasn’t for me. I needed to go and build something. To your point, I’m a builder.
The interesting thing about builders is that they’re elliptical. The people you’ve written about—Steve Jobs, Elon Musk—I think you’ll find in every case a combination of an outsider with an aesthetic proclivity that’s as much humanities as tech. I think, objectively, although most people in tech would disagree, it’s more aesthetic world importance and business acumen than tech. That’s why tech companies are built here and not, unfortunately, in Germany or France. It’s not that we’re smarter, but we are producing that shape of person. And part of it is that most of the people building things here—or their great-grandparents, parents, or they themselves—opted to come here. That’s already a big step.
4. Why Karp Built Palantir
Why did you found Palantir? What was the point of it?
My version of why I founded it is that I knew Peter in law school.
Peter Thiel.
Peter Thiel. We were in the same section and the same dorm, which was the dorm you were in if you didn’t have a lot of money. We were both intellectually very curious. I was doing no work; Peter, I think, was doing much more work than I was.
He came to me and said, “We have this idea to bring a methodology that was used at PayPal to the intel space,” because it had worked to cut down cyber fraud, and they obviously weren’t good at that kind of tech. I thought it was wildly interesting. Peter is wildly compelling, and I just said yes.
I joined because I thought the idea was interesting. You can take a jingoistic version of America, like everything. Of course, we’ve made mistakes, and some of those mistakes are hard to understand in retrospect: mistakes around integration and respect, slavery, not paying enough attention to class, and not having a stronger national project. There’s a long list. But in my view, and I think in the view of most Americans, there’s no project at scale. So you can’t really compare America to, say, Sweden in the ’70s, which was a completely different, completely homogeneous culture.
Are you saying the diversity of America makes a difference?
The size? It’s the size and grandeur of our project, and the fact that we have this Constitution that is still fresh. The First Amendment, in my view, the Second Amendment, and the Fourth Amendment—so, right, free speech, which is basically also a stand-in for the freedom to decide your own religion, which is very radical.
The Second Amendment: I don’t believe we should allow plastic .50-caliber guns roaming through the streets. But I do believe it’s crucial to allow citizens the right to bear arms, because it’s often reframed as self-defense, which I’m also in favor of, but it’s really about putting the government on its back foot and mandating what official policies would be.
Well, the Fourth Amendment pushes you back a bit.
I know, and I’m going to get to that. I just want to underscore that I’ve never been a neocon. I never thought it made sense to make the Muslim world like us. We have many Muslim country partners, and it runs very well. Of course, I want a different kind of government. As much as I’m on the other side of the CCP, I want America to win more than I’m interested in how they run their world. I’ve never been of that school.
We have the ability to say whatever I want here, and there’s no government sanction for what I say. The Second Amendment does not exist, to my knowledge, anywhere else in the world. The First and Second Amendments are still enormously radical. But as necessary as the First and Second Amendments are, they’re not sufficient if you don’t protect people’s rights to privacy.
You could say, “Well, the Fourth Amendment—you don’t need the Fourth Amendment to protect your right to thought.” People would say that abroad, in many countries. But it’s not true, because we’re language-driven beings. If you can’t articulate your thought in some form, even as a dyslexic, is it a thought? What is a thought that can’t be articulated? Can a thought be extended, manipulated, changed productively or non-productively without its articulation in the thing—the formative thing—that makes us human and has allowed us to advance?
Spoken like somebody who’s read a lot of Heidegger and Wittgenstein.
And where it’s true. This is why the Fourth Amendment, the ability to protect privacy, is so crucial. One of the most vile and stupid attacks on Palantir has been that we’re somehow anti-Fourth Amendment. No, our products are actually built around protecting these rights. The only way you could use AI now is with one of our products, where you protect these rights.
They’re vital to what makes us human. That’s exactly why regressive political movements go after the Fourth Amendment. They want to eviscerate these things because their retrograde ideology would force everyone to be standardized, cut out, the same—non-neurodivergent, non-sovereign, non-individual. That’s why they cannot stand America or the way it’s set up: our principles are set up to build builders. One of the major reasons we have a tech scene unlike any other country is that our Constitution is built for builders.
5. The Fourth Amendment Debate
We’re talking about the Fourth Amendment, the right to privacy. So many people would talk about Palantir, the cameras gathering information on us. Doesn’t that impinge on our right to privacy, or do we over-index on privacy?
No. One of the things I think Europe got right was that the European constitutional order focuses on privacy in both the commercial and government contexts. Our Constitution protects against government overreach. Culturally, I think we’re not sensitive enough to commercial overreach.
All my students say they don’t mind Google knowing everything about them, but they don’t want the government to know everything about them, which is odd.
I’m in favor of them not wanting the government to know everything about them. They’re underestimating what they’re often imputing: “I don’t care what commercial companies know about me because that’s my choice,” which, by the way, it is, obviously.
But they’re thinking of their opinions as wholly their own. At best, your opinions are a dialogue between your ideas and the outside consensus view. They’re underestimating how much their thoughts are being influenced by the tech products they’re interacting with. Because they underestimate that, they don’t understand why they should be more concerned about privacy.
I’m sympathetic to some of our critics of Palantir. They’ve got the Palantir part wrong. You’ll see things like “Palantir surveillance state,” blah, blah, blah. The Palantir stuff is technically and in practice the inverse of what they think, by the most extreme amount of any other tech company. The part they’re actually right about is that if you have companies controlling all data, they will control what you think, or at least heavily influence what you think.
But isn’t Palantir part of a surveillance culture? There’s an upside to that, but also maybe a loss of autonomy?
Of course, you have to get a definition of what we mean by surveillance. If you’re an enemy of the U.S. shooting at the U.S., planning to shoot at the U.S., or trying to stop people who are coming into your country to dismantle something you’re doing that’s nefarious and bad, okay, there are a lot of people who don’t agree, but okay. Yes, our products are used for that.
But what about when I’m walking down the streets of New Orleans and you have a system there that says, “Okay, here’s where Walter was”?
Domestic surveillance is where the issue comes in. In a weird way, Palantir is used heavily in France and in other places in Europe where we’re contractually not allowed to mention that. There’s a difference between surveillance under the context of a legal mandate and surveillance where there’s no legal mandate. This distinction is politically, morally, and technically important.
Most people get the political morality. But it’s very hard to build a product that, for example, in the European context, can only keep data where there’s a warrant, and then purge it when the warrant expires. The hard part is: you have a warrant, the warrant disappears—what happens to the data? Our products allow you to work in conformity with that.
Some people say that’s still surveillance: you’re surveilling someone the courts have decided is a criminal. I think there you get into this: you have people who are so pro—what I would call the right-wing version caricatured—where they’re like, “Just kill the terrorists and criminals.” Okay, that’s not a society most of us want to live in. You have the left-wing version, which I think just brings the right: “No, we’re not going to surveil known hardened criminals.” Our product is used most effectively in Europe on criminal gangs, professional criminals, meaning they’re organized and trained.
I imagine on border enforcement, because it’s all done by professional criminals to get people in, and on anti-terrorism, in terrorism cases. There, in the case of domestic intelligence, you have essentially something like probable cause and a construct that allows you to do it legally. So, in practice, it’s a form of law enforcement that I think 90% of society supports. I would say the 10% that doesn’t is basically going to get us some form of fascism.
If the terrorist attacks in Europe that we had prevented had happened, you would have some form of far-left or far-right fascism. It would amount to basically the same thing. You are tracked all the time.
Let me broaden it a bit, more philosophically—not just to Palantir, but to the technology we have today. It has enormous data-gathering capabilities: cameras, visuals, facial recognition. And now, with AI, it has an enormous ability to draw things from that data. Do you think this whole explosion of data collection, combined with AI, in some ways impinges on personal autonomy?
Well, look, obviously it can, and it will if you don’t push against it. Now we get into the AI debate. AI is, in my view, best conceived as a raw material that can be processed for good or bad, broadly speaking. One of the most negative ways you could harness these powers is for surveillance. I would argue that most of that is going to happen in the commercial context, not the government context, but it could happen in the government context.
I’m very in favor of actual restrictions on some of these technologies. For example, if you take the AI that we use on the battlefield, we should have legislation that prevents that technology from being used domestically. We need it on the battlefield. We have no choice, because our adversaries are building stuff, and if we don’t build it, we’re going to be demolished.
I would say that if you’re anti-war, in my humble opinion, looking at any part of history, you have to have a strong military. The idea that you’re going to weaken the military and have more peace is a fantasy of people who are half idiots, or people who actually want war. But the domestic part that you’re flagging is a very real issue. What are the technologies we’re going to unleash?
The famous thing about fascist countries is that they have no crime—no crime. When the Nazis had power, the only crimes that were committed were by them. That’s how it works. There is only one criminal band: the government. There will be no other criminal band, and they will take people out and shoot them, and lo and behold, there’s no crime. To do that, they shoot a lot of innocent people, too. Criminals are rational actors, and probably they’re the people who get shot the least. The people who get shot the most are people with political views like me, honestly.
That ties this back to my outsider thing. From the time I can remember thinking, I always thought I would be the first person against the wall. I believe I will be the first or second person against the wall if we have no crime. The first thing they do is define you as one of the people with an actual view. I’m talking about things I say that are outside the norm and are transgressive, building a company, or having too much money. If you don’t protect the rights of the people you don’t like—the unapologetically, unambiguously opinionated people who also have money—you’re going to end up with a country run by, and heavily influenced by, the extremes.
6. AI and the Wealth Divide
Take the wealth tax. There are class inequities in society that have to be addressed, and AI is going to exacerbate them. I believe AI will create more jobs or higher-quality jobs, especially for working-class vocational people, and many people are going to get retrained. But even if their salaries go up twice the rate of inflation, they’re unlikely to keep pace with people in tech and AI, because those people could be 10, 20, 50 times wealthier.
Okay.
Then the reaction to this is to impose something that just hurts the wealthy; it doesn’t help the poor. And, by the way, it’s going to hurt the middle class for lots of reasons.
Well, wait a minute. If you agree that AI is going to lead to greater disparities of wealth, then go back in history: in the 1890s, we had just as big a tech boom, which led to great disparities of wealth. Isn’t there some reason to say, “Let’s put more in the commons”? Let’s reduce the disparity of wealth?
That’s what I’m saying. You have to do something. What do you do? My version would be to say you have a 50% death tax. I would go to people and say, “Okay, for every point you pay in, you can have some point.” You could pay 25% now, but the government can only collect by doing half as much waste identification.
The legitimate critique I agree with, which most people agree with, is that you can’t have people becoming 100 times wealthier when they’re already billionaires and everyone else is becoming 5% wealthier. That’s legitimate. The legitimate critique from people who are economically literate is, “Yes, but this isn’t really going to help people. Half of it is being wasted.” So the way I would propose it is: great, you have to tackle both. It’s the same thing as having to protect civil liberties and fight terrorists.
The other problem with the wealth tax, which for some reason people haven’t talked about, is that to enforce a wealth tax, you’re going to need surveillance. I’m going to need to know: What is the bench in your house in New Orleans worth?
Let’s leave aside the practicalities, which seem very impractical: assessing each person’s wealth and taxing it.
Sorry, let me just push back on this one point. They’re not impractical with AI. It’s exactly the point you were making. They’re not impractical with AI. AI could impute every single thing you own, especially with all the data about it. It can impute the things—
Well, then what’s wrong with taxing people’s wealth?
Because, first, it doesn’t help the poor. It breaks consensus in society. It breaks the tech culture because people are illiquid. Those have all been talked about. And the logical consequence of it is surveillance into your personal life.
So give me a good Palantir way to say we have too much economic disparity. We want to have more common ground in our society. How do you do it, besides maybe more government efficiency being the trade-off?
You basically make a trade-off where taxation goes up. I gave you my version, but under any other version besides a wealth tax, taxation goes up on the condition that for every unit of tax you raise, half a unit of waste is identified. That transfers a lot more to the government, because you’re trying to solve 2 problems: inequality, which is a problem, and inequity, which is also a problem—how do you make the life of a poor person—
Better?
They’re not the same problem. One problem is that it’s not fair for me to become 20 or 50 times wealthier and for you to become 5% wealthier. That’s not fair.
Okay.
That’s one problem. It’s not at all correlated to the other problem, which is that neither you nor I are working-class people in the city. That’s a problem that has to be solved, too.
The interesting thing about the proposals, mostly from the left here but also, honestly, from the far right, is that they only care about problem one. When a proposal is focused on only one side—whether it’s civil liberties or anti-terror, or surveillance and AI without considering the good and the bad—the problem is, it’s always both. AI is both a force for good and bad.
The reason we have to do it is because we have adversaries, and our adversaries are going to do this. If they own this technology, they will change how the world works. That’s not good for humanity. It’s not good for us. It’s the same thing with surveillance. Does anti-surveillance mean we can’t have a border? Does it mean we can’t stop trained, hardened terrorist networks? Does it mean we can’t identify spies who are stealing our IP—trained adversaries who’ve gone to school and are at the best spy service in the world? They’re stealing our stuff. Can we not stop them? I can tell you that isn’t going to work without technology. You’re going to need to know how to do this.
Okay.
And then you have the same thing with wealth. How do you do this so you actually have a fair society, where wealthy people aren’t becoming 100 times wealthier and the delta isn’t so large? How do you do it as an addendum, which is not talked about enough, where you don’t expand the surveillance state? And how do you do it, last but I think most importantly, in a way no one seems to care about: does taking wealth from the wealthy actually help the poor in any material way? Somehow no one cares about that.
These things are very much linked. I’m in favor of people like me paying a greater share, under the condition that it doesn’t destroy American tech, which a wealth tax would do; doesn’t expand the surveillance state, which it absolutely would do; and lands in a way that helps the working class, the upper middle class, and below. If you pick one, you just divide society. The people who pay for that division are the people who have no voice. De facto, you have no voice in society if you’re actually poor, because the people repping you are interested in killing the wealthy. They’re not structurally interested in helping the poor.
Let’s talk more about data training and information surveillance. In some ways, you’ve been talking very strongly about digital sovereignty now—that the big LLM companies are not only sucking up all our data and training on it, they’re doing it without our permission.
They’re building on our backs. Well, U.S.
Enterprises are livid, enraged, apoplectic. Because what they’ve discovered is they’re paying a lot of money for a product that hasn’t created a lot of value.
You’re talking about LLMs at the moment?
Yeah, LLMs. And then, lo and behold, you find that while you’re paying them a lot of money, you’re actually paying for them to train the model. How are they training the model? They train the model on your data, which is the most valuable data because it’s proprietary. De facto, this is what happens if you use any of these products.
Sounds in some ways like the old McKinsey & Company management consultants.
Yeah, but think of it times 5 million, and then it actually shows up in your business, and your business is theirs. They’re particularly angry at Anthropic. It’s funny. I feel like I spend 10% of my day explaining to enterprises, “Dario is not the caricature that you think. He’s not Lord Vader.” And they ask, “Well, if they’re so ethical, how can they take my data and turn it into a product that’s competing against me while I’m paying for it?”
So how does Palantir fight against that?
Well, first of all, the secret of Palantir that’s most obvious and hardest to replicate is that we actually are aligned with the interests of our partners. Mostly, we’re educating. People ask us, “How do we do this? What do the contracts look like? How do we interact with closed models, open models?” We’re publishing document after document, not just for our clients and partners.
With our clients and partners, they use what’s broadly called an ontology now, a concept that we developed that was also derided and that now everybody talks about. You can think of an ontology as a contextual framework, where that contextual framework is de facto what you’re doing. You’re taking random things I’m saying and putting them into a frame where somebody would find this interesting and intelligible.
Basically, the ontology is both an alpha creator, meaning it takes these raw products and makes them valuable, and a protective layer, which means you can interact with it without being burned. But the larger point of the sovereignty thing is that we believe in sovereign institutions and sovereign individuals. How do you apply that? Well, that’s complicated, but in the end, whether it’s vis-à-vis the consumer or vis-à-vis the government, there should still be an imputed version of a First Amendment and a Fourth Amendment in your life.
Then there’s a national version, like the stuff we’re doing at the Department of War, which is very powerful, or the stuff that’s being used in Ukraine or in Israel. They have ways of doing things, datasets, ways of understanding the data, and ways of implementing it on the battlefield that no one else should understand.
Do you think, by the way, that more American companies should be willing to enlist in our military and national ambitions? And what do you think when Anthropic, or even Google, or people say, “No, no, no, we don’t want to be a part of this”?
Silicon Valley is a wondrous, unique place the world is jealous of, for good reason. It’s a highly insulated island. The only institution that a supermajority of Americans think is indisputably noble in this country is the U.S. military.
And the Valley was born out—
Of the U.S.
—military.
Well, it’s a hypocrisy. It would not even exist without the U.S. military. Thank you for reminding everyone of this. The Valley was built as a hybrid: basically, university, government, capital.
A Vannevar Bush model. Yeah.
It was mostly Stanford, the government, various departments, and people who came up with the venture model.
And so what happened?
Well, a lot happened, and we made a lot of mistakes. I do think the government lost its coolness. I think Elon, Palantir, and others—Palantir graduates—have helped bring that back in the context of defense, defense tech, and defense-tech-adjacent work.
But I think a wave entered this country that’s unfortunate. One of the ways you can signal to your community that you’re smart is by being craven, meaning you make other people feel smart while you transfer their IP to you. That model has been very, very successful, and the unit economics of it are very good—unusually good. If you’re a venture capitalist or you’re involved in this, you can make 100 or 1,000 times your money.
By the way, that also leads to things that are legitimately good. If there wasn’t so much wealth creation in the Valley, the bastard misfits—Palantir being the best example of this—would never be able to raise money. We raise money on the back of the fact that everyone’s getting wealthy, but we are the last people to get the money.
Go back to this notion of technology engineering being in service of a greater good.
But by breaking the explicit, or definitely tacit, covenant between wealth creation and creation of value, you break other tacit covenants in society. You get to get wealthy, and we’ll leave you alone with your vast wealth because you’re also creating value in society. If you break that, you end up with a fracturing of society.
So show me how we broke it. In other words, was the wealth creation done on frivolous, dumb things that didn’t add a lot of value?
I think these things do create value. But if you’re the average person—say you’re sitting in Louisiana and you’re a carpenter—do you feel like your life is better, materially better, than it was 30 years ago? I don’t think so. So, right or wrong, I think the perception is more right than wrong: a small group of people have gotten 100 or 200 times wealthier, and your life hasn’t changed that much.
And it’s based on consumerism and ads.
Consumerism, ads. But in the end, these things become micro-subcultures, and they’re very divisive. Everyone in that microculture is in it because they’re the same. But the differences that get you kicked out of the culture get smaller and smaller and smaller.
Witness how many Democrats refused to accept me as a Democrat because I thought they were part of the party that was going nuts. But we’re struggling to find commonalities. Okay, so that’s the backdrop. Then AI comes, and you’re taking a fire that’s already burning and dumping gasoline on every part of it.
The people who are running these labs are saying, “Your life is going to suck.” In a normal political movement, you’d have the people doing the movement saying, “Hey, this is going to be good.” No, you have the people in the movement saying, “Your house is going to suck. You’re going to have no job. And when you have no job, you can rely on me to give you money. Oh, but by the way, I’m going to be worth it.”
Well, it seems we’ve actually lost 2 things: one, the idea that your life will be much better, that we’re not just creating some consumerism- and ad-driven stuff; and secondly, a national ambition, a national greatness. Both those things seem to have been lost.
And they’re very linked, because if you have a national ambition, you have to talk about things that will actually make your life better today and tomorrow—not about how bad your life is going to be, but about building a dream that you don’t belong to. They’re linked because part of the reason the Valley goes so wrong on these things is that they’re not talking to anyone in America, and they don’t care. It’s an island. On that island, you’re cool.
Let’s talk about the anti-government sentiment, especially coming out of the ’60s and ’70s in Silicon Valley. But there’s a strong anti-government sentiment on the right now. Where’s that come from, and is that dangerous?
Well, again, we at Palantir are always interested in what the legitimate version of maybe a retrograde argument is. Let me start with this: If you have a rise of a powerful technology that could be used to abuse you, to undermine the imputed rights that you actually have under our Constitution, directly or indirectly, and you’ve seen mismanagement of issues—call it COVID, which I think most people would say in some form or another was mismanaged—and you have no national purpose, maybe the lowest I’ve ever seen, then you have high mistrust in government.
You have low national purpose. You feel like people who claim to believe in things are just liars. And the people who are doing the best seem not to be so tethered to you. You’d be a little bit weird if you didn’t have skepticism of the government.
Where I think it’s very nonproductive, again, is on the right and on the left. But, just for clarity, I think the right you’re referring to and the left I’m referring to are not actually right. I think that right is actually left and the left is actually right. That’s one of the reasons fascist movements tend to do this. It’s somehow very charismatic.
Did you learn that in Germany?
Well, I spent a lot of time with Nazis, ex-Nazis, when I was in Germany because I wanted to understand real ex-Nazis. You had to go to the countryside to meet them, because in the city, there were no ex-Nazis. Everybody was like, “We hated Hitler.” Just in the country, people were honest about it.
I actually think anti-Semitism was a motivation, but honestly, for the record, I think a lot of the motivation was, “That’s not our battle,” rather than overt anti-Semitism. It was the charisma of Hitler to them and the national abuse of a national mandate, which is justified.
And abuse of patriotism, which is necessary to humans. You see this in the reactionary left and the reactionary right. Why are they very different movements? Why are they so charismatic to their peers? It’s because the right has been indoctrinated with “capitalism is always good,” which is simplistic. The correct narrative is that it’s better than the alternatives. That’s very different from saying it solves all problems.
Classic right-of-center institutions sell capitalism as if it solves all problems, when that’s obviously wrong. It solves problems that other systems don’t solve, and it solves them in ways those systems don’t. Therefore, it’s better, but it has negative externalities.
Such as?
Egregious wealth dislocations that obviously impinge on political power, and a dislocation of societies and narratives, because you’re not living in the same world if you have a billion dollars or $100 in your account. Capitalism inherently assigns a dollar value to human life, and that dollar value is a derivative of your net worth. If you get hit by a car and go to court, the driver who hits you will presumably pay a different sum depending on whether it’s you, a working-class person, an unemployed person, or me. So it inherently does that.
Let me end with a question about what you just talked about: What does it mean to be human? You look at Creech Air Force Base, where they now have drones, and you’ve worked with that quite a bit. Humans are almost not in the loop anymore. That seems like a metaphor for what we’re facing with this new technology. To what extent do humans not only have to be in the loop on things like systems, but what is the role of humans when there’s intelligence, and maybe even consciousness, in machines?
What is special about humans when our core capacity and our claim to superiority have been our cognitive intelligence? Yeah, but what happens when we birth a species where that’s no longer true? This is where I again think the Founding Fathers of our country were people with something beyond genius-level wisdom. You have a different setup in Europe, where the constitution is birthed by parliamentarians. Our rights, according to our constitution in this country, are given to us because of a grant and the presupposition of a higher figure.
That means there is something in the human spirit, protected under law and internalized by Americans. It’s a reason why I’m very skeptical of open migration, because it’s very special. We believe deep in our hearts, because of what we were given in this country, that there is something in us above and beyond just our ability to speak and our cognitive abilities, which at some point will be surpassed. It’s something called human, which is artistic, individual, particular, special, more valuable, and needs to be protected. That setup, for lack of a better word, is unique to this country. We should make sure we fight to keep it.
Alex Karp, thank you so much.
Thank you.
Thank you. Alex Karp is hard to pin down. In some ways, he’s grouped with the other tech bros who are on the right or libertarian, but he’s not that way. In many ways, his philosophy and his politics are unpredictable, but he has a strong sense that when we build new machinery and new technology, it’s got to be in service of our republic, in service of our nation, in service of our democracy. That’s something that hadn’t really been part of the ethos of Silicon Valley for the past 50 years.
It’s clear to me that Alex Karp is different, and I think he’s being sincere. But what I keep coming back to is the trade-off that people like Karp force on us. He’s intellectually honest about surveillance, about Palantir, about the government using these tools. But the thing I’m left wondering is whether a man this rational, this comfortable with trade-offs the rest of us might flinch at, fully appreciates the fears people have about the surveillance state.
That’s it for today’s episode. I’m Walter Isaacson. I’ll see you next time on The Builders.
The Builders is a Kaleidoscope and iHeart production. This series is hosted by Walter Isaacson. It’s executive produced for Kaleidoscope by Oswald Lishen, David Weisbord, and Mangesh Hetikadur. Sheena Ozaki is the producer. Julia Nutter is supervising producer. Eva Zorich is our consulting producer. Fact-checking by Eamon Whalen. Research by Gloria Thiel. Maria Paz Mendez-Hodes is head of video. Marketing by Christy Mirabal and Elle Liu. Production support by Dara Look Potts. Music for the series was created by Spencer Stevenson, a.k.a.
Botany, and it was.
Mixed by Steve Bone. From iHeart, the series is EP’d by Alistair Perry and Katrina Norvell. Special thanks to Bob Pittman, Conal Byrne, Will Pearson, John Sykes, Nikki Etorre, Carrie Lieberman, Binky Urban, Julia Prosser, Jeremy Blum, Alison Goldberger, WYES, and ASL Studios. For more from Walter Isaacson, be sure to check out his newest paperback, The Innovators, How a Group of Hackers, Geniuses, and Geeks Created the Digital Revolution. It’s available wherever fine books are sold. And for more videos and podcasts from the Frontiers of Discovery, be sure to check out kaleidoscope.nyc.