Joe Lonsdale — co-founder of Palantir, Addepar, and managing partner at 8VC — joins Shaan for a wide-ranging conversation recorded after a morning cold plunge. Joe shares frameworks from Peter Thiel (blended reasons, power law thinking, talent over expertise), walks through the founding of Palantir and Addepar, explains how he identifies conceptual gaps to build new companies, and makes the case that unfair advantages compound with each win. The episode ends with Joe’s answer to the one question that matters: what separates founders who build multiple billion-dollar companies from everyone else.
Speakers: Shaan Puri (host), Sam Parr (host), Joe Lonsdale (guest, co-founder Palantir, Addepar, 8VC)
Introduction and Blended Reasons Fallacy [00:00:00]
Shaan: Okay, Joe, thanks for doing this, man. And thanks for inviting me into the morning workout, the cold plunge, and everything in between. Good to have you here.
Joe: Good to have you here, Shaan.
Shaan: So I want to start with lessons learned. One of them we were talking about this morning during our workout was making decisions on blended reasons — having a blended strategy where you say “we’re going to win by doing these five things.” Why does that sound like a good idea to the average person but is actually a very bad idea? Can you explain that?
Joe: Yeah. I wrote a piece online — nine lessons from Peter Thiel — that I sent to all of Palantir and all my other companies. It’s one of my favorite pieces, and that’s one of the lessons in it.
It’s one of those logic things where usually there’s one reason that dominates everything else. That’s just the nature of the world. When something’s successful, it’s kind of like the power law rule — even in a venture capital fund, there tends to be one thing that’s successful that’s bigger than everything else. And that’s true inside companies too.
It’s extraordinarily unusual for there to be two similarly equal products or revenue lines, especially early on. So usually if you think you have two or three reasons for doing something — or five reasons — what that means is you actually haven’t figured out the actual reason. You have no strong reason yet. Five weak reasons is not good. You want to find one really strong thing you’re driving toward. That’s how startups work.
Shaan: It’s a seductive trap.
Joe: It is. First of all, you have to be taking big risks to even do a startup. Most people do incubators because they’re like, “Oh, I don’t want to do one company, I’ll do ten.” But that’s much less likely to produce a big success because you’re not focused. And then even within a company, people do the same thing — “here’s five ways I’m going to make money.” No, no, no. If you really want to do a company, figure out the way you’re going to make money. That tends to be how it works.
Shaan: So eight revenue streams means you probably don’t have one awesome revenue stream.
Joe: Exactly. Growth strategies too — if you have seven things we’re going to do for growth, it means you haven’t figured out which one is actually going to work. The one that’s going to compound like crazy. It may change over time, but what’s the thing that’s working now that you’re pushing on?
The other version of that is blended reasons for doing something. I read once that Reid Hoffman uses the same philosophy. He gave the example: if there’s a trip to China and he’s deciding whether to go — it’s far away, two weeks — and his chief of staff comes in with a list of pros and cons with multiple pros and a couple of cons, that’s a guaranteed way to make mediocre decisions. If you don’t have one key reason to be there, just say no.
Shaan: Exactly. He said if we don’t have one reason that alone is worth doing, we should just say no.
Joe: Is that something you do? Reid was one of the key minds at PayPal, of course, along with Peter Thiel and others. So who knows who actually came up with these ideas — it may all be Reid’s ideas that I’m recycling. He’s a very smart guy.
Shaan: Well, I say them as mine now. The highest compliment is to meme an idea and let it spread.
Recognizing Exceptional Talent Early [00:05:30]
Shaan: One thing I wonder — you’ve been around a bunch of special people. Peter Thiel, Elon Musk. Is it clear in the early days when someone’s different? Do you remember the moment where you’re like, “Oh, this person is five standard deviations out”?
Joe: I mean, we all try to set strategies for ourselves and goals. I remember in college thinking, “This is a really good opportunity at Stanford to reach out to a bunch of successful people — learn from them, see who I admire, who I want to be like.” I met some legends of global macro finance, a bunch of big business people. But Peter had by far the most interesting intellect. He was clearly just one of the smartest people I’d ever met, and he was interested in some of the same things I was interested in. He’d taken some of the logic farther than I ever had. I always learned the most from conversations with him. That to me was one of the reasons he was the most attractive person to try to work with and build with.
Clarium Capital and Global Macro Finance [00:07:00]
Shaan: Before you start multiple other companies, you do Clarium — his global macro fund. What even is global macro?
Joe: It’s the most fun part of finance. Global macro is looking from the highest possible perspective at everything going on in finance — bond yields around the world, the flows of money, how equities are being valued. It’s just a very top-down view of what’s happening in the world of finance and how you map it out.
You look at relationships, understand correlations. The Australian currency is going to be really correlated with the price of metals. When they get way off, you can create a trading system around the reversion. But then you say, “Normally I’d trade this, but China just did X, so it’s not a good time.” It’s a really fun map of the world.
We were really interested in long-dated oil prices. We were very interested in Fannie Mae and Freddie Mac — the two giant housing GSEs that had gotten much bigger in the early 2000s. They had these giant mortgage-backed security portfolios worth trillions of dollars, and how they were hedging those portfolios was completely changing the global fixed income market. We found some really interesting ways to play that.
Shaan: Give me a flavor of what it actually felt like day to day.
Joe: My favorite story is this. We’re working at the hedge fund, market opens at 6:30 San Francisco time, so traders are in at 6:30 — except the first Friday of every month you had to come in at 5:30. Non-farm payrolls come out at 8:30 AM Eastern the first Friday of every month. This is a very important number because it shows what’s going on in job growth across the whole economy.
Kevin Harrington was our head of research — this genius, crazy physics-biology guy who later ended up on the National Security Council. He had mapped out something I helped him with where we realized the Bureau of Labor Statistics was adjusting the numbers wrong based on seasonal adjustments. So we were able to predict whether the number would hit or miss.
This thing would move bond markets by hundreds of billions of dollars. We figured they were systematically mis-doing it — totally legal. We figured, “This is a mistake 75% of the time, we can make money trading this.” We’d take a big position the night before, then make a lot of money in the morning when it worked.
Shaan: And what did the room feel like? Was Peter sweating?
Joe: Everyone was like “what’s going to happen?!” It was so exciting. I remember so many celebrations — we’d all go to our favorite breakfast place afterwards if it worked out. It was pretty cool.
Shaan: What’d you do if it didn’t work out?
Joe: Sometimes we went there anyway. For some reason my mind — maybe this is who I am — I remember all the times it worked. I can barely remember the times it didn’t. I know it didn’t sometimes, but I just see the good memories. Maybe that’s a serial entrepreneur thing.
Shaan: That’s an entrepreneur survival tactic. You’d have never built all these companies if you had to force yourself to think too hard about the traumas.
Joe: Exactly. You have to kind of force yourself to think about that stuff a little bit, too, you know.
Shaan: The way you just described it — you’re pretty young when you’re doing this. That’s not what you were doing just before.
Joe: My little brother got me reading economics and finance stuff when I was young, so I had a weird combination of computer science and finance background and I had opinions about it. Those were like my two things. I knew that world pretty well. And yeah, you learn a lot being in it. You just have to be in it.
Shaan: Credit to them for betting on you. They could have hired somebody who’d been doing it for twenty years.
Joe: We talk about lessons from Peter Thiel, and I think one of the most important ones I’ve taken to heart is: I’d much rather hire for talent and ambition and hard work than for expertise. There are exceptions — if you have a company that’s already really working and scaling, and there’s a machine that has to be run, you want the guy who knows how to run the machine. But when you’re still building the machine, you want raw IQ and hard work and iteration.
How to Filter for Exceptional Talent [00:15:00]
Shaan: What’s your way of filtering for that IQ? How do you interview or what are the signals you look for?
Joe: This is super hard. We used to throw hard math problems at people. We had really hard tests at Clarium. Once I had a bunch of really top talent around me, it’s like they knew people — “who’s the smartest guy in your physics PhD at MIT?” So you get a network that’s already really solid and then you go out from there.
Shaan: Ben told me some number — I don’t know the exact number, twenty or thirty people who’ve worked at your companies have gone on to start unicorn companies. What is that number, roughly?
Joe: I don’t actually keep track of the official number, but there are over twenty people who’ve worked directly for me who’ve started huge companies. And it’s something I’m really proud of. In some ways I like to tell myself, “Oh yeah, I’m a great guy, I shaped them.” But in some ways they were already really amazing and I was maybe just good at convincing them to work with me ahead of time. So there’s a nature-nurture thing.
I do like to think we’ve given them useful frameworks — just like I learned from Peter, they might have learned some things from me too. I’ve backed some of them. Some went off and did it themselves.
Shaan: That’s the worst — when you didn’t get the call.
Joe: Actually, I’m still proud. There was a guy who came right out of school to help at Palantir. He became president of the company — super talented — and I didn’t know what I was doing even though I’d already done Palantir. There were mistakes I was making, things I didn’t realize I had an advantage on. Palantir’s a multibillion-dollar company now, which is great, but there were things that took longer. And this guy eventually partnered with someone and started a company that became worth $3 billion in like four years and he sold it.
Shaan: Wow.
Joe: And I was like, “Oh, that’s pretty freaking cool.” And you know, they didn’t bring anyone else into that company, but I was still really proud that I’d worked with this guy and he went on and had that crazy success. It’s a cool feeling. But for the record, I’m always more proud if I got to invest too.
Shaan: Ha. Come on, take care of your buddies.
Joe: Yeah, 95% of them do bring me in and I’ll help coach and work with them together.
What Is Addepar? [00:20:00]
Shaan: Let’s do big hit number two — Addepar. Explain in plain terms what it is.
Joe: Addepar is the leading wealth management technology company in the US. There are tens of thousands of registered investment advisers — RIAs — and they help you or your parents manage their wealth. They give reports, handle trusts and estates, all the wealth planning. Addepar is basically the software platform that makes all of that work.
It was really freaking hard to build. We hooked up to Schwab and Fidelity — the two biggest custodians — and we had all this wrong data in the reports. We couldn’t figure out what we were doing wrong. Turned out the data coming from Schwab and Fidelity was often wrong. I’m like, “What is wrong with this space? The whole thing’s a mess.”
It was a hard company to build. Took about three or four years before we really got to the point where it was useful to people.
Shaan: You didn’t give up during that time?
Joe: Oh no. That’s the last thing you do. You have to have a personality where you’re not going to give up. I’ll give up if it turns out someone else has already solved it perfectly — if Addepar already existed it would have been a stupid thing to build. But the fact that there’s a hard problem? You just double down and fix it.
The upside of hard problems is that once you do solve them, you’ve got a giant moat. Addepar has the biggest moat. If you want to try to copy it, good luck — you’d need a billion dollars and ten years. No chance.
Shaan: You talk about hard work. I called somebody on your team and they said, “He’s always on. I’ve never seen him just go dark, turn his brain off, just want to chill on the beach and drink tiki drinks and not do anything.” Is that accurate?
Joe: This is my game. You shouldn’t play games you don’t like. If you’re not enjoying the game, you’ve got to switch it up. This is what I’m doing because I really enjoy it. It’s fun. I’m really good at it. Aristotle would talk about how one of the highest pleasures is using your mind at its highest capacity. If you’re good at both chess and checkers, you enjoy chess more — it has more levels, it’s more complex. You want your mind engaged in interesting, complex, useful things that are getting things done. That’s just how humanity works.
Shaan: How do you still recharge or get clarity of thought? One of the challenges with being always on is you might not have the chance to step back and reassess. You need open time for creativity.
Joe: I’ve noticed when I over-schedule myself I’m not nearly as creative. And I think the cold plunge helps. I don’t know about you — mind that was a quick 51-second reset for you. I think you stayed in for the phone call.
Shaan: You did two plunges. That was pretty good.
Daily Schedule and Finding Conceptual Gaps [00:25:00]
Joe: My number one thing is 8VC’s build program, which is my favorite part of it. A third of our money now goes toward building new companies. My favorite thing is the strategy for recruiting talent and then the strategy sessions to figure out what we’re going to build.
Shaan: What do you mean by “strategy for talent”?
Joe: The way I look at opportunities for new businesses — you’re looking for conceptual gaps in the world. Here’s where this industry is now, here’s how this process is done in the economy, and here’s where it could be if it was using the right new technology or the right new incentives or effective cultures or whatever it is that could make it better.
The question is not just identifying the gap — it’s actually very easy to find conceptual gaps in the world. If you spend some time with a few smart friends and study an industry, you’ll usually find, “Wow, this thing could be done this way.” The harder question is: what’s the path to get there? What could we put in place to get it there? How do we build this thing, how much is it going to cost, how do we make the people want it?
There’s a lot of things in healthcare and hospitals that are done completely wrong, but it makes hospitals more money — so even though it’s clearly not the right way of doing things, they’re not going to just fix it. You have to find some way to get from here to there with the actors being willing to do it.
Example: AI-Powered Loading Yards [00:28:00]
Shaan: Give me a concrete example — maybe something you’re thinking about right now, maybe not all figured out. A gap and a path you’ve kind of worked out.
Joe: So it turns out there are tens of thousands of loading yards around the country in logistics. These are places where trucks come, get in line, drop things off, move things around. It’s a pretty interesting area. The biggest warehouse company in the country is Prologis, and the founders’ investor is a friend of mine based in California. There’s also a handful of the biggest carrier companies like Ryder — we actually sold a company to Ryder, so we’re close to those guys.
Because we knew a bunch of these people, we started analyzing: what can we do with AI to make these yards better? What we figured out is that if you put just a couple cameras in — we first thought we needed six, but just a couple — you can watch and map out where everything is, how efficient the gate is, create accountability, incentivize people to go faster, create ways of pricing the carriers if it’s backed up to use other nearby loading yards. Basic little ways to improve the margins by timing things and creating a bit of competition — people always respond well to that if you do it properly.
We ended up co-founding a company with a really strong AI team. We actually co-founded it with the warehouse companies and carriers based on their needs. The really cool thing about how we built it is we have a few hundred million of revenue already built in that we can scale up into their existing yards — and then we’ll have something all the other yards want to use too.
Shaan: Let’s do more detail. Where does the nugget of insight even come from? Take me all the way to the beginning.
Joe: My partner Jake’s in charge of logistics and supply chain — he’s the biggest driver here. He’s done strategy sessions with our friends running these different companies. I have a vineyard in Napa Valley and every year we bring about a hundred top logistics CEOs. We drink and strategize and hang out. You ask them questions to try to find their pain — what are your problems, what are you spending money on, what are you seeing? We bring new entrepreneurs we’re backing, get their take on things.
So loading yards became a big theme. Let’s go talk about what we could do with AI to apply new things to it. In this case there was a really strong AI team in the UK we’d met through other friends, people who’d worked on really cool computer vision problems. We ended up recruiting them as co-founders.
We’re always separately doing napkin math — if we can improve the loading yards by X, that’s the size of the prize, and that clears a hurdle.
Shaan: And California regulations gave you a little tailwind.
Joe: I don’t love depending on that, but it’s definitely helpful. California had some new regulatory rules about things you had to map out anyway. Why not create more value based on something they’re going to have to do in the next year regardless? Logistics has had a few of these — there was a rule on telematics where you had to record things differently for driver hours. That was an insertion point for a couple of fleet-tracking companies we helped back.
And then you sit down, you map it out. You’ve known the people for a long time. One of the unfair secrets of business is: the more success you have, the more it’s a platform for a bigger thing next time. You just have to stay in the game.
I used to think when I was young, “Oh, I could start a business, make a lot of money, go to the beach, that’d be fun.” That’s not nearly as fun as the fact that once you’ve started something successful, you get to do bigger things more easily. That’s why when you’re young, you should partner with someone who already has an unfair advantage. Help them use their unfair advantage, build something with them. Now you have all these extra unfair advantages to do it again.
Every time you have a success you now have the respect, attention, and understanding of all these people who will then help you do the next thing.
Example: Defense Tech and Epirus [00:38:00]
Shaan: Can we do another example? Another gap or origin story?
Joe: I’ll give you a crazy one. I was pretty bullish on China like fifteen years ago. A lot of us were — reading Milton Friedman, free economy stuff. I thought, “This is going to be great.” We started making bets there, doing stuff there. Then Xi Jinping takes over. I’m told by all these senior people there, “Joe, none of the stuff you want to work on here is going to be allowed, it can’t be foreign.”
So I stopped working there as much. Then I start hearing terrible things are happening — a couple of people I know disappear. I hear he’s making all these engineers work on defense. I’m getting kind of nervous.
At the same time — Palantir started in 2003, so this is now 2015-2016 — I’m not running Palantir anymore but I’m still an adviser, watching. And a few of my guys there are noticing that the defense world in the US has gotten a lot more dysfunctional. What happened in defense is all these companies consolidated in the 1990s after the Cold War. Spending went down, they all merged and became these giant primes — and these primes started acting a lot more like government agencies themselves. Not super nimble, not super adaptive. They get contracts by having the right lobbyists, not necessarily by innovating. We call it “innovation theater.” They have to pretend they’re innovating, but they’re not actually great at it because they’re so big with no competition.
So we’re looking at this going: none of the best engineers are going into these primes, their stuff is getting worse, meanwhile China has the best engineers and they’re getting better. We better do something.
Three of my best ex-Palantir guys partnered with Palmer Luckey and started Anduril. I backed them early along with Peter. And then we mapped out: what are twenty new possibilities in defense? We chose the most important one.
One area that looked promising was directed EMP — electromagnetic pulse. What an EMP does is fry electronics. It turns out that if you can direct these pulses — you’re taking microwave energy and shooting it in a blast — you can turn things off. The government had spent billions on this but hadn’t applied any of the new chips. And you want this to protect against drones, for example.
With gallium nitride emitters and the right talent, you can shoot down things ten times farther away. Theoretically you could use it against missiles, turn off their guidance systems. Or against a truck coming fast to your base — blast it, turn off the engine. It might be a terrorist, might be a pregnant woman trying to get to the base for medical help. You don’t want to blow it up. You turn the truck off, you’re not killing the person.
Long story short, with less than $30 million in the second round, we got to the point where we were invited into a contest against the primes.
Shaan: How did you even get into the contest? The primes don’t want competition.
Joe: You push Congress to make laws that require competition. There are competent generals and admirals who, thanks to Palantir and SpaceX, had seen outsiders massively outperforming. They’re more open to the ideas now. We’re pushing things in the right direction.
So they do this contest. They didn’t take us seriously — the other guys had spent billions on this. But long story short, we ended up shooting down hardened drones nine and a half times farther away than the other guys. They had to use binoculars to watch us shooting them down because it was so far out of range.
Shaan: That’s the name of the company — Epirus?
Joe: Epirus. It’s from the legend of Theseus. Theseus started Athens and his bow had infinite arrows. This is kind of like a bow with infinite arrows because you can keep firing the pulses.
Shaan: Palantir was Lord of the Rings?
Joe: Palantir is Lord of the Rings. Epirus is more real mythology — Palantir is fake mythology. We have quite a few things named after Tolkien as well.
The Palantir name — in the story, the palantíri are these crystals built by the lords of the Uttermost West, created for humans to defeat the bad guys. But 2,000 years later a bad guy takes over the technology and uses it for evil. So the idea with Palantir was we’re civil liberties-obsessed. We wanted to build it in a way that would go after the bad guys while also watching the Watchers — being really careful about how the technology is used. We’ve enabled killing thousands of terrorists. We’re happy with the results.
Why Go Where Others Aren’t Fishing [00:48:00]
Shaan: What’s fascinating is not only have you had multiple hits, they’re in areas that are really different — and just hard areas. Loading yards, EMP pulse technology. Is it a strategy to go where people aren’t fishing, or is that just what happened?
Joe: It’s definitely a strategy of sorts. You don’t want to build yet another restaurant app or yet another to-do list app, because that’s what every other 21-year-old is thinking about. That’s probably why I missed Stripe early — I’d been at PayPal and had to sit through every payment app pitch. There were literally hundreds of payment app companies. Peter would make me take the meetings after a while and I didn’t want to take them anymore.
There are parts of the world that are big parts of the economy — very important parts — that are not as exposed to technology cultures. Our advantage is we have a really tough technology culture. Let’s figure out how to expose it to something that hasn’t seen that before.
Idea Generation and Pitching Opportunities [00:51:00]
Shaan: What are some ideas you’re not doing just because you can only do so many things at once?
Joe: So many things. In the wealth management world — the process of starting a trust, the actual legal work around estate planning could be very automated with AI. You could have a smart lawyer and a few smart engineers and build a whole new trust company, a whole new estate management company. We have SaaS software for estate stuff you’d plug into, and I have a lot of distribution for this ready. I just need the right engineers and AI people. I’m not sure if it’s bigger than a billion or two billion dollar business, but it’s still worth doing, and it saves a lot of time.
But I’ll give you a really big one. Local government. We just sold OpenGov for $1.8 billion. Lots of lessons from that company — a lot of mistakes, it took us twelve years. Zach Bookman the CEO, my co-founder, did a great job. We learned that you can really only do things local governments already have a budget for. There are about ninety things local governments pay money for software to do, and OpenGov has about forty of them.
The insight is that there are trillions of dollars spent by local governments, and a lot of them can’t find the right staff for certain processes. A lot of their people are retiring soon. I think a business process outsourcing company for certain tasks for local government — augmented by AI — could be a huge business. Basically helping them with licensing, permitting, asset tracking, budgeting — faster and cheaper than hiring. The CFO doesn’t have to hire three people, they can hire one person, which is you.
It’s a very unnatural thing because it’s not how they work right now. But there’s a secular trend toward local government not having enough money. If you can find ways to save them massive amounts of money and make it higher quality service, there’s something there.
Shaan: What would the first six to twelve months look like?
Joe: You’d basically have to find early partners willing to trust you and outsource one of those ninety processes. Something like augmenting their ability to do licensing and permitting faster, tracking their assets, doing budget work automatically. You go in, figure out which processes cities are even willing to let you do. It’s one of those things where you have to just get that first beachhead.
Agree or Disagree [00:58:00]
Shaan: All right, I want to play a little agree-or-disagree game. First one: get your nut first, then work on your noble mission. A guy came on the podcast who said: first you go develop your skills, make a few million bucks, get personal autonomy — then pick a noble quest and go on it. Agree or disagree?
Joe: I agree in general. You want a solid base before you do something that’s a huge crazy risky quest. You don’t want to be poor for twelve years while you’re doing something if you can help it. I mean, that might be the path for some people, but if you can do what he said, that’s a really great way to do it.
Shaan: Next one: ideas are cheap, execution is everything — that philosophy is BS. Agree or disagree?
Joe: That’s a tough one. I’ve seen a lot of good ideas that weren’t worth that much because the people behind them couldn’t actually bring together the talent to execute. Ideas are necessary but not sufficient. You need both. Execution is the harder part. But you’re right — I’ve seen people with amazing execution go after the stupidest ideas and it yields nothing. You need both. Ideas by themselves are not that valuable.
Shaan: Focus is the superpower and should be the strategy — agree or disagree?
Joe: Focus is clearly the right strategy for the most successful people in the world. If you look at the wealthiest people in the world, they nearly all focused and compounded over a very long time. Palantir and Addepar are the two companies where I focused full-time — they’re my two most important companies. As an entrepreneur, as a business person, especially if you haven’t already had a giant win: focus, get that giant win, compound.
I think there’s some room where philanthropically, coaching entrepreneurs — I think I’m creating a lot of value in those areas too. But as a general rule, especially before you’ve had that first big win, just focus.
Shaan: You’re friends with Elon. What’s something you’ve learned from him or admire about him that the rest of us don’t know just from the outside?
Joe: I really admire his boldness. He’s really good at focus — really good at just saying no to anything that’s not tied to what he’s passionate about and focused on to win. Really bold at clearing nonsense out of the way and going right for the thing that matters.
Most people in my life would think I’m maybe too bold or too direct or too extreme. Elon is a whole step further, and I think that’s great to see. It reminds me there’s more room in that direction. I could be more myself in that direction and I’d be more successful.
Shaan: Is he focused though? He’s got Twitter, SpaceX, Tesla, Neuralink, Boring Company — so many things. How do you square that?
Joe: I think he was insanely focused for a very long time on the things he was doing and on scaling them. And yes, sure, there were two companies for a very long time, after doing one company for a very long time. And at the level of success he’s at — societally there are things that really matter for society that only he can do. It makes sense for him to expand.
I told you earlier — my partners are allowed to work on up to six things at once. When you’re coaching and managing, that feels about right. If you’re just operating, you want to do one or two at most.
The One Question That Matters [01:04:00]
Shaan: Last question. I think we’ve got to go, but my whole thing when I come to an interview like this is: what’s the one question that really matters? All the questions I’ve asked have been beating around the bush of this one thing. So now I’m going to ask it directly.
You have now created — by my account — at least four billion-dollar companies directly, probably more than that. I call that the Entrepreneur’s Grand Slam. You have done something that anybody who wants to make one successful company would be envious of. What would you say you are doing different from the average good founder?
Joe: So I have some unfair advantages. And it’s actually not that hard, once you’ve been really successful, to do it again. Routing it for the first time is the hard part.
But what am I doing differently? I have a massive talent apparatus. We have people at dozens of universities. Fellows programs, parties, a huge talent scouting department that works in different ways, different places. Each of my partners and I — that is our job: scout for talent. The more wins you have with smart people, the more you get to meet their smart friends. So we have a huge advantage in talent.
We also have a huge advantage of knowing the people who run a lot of big industries. We have good relationships with them, they’re advisers to us, they’ve been part of things we’ve done. So we can understand what’s happening in those industries, get quick feedback, and iterate on ideas.
One of the biggest mistakes founders make is they have an idea and they’re afraid someone’s going to copy it. They think the idea is too valuable. That’s really dangerous, because then you just don’t share it. You think it’s yours.
One of Peter Thiel’s thought experiments: imagine there are five founders who each have an idea. Four of them are secretly working on it, and one of them is talking to everyone and iterating. Which of the five is going to win? Clearly the guy who’s iterating with tons of people. We’re able to do that — iterate faster than anyone else because we know everyone running everything.
And then whenever we start a company — maybe this is a mistake because maybe I could have had a lot more money — I’ve always been a minority owner. When we start a company out of our build program, our fund might take twenty percent and my team and I might take ten percent, so we have thirty to thirty-five total. We’re very generous with the founders, the early employees, the advisers, the partners.
So from the very beginning I don’t have to wake up being the only one worried about this thing. There are thirty smart people waking up for whom that’s their focus and they’re on it. That’s such a huge advantage — to have other people obsessing over this thing. People who truly feel like co-founders.
I think one of the big mistakes is: “I’m the founder and you guys are the people working for me.” What are you getting from that other than your ego? I’d much rather be one of many. There are companies where people don’t even know I’m the founder, and that’s way better — because those people are proud founders, they’re out in the press doing it, they’re going to sleep worried about it because it’s their reputation and their focus.
How to Make Your First Million [01:10:00]
Shaan: In the intro I’m going to say something like “This guy has started more billion-dollar companies than anybody in America.” But forget a billion — let’s talk about a million. The name of the podcast is My First Million. I went and read a Quora answer you wrote. Somebody said, “If I want to make a million dollars, how should I invest my money?” And you said something like: if you want to make a million dollars, it’s not about how you invest your money, it’s about how you invest your time.
How should somebody who wants to make a million dollars invest their time?
Joe: You can’t just make a million dollars from investing if you’re the average person — maybe over fifty years. My grandmother passed away at 103 last year and she was a multi-millionaire, having grown up in the depression, just from saving money and cutting coupons. She bought one stock — my grandfather worked for Abbott and she held it for sixty years. It worked really well. She was a wonderful lady. But that’s probably not how you want to make money.
Especially these days. There are so many places where with your time you can learn how to help build things, help create things, be useful to other people. That’s what you want to be doing. You want to find what you’re passionate about, what you’re good at. You have to be willing to say, “This is what I’m not good at, this is what I am good at.” Then find who else is good at this — who can you learn from, who do you want to be like? Spend your time trying to find a way to help those people. Try to find a way to learn how their businesses work. Be part of a business that’s growing.
If you’re talented enough, the people you succeed with are probably going to be able to help you do something new later.
Shaan: And that’s what happened to you. You went to Stanford, went to PayPal, and now here.
The PayPal Intern Story [01:14:00]
Joe: Well, I tried a couple of times — they rejected me the first time I applied. I was a freshman at Stanford, I remember being at a whiteboard with Max Levchin and just arguing about some tech thing. Was PayPal already a big deal?
Shaan: Is that why you wanted to work there?
Joe: It wasn’t that much of a big deal yet. Peter Thiel wasn’t who he was, Elon Musk wasn’t who he was — their companies had combined to start it. But I saw all the smartest people, all these iconic people going there. People who thought differently. And so you try one year, get rejected, try again the second year.
Shaan: Second year, you’re an intern. What are you doing?
Joe: Not very glamorous. They basically had me setting up the PeopleSoft instance, which was terrible. But I sat next to this guy who was managing all the money on Bloomberg — a treasury function. I got to learn finance and treasury.
And then the big challenge was: the company was going bankrupt because Chinese and Russian mafia were stealing all of our money. So I ended up hanging out after work with all these guys who were working on how to stop the bad guys.
PayPal is a honeypot — it’s got all the money. A 7-Eleven clerk maybe having a bad year writes down five hundred credit card numbers and sells them online for five to ten bucks each. The Russian mafia buys them up. All of a sudden you get a chargeback that says $200 on PayPal for something you didn’t do. PayPal is left with the bill. This was costing PayPal several million dollars a month. A lot of our competitors went bankrupt.
So we learned how to basically stop the bad guys. It was a fun problem.
Shaan: One piece of advice I give on the pod: work on the A+ problem at any company, regardless of your role. Know what the real needle-mover is and try to be helpful on that.
Joe: I didn’t do this on purpose. I actually wanted to quit at first because the PeopleSoft thing was so dumb and I couldn’t figure it out. My dad convinced me — don’t quit, just do what you can to be helpful. That was good dad advice.
Shaan: I almost quit college for the same reason. My dad was like, “That’s exactly why you should be there.” You want to be surrounded by people who are better than you.
Recognizing Talent Within Two Standard Deviations [01:19:00]
Joe: One of the insights Alex Karp had — my co-founder at Palantir — is you can’t really appreciate a talent unless you’re within two standard deviations of it. You think of talent as being on a normal curve for any typical skill. The very best people in the world at something might be at the fourth or fifth standard deviation. But if you yourself are at the second or third standard deviation, you can start to appreciate that.
Whereas if you’re not great at physics, you’d have no idea which one is Einstein versus a top-tier professor. You can’t differentiate.
Shaan: Exactly.
Joe: But once you can recognize that talent, you can harness it for what you’re building. I got A-pluses at Stanford in computer science. I had friends who were way smarter than me at computer science, but I was good enough to know who those kids were, how to harness them, how to hire them.
From PayPal to Palantir [01:22:00]
Shaan: So now the next thing you do — you go and start Palantir?
Joe: Well, first I worked for Peter’s family office for a while. He was the first investor in Facebook — it was going on in his family office. There was tons of crazy stuff going on. A really expensive restaurant that went bankrupt, a spam company someone was starting. Peter is a genius but he’s not really a manager. So I started just taking charge — hiring people, helping him run things. Some of the older guys were like, “Who is this 21-year-old kid?”
That summer we worked on a project: what if we took all the stuff we did at PayPal with anti-fraud and generalized it for the intelligence community? What would that look like? My roommate Stephan and I from Stanford were sketching up products and building prototypes. That was kind of the genesis of Palantir.
Shaan: So PayPal showed you a sliver of the problem.
Joe: Basically, I got to know all the guys at PayPal who were in the Secret Service and FBI — the two groups in charge of arresting the mafia guys. They would come to us and ask for advice on other things. It was fun, like cops and robbers — every young man likes that, right? So we were teaching them about internet fraud in 2000-2001, and they started coming to us.
Then 9/11 happened. The government started spending billions of dollars on supposed solutions to catching bad guys. We knew these guys and we watched what they were spending money on, and we were like, “Oh my God, this is stupid. This is way behind Silicon Valley.” It’s not that I’m so smart — Silicon Valley had gotten way ahead just because it brought talent from all over the world. But what was being done in the Beltway with the old integrators was still stuff from twenty years ago.
We said: our nation is under attack and we’re wasting billions on stupid stuff. So let’s draw up what we could actually be doing if they did it better.
Shaan: What happens next? Palantir today is a $50 billion market cap company. That’s amazing. But it’s also twenty years later. Walk me back to the early days.
Joe: There are lots of different ways to build successful companies, so I don’t want to say this is the one path. But our path — I was very lucky to be sitting there watching all these companies come out of PayPal: IronPort, LinkedIn, YouTube, Yelp. The ones I could tell were more successful, it was obvious even at the time — they were the ones attracting the very best engineers. It was like the engineers were lined up trying to get into these companies.
A lot of people have a business idea and think “how do I hire an engineer?” It’s the opposite. The ones that are crushing it are like a nightclub where engineers wish they could follow those people in. Something about the culture is so dynamic that they’re able to try ten times as much, really quickly.
So we decided: we need to build a really top engineering culture. That was one of our core principles early on. We used to say “the engineer is king.” How do you have a culture that attracts the very best and respects them and makes them partially in charge of the company?
It was a bit chicken-and-egg. I was lucky to have a bunch of friends from growing up who are national math and chess champions. So we built this really core culture and then convinced some top PhDs to join.
Shaan: What was your pitch to them?
Joe: You need a magnet to bring talent in. The pitch was: the government is spending tens of billions of dollars in this area, they’re way behind what we figured out how to do in Silicon Valley. There’s clearly going to be a multi-billion dollar company that solves this problem. You’ll be working in close iteration with national intelligence and defense agencies. And if we succeed, we’re going to save thousands of lives and protect liberties.
It’s a fun, big mission.
Shaan: One of those things you said was “we’re working closely with these agencies.” How easy was it? You’re young — how do they even take you seriously?
Joe: Maybe it wasn’t entirely the case that they were iterating as closely as I’d like for the first year or two. Alex Karp was very good at this. Alex was at first just an adviser to us — helping us — and then Peter kept trying to get us to hire a CEO, because Stephan and I were twenty-one. You couldn’t really have a twenty-one-year-old running something and working with senators and the defense department. These generals and former admirals would come in and they just did not understand tech culture.
Finally, Alex Karp — who had been helping Peter raise money, they’re friends from law school — was a PhD in philosophy from Germany. In Europe, a lot of billionaires like to pretend they’re philosophers. It’s the high-status thing to do. Kind of like here they like to pretend they’re podcasters, I don’t know.
Shaan: Ha.
Joe: He knew all these wealthy people. He started advising us on how to talk to them, how to meet them, how to think about how institutions work. He had this whole thing about casting — how you approach people to be taken seriously.
Shaan: What do you mean by casting?
Joe: If we show up and we’re these quirky twenty-one-year-olds from Silicon Valley — that’s the wrong casting. Whereas if you show up, the former head of the CIA has made an introduction, Peter Thiel is backing it, and senior people they respect have already explained to them that this is a project based on a major advancement made in Silicon Valley and they want your feedback — that’s a high-status way to come into the room.
Shaan: Versus walking in wearing a hoodie going “look how smart I am, look how right I am.”
Joe: Exactly. The substance is important, but you can’t lead with substance because substance alone is not taken seriously. You have to have the substance to win, but you also have to have all the trappings of why they should take it seriously and why they’re comfortable doing it.
Alex just had this wisdom about people, institutions, and how to navigate that — while the rest of us had the best tech culture. So we convinced him to pretend to be CEO for a few years so that Peter would stop trying to hire some senior military guy. But then he was such a good CEO that he’s still CEO twenty years later.
Peter Thiel’s Role at Palantir [01:38:00]
Shaan: You mentioned Peter. What was he doing day to day?
Joe: Peter was first of all financial support. No one would give us crazy kids money. A lot of people around him were like, “This is the craziest thing you’re doing. It’s one thing to waste money on a restaurant, but you’re giving these kids money to pretend to be spies and build things? There’s no such thing as a company that does this.” You have to give him credit for being willing to bet on a crazy idea. He put a lot of money into it early on. That’s a huge amount of money for crazy kids.
And then Peter was really good for strategy. He stopped us from a couple of early partnerships. A lot of startups, when you’re getting going, you get really excited about a partnership on the table and you think it’s going to help you. But a lot of partnerships will cap your upside, limit your ability to iterate and try other things. Because he had the experience and was going for a multi-billion dollar win, he was really able to steer us away from a lot of big early mistakes.
For example, we might have spent all of our time selling our interface in Europe and giving them all the rights to it and flying over to support things they were doing — random things that sounded really good when you’re desperate for that first $500K in revenue. But you have to make sure you get it in a way that’s going to actually scale into something much bigger and not distract you.
Shaan: And the Facebook thing — were you around when he wrote that check?
Joe: Yeah. I can’t say I thought it was going to be as big as it was. I don’t think anyone realized that. But it was clear they were really smart, really interesting guys who were very aggressive. One of my good friends who actually helped Peter write the check is played in the movie by an African-American actor, which he jokes about.
Shaan: He’s a five-foot-four white guy.
Joe: Exactly. A little Jewish guy. That’s the ultimate compliment. If anyone out there wants to rewrite me as a super handsome Black guy, I would love that — thank you.
Closing [01:45:00]
Shaan: Thank you for doing this. I’ll give you the highest compliment, which is: I learned a lot today. That’s my goal with all of these conversations — what are some golden nuggets that come out of this. So thank you so much for doing it.
Joe: It’s fun getting to know you, Shaan.