Palmer Luckey & Mike Solana On The Future of War, Israel, TikTok Ban, & U.S. Manufacturing

It's even more involved than that. I often get asked by people who are not experts in it, or experts in anything like much of the media, if I would sell to X country, Y country, or Z country, what about this conflict, what about this hypothetical? The good news is that it's actually not up to me. The United States government has rules, so it doesn't really matter if I would be willing to sell to North Korea or willing to sell to China. It is prohibited by the United States government, and I'd go to prison for doing so. That's a point worth harping on a little bit because it's important that it doesn't matter what my opinion is. You don't want to live in a country where our de facto foreign policy is determined by corporate executives deciding who they want to work with and who they don't. People in my position, in a private corporation making weapons systems, shouldn't have the ability to decide, "Yeah, I would love to sell to North Korea" or "No, I personally would not, but the guy down the street would." That's not a tenable situation. So luckily, the people who are making these decisions are to some degree accountable to the populace through the democratic process. If someone is doing things they shouldn't, you can vote them out. Companies like mine have to work pretty closely with the government. There's something called FMS, the Foreign Military Sales process, and pretty much everything that is exclusive for military use has to go through that process. Luckily, the United States has gotten better at facilitating those transactions. We could talk about how they could get to a good place; I'd say they're better but not good.

I can't talk about everyone that we're working with, especially where it would disclose things that are operationally sensitive. But things I can talk about: we of course work with the United States, we do a lot of work with the UK Ministry of Defence, we do a lot of work with Australia. We've sold to a few other European countries. We are doing some stuff in Ukraine, have been since the second week of the war, and that continues to be ongoing. We're part of one of the aid packages getting some more loitering munitions into Ukraine. We've cast a pretty wide net. When we started Anduril, it wasn't just about the United States; it was explicitly about defending the West. I think that even goes back to the name of the company. We wanted to take the position that countries aren't just different, that there are countries that are better or worse, and that we were going to be on the sides of the ones that were better, as determined by the handful of values that still remain true between those countries today.

I think it depends on the technologies you're talking about. There are a lot of things that certainly accrue more to the defender than to the invader. For example, if you're talking about the ability to manufacture large numbers of autonomous systems, it really doesn't matter so much who's the defender versus who's on offense and defense. It's about who has the larger industrial capacity. If you have China as this huge major industrial power, they're going to have a huge advantage in a world where autonomy enables them to win against a much smaller nation. But on the other side, especially as you start to look at things that are enabled by AI, there are a lot of advantages that are going to accrue more to the defender. I kind of feel this way about bioweapons, for example. People point out that large AI models might allow for the creation of really dangerous bioweapons. I think that's kind of scary, but at the same time, AI biodefense is actually going to get a stronger accrual of advantages. In other words, if I can use it to build systems that keep people free of pathogens — I'm going to go way off into science fiction land here — but if we can get to building nanobots that are going through our blood and getting rid of everything that is not matching a bio signature that it's seen the week before, that's an advantage that is going to work against every potential pathogen you could imagine making. There are thousands of ways you could try to beat that, but the one thing you've made is now defending against all of them. And it's worse than that for the aggressor: they don't just have to go up against one brand of nanobots; maybe I'm going to have 10 brands of nanobots in my body, including an open source one, and they're all going to be continuously updated and competing against each other to stop these pathogens. That's an example where I think the defender does get the advantage.

They're shooting down missiles with missiles. It's fundamentally a counter-ballistics system, so it's not necessarily designed to go after maneuvering targets like aircraft, but it's really good at going after the threats that Israel has been seeing. Iron Dome was a collaboration between the United States and Israel; there's a lot of US weapons technology in there and a lot of Israeli know-how. The way I look at Iron Dome is the ultimate manifestation of the future of the United States's role in future conflicts, which is not to be the world police but to be the world gun store. It's not that we're going to be setting boots on the ground, marching around and dying for other people's countries. Our future role is going to look like making things like Iron Dome more accessible to other nations, making defensive tools that allow people to defend their homes, making these countries into prickly porcupines that nobody wants to step on. That's probably the best way for us to shape a lot of these foreign wars: pick who we're going to sell defensive tools to, double down on getting the right tools, and let them die for their countries, giving them the tools to do it themselves rather than taking on the load ourselves.

The rockets are janky, but they're carrying pretty significant payloads. The reason they're not such a huge threat anymore is only because of this highly advanced defensive system that's been optimized against them. They were a very powerful asymmetric capability against a lot of people. Put another way, you could have a lot of those janky rockets level Washington DC in a matter of a few days, because we don't have a system capable of a prolonged defense against a barrage like that. That's true of most nations in the world. Israel has been forced to confront the reality of violence as a part of their existence, so they've had to build things like Iron Dome. But I think the future is going to have a lot of these specialized systems. In the past, the United States developed these very exquisite, powerful weapon systems that were often optimized around projection of force to the other side of the world, and then we would sell them to countries through the FMS process, selling them things that were way too expensive for the particular threat they face. A critical example: Patriot missiles cost a couple million a shot, and you often need to fire two of them to make sure at least one hits. They're reasonably reliable, but if you really need to take something out, you're not going to bet just one. They cost a couple million bucks a pop, and there have been cases where we have allies shooting down drones that are quadcopters costing hundreds of dollars with multi-million dollar missiles. The average American is not quite aware of how ridiculous it is that we're selling these tools to people that are hugely overmatched for the real threat they have. In the future, you're going to see the US doing a lot more work like what Anduril has done, where we work with customers to build things that are tightly tailored to their particular problem. The thing that solves Israel's problem is Iron Dome to a degree, but even Iron Dome doesn't solve the problem of a dozen other US allies. We're going to need to build the things that allow them to cost-effectively fight their terror.

There's a whole bunch of stuff I can't talk about, but there's one thing — I can't talk about the details of it — we have a system that's a reusable counter-effector. It allows you to, rather than throwing away a missile every time you need to shoot down a large aircraft, use this tool again and again and again. That hasn't really been the way that cost-plus contractors have thought about the problem. As long as governments are willing to buy single-use missiles that are really expensive, you want them to buy a whole new missile every time. But if you start to think about driving down the operational cost of using this for long periods of time — think about dollars per kill rather than dollars per missile fired — you're going to make very different decisions. Keep an eye out on some of that stuff. I think we are thinking about it pretty differently because we have different incentives. It's not because I'm smarter; I think a lot of these companies could have come up with this stuff, but they didn't have any incentive to do so because of cost-plus contracting. I know you've talked about it so many times, but for this audience: cost-plus contracting is the way most procurement of new weapon systems is done in the United States. The background here is that 80% of procurement money goes to just five companies, and 30% of major weapons programs have only a single bidder, meaning no competition at all. A cost-plus contract is where the government agrees to pay you for all of your costs and then a fixed percentage of profit on top, let's say four, five, or six percent — usually a pretty low margin compared to the consumer space or B2B space. The problem is that it incentivizes your costs to be high because you only make money on the plus, and there has to be more cost for there to be more plus. So you actually want your programs to run as long over schedule as you can get away with before it gets canceled. You want to use the most expensive bolt, the most expensive engine, the most expensive sensor you can possibly justify. You're never going to propose a system that can solve a problem for $10,000 if you could instead build a system for $100 million that does the same thing. Where's the money in it? You can't make money in a cost-plus world if you're too cheap. You have to fundamentally think about the problem, and maybe that means defining the problem as larger than it needs to be. For an Iron Dome example, what if they claimed Iron Dome didn't just need to shoot down these janky ballistic missiles but had to work against high-speed maneuvering air targets as well? That's the type of thing a contractor would say because they can't make money on the cheaper system. It's a horrible set of incentives.

We always say to trust that the customer knows what their problem is, but never trust that they know what the solution is. These guys know how they're getting their asses kicked and they know what their weak points are, but they don't necessarily have enough awareness of the entire engineering possibility space to know the right way to solve their problem. And they certainly don't know what building blocks we are incentivized to reuse because we're not a cost-plus contractor. We're not trying to redo work over and over again; we want to reuse the exact same ingredients to build new things, just like Taco Bell does for their menu.

If you really want to get spicy, I'd say it's not that people don't remember it's that people have erased it. You'll see people strongly arguing, "Here's why it really wouldn't have happened, here's why they wouldn't have fought to the death." There was a recent high-profile internet talking head journalist who said, "Do people really imagine that the Japanese would have fought to the last head?" It's like, dude, look at the outer Pacific islands. They did fight to the last person and they ran off cliffs when it became clear we were going to win. People have erased this because I think people don't like to admit how different cultures can be. They act like the modern multicultural global culture is what you need to look at World War II with, and they're just wrong.

There are lots of people, different religions in different ways, all jihadists.

I've been thinking about not exactly this problem, but a much broader problem for the last few years. I've talked about this publicly, but nobody ever really talks about me talking about it, so I guess I'll just keep saying it and eventually people will pay attention. Very big picture: I believe that the next war fighting domain is the subterranean domain, in the same way that air, land, sea, space, and arguably cyberspace are. People think I mean tunnels and caves — that isn't what I mean. I mean using the entire volume of the Earth as a three-dimensional space that you can maneuver in and fight wars in. That sounds a little crazy, but then you remember that the United States and the Soviets believed this during the Cold War. The US and Soviet Union both had subterranean programs trying to build nuclear-powered underground vehicles that could bore through the Earth as easily as a submarine moves through water or an aircraft moves through air — maybe not quite as easily, but the same principle. The idea that you could go anywhere arbitrarily and deliver big payloads, big effects, using the Earth as the medium through which you traveled and hid. Unfortunately, when the Cold War started wrapping up, all of this kind of collapsed. The Soviets actually built a working prototype — it wasn't just a concept — but with the collapse of the Soviet Union and the recognition that nuclear was not going to be a significant part of our civilian economy or wartime strategy, we banned the idea of tactical nukes. We basically said strategic nukes are all we're going to have, we're going to keep nuclear-powered ships and submarines, but we're not going to find new things for nukes to do anymore. That was a huge mistake — maybe the biggest mistake our country's ever made. We could have a whole podcast on that. Recent advances in technology, especially on the power density side, air-independent propulsion side, and the ability to put useful payloads into very small things, have actually brought back this idea in my mind as the next future of warfare being underground. I've actually built vehicles here at Anduril that are capable of tunneling underground and delivering a variety of electronic and kinetic effects to wherever you want them to go. I wish I was further along on it. The tunnels in Gaza are just one example. North Korea has a huge tunnel system built into mountains and the ground. Iran is very smartly locating all their critical stuff deep underground. It's one of those cases where you go to war with the tools you have, not the tools you want, and right now we do not have the tools to fight in the subterranean domain. I think eventually you're going to have a US Subterranean Corps — I don't know which service that will live under. I strongly believe that if we had robots and vehicles that could move around underground as easily as a submarine moves in the water, this tunnel problem would be so much easier to solve. It would be easier to map them, easier to take them out, easier to do so without having to go through the air and then through the ground and then into the underground domain. Bombs from the air are a really high-power, imprecise way to take out tunnels deep underground. It's really the only option the Israelis have, and the only option anyone has. But if you look at it from a first principles perspective, bombs that have to go through the air, then through buildings, then through the ground, then into a tunnel are a lot less elegant than something that could just go into the tunnel.

What I'm saying is more as an example of a broader shift in what we were to do with nukes. The idea being, nuclear weapons on the battlefield used to be a very reasonable thing. We imagined we were going to use them to bust tank formations. We had nuclear anti-aircraft missiles that could take out a squadron of bombers. Nukes were envisioned as an enabling technology for all types of war fighting across every domain. We've decided that's not the case. Nuclear weapons are basically the domain of ICBMs — strategic mutually assured destruction weapons — and nuclear-powered submarines, which are themselves primarily justified through the existence of MAD nuclear weapons. We also had built nuclear-powered aircraft carriers and weren't going to go back on that because it's the only way to make a modern carrier group work. But there was a recognition: we're done. We're not going to build nuclear-powered fighter jets, nuclear-powered tanks, or nuclear-powered boring vehicles underground. That's off limits now. The shift away from tactical nuclear weapons is just part of that shift away from having nukes on the battlefield as part of an everyday conflict. It wasn't necessarily a ban; it was a policy change. We said we are not going to have tactical nukes as part of our strategic doctrine, whereas it's worth noting the Russians do to this day. The Russians train for using tactical nukes on the battlefield; they imagine deploying a nuclear weapon to destroy a battalion of tanks, blow up a naval formation, or get rid of an airfield. That's why, maybe a year ago, Ukraine was at risk of this. They moved the unit in charge of tactical nuke deployment into high alert and had them doing day-to-day training on scenarios. It's part of their doctrine; it's not part of ours. Also, Russia has made similar good decisions in using nukes for non-military applications. Did you know that in Russia, there are multiple privately owned civilian nuclear icebreaking and container ships? How crazy is that? That's the type of thing we should have in America. We should have zero-emissions, zero-carbon, dirt-cheap-to-operate nuclear container ships, and instead it's the Russians. You can even book a tour on one as a normal person; you can go to the North Pole on a nuclear-powered icebreaker run by a civilian company. They did not abandon the atom the way we did, which is again one of the worst decisions we ever made.

We've chosen that we don't want it to be part of the way we fight wars. The extra power is what you would need to bore very easily. That gets down to the payloads problem I talked about. Back then, they imagined these would be manned vehicles, so they had to be large enough to have crew quarters and room for pilots and operators. To do that, you need fission or fusion to make enough energy to displace enough soil at a fast enough speed to be relevant, so they had to be nuclear-powered. Today, because you can automate so much of that, you don't necessarily need a room full of signal operators running your electronic warfare or data collection systems. You can fit all this stuff into much smaller diameter stuff. To the extreme, a vehicle that is 100 kilometers long and a millimeter wide is not going to take very much energy to push through a hole, because at the end of the day, you bore the hole. Length is free once you've made the hole; everything going through the hole after that is negligible energy cost. Making that diameter in the first place is what's expensive. I don't want to give away my whole scheme here, but it totally works. This is not crazy far-future science fiction. If I had started working on it five years ago, we would already be deploying it today. Unfortunately, I only started working on it two years ago, so we would be deploying our own boring technology.

There are a ton of ways. The first thing you do is get tactical situational awareness. You need to know where these tunnels are. First, you send these things in and map them all out using ground-penetrating radar and a variety of acoustic schemes. It would not be that hard to build a pretty high-fidelity map of what's going on there, especially if you had built that map ahead of time. If we had these tools, you would be building it as the war starts; you would have continuous awareness of where every tunnel was. Once you have that map, there are a variety of effects you can deliver. The boring payload — pardon the pun — is just explosives to collapse the tunnels. Another good idea is to flood the tunnels with seawater from the Mediterranean: bore a hole from the Mediterranean to a pump station, then to a whole bunch of holes, and just fill them with water. Then there's my personal favorite, which I haven't seen anyone talking about, but it's something I've always thought was underutilized: using carbon dioxide as a structure denial weapon. CO2 gas is really easy to move around as liquid CO2 or as dry CO2, with really good expansion ratios. A fun fact about the human body: it does not have the ability to detect lack of oxygen; it only has the ability to detect the presence of CO2. If you've ever been in a room that gets really stuffy, that's the CO2 building up, making you really want to get out. If you're in a room with enough CO2, it feels like you're holding your breath and suffocating, even though you're totally fine in terms of oxygen going to your brain. I think we should just fill the tunnels with carbon dioxide. That wouldn't be perfect — they could still move through them using self-contained breathing apparatus, so these guys could use their scuba gear and walk around. But it would make it a huge pain in the ass to be in those tunnels if you couldn't take even one breath without feeling like you're suffocating, even though you're not actually suffocating people. If you filled the tunnels with nitrogen gas, people would walk in and just fall over dead from suffocation. You pump CO2 in, they feel like they're suffocating — it's a immensely powerful physiological response, but you wouldn't necessarily be killing them. I actually started experimenting with this a long time ago, doing CO2 delivery from drones. The idea being, in a hostage situation where normally you'd have a SWAT team trying to rescue people, you would just pump the building full of carbon dioxide gas from a liquid CO2 container. In a few minutes, everyone in that building would feel like they're in the stuffiest room they've ever been in, and an immense panic and need to get out. I've tested this on myself; it definitely works. It's not great.

This is all a game of cat and mouse. Very big picture: you're never going to win forever. You just need to win more often than they do at the important breakover points. If I can build something that keeps me and my people safe for five years, and then they're able to break it, but then I'm able to fix it again with something better within a day — that's a huge win. That's really the best we can hope for; we're not going to get to absolutely perfect coverage. It used to be that the United States could build things like stealth fighter aircraft and we were the best in the world by far, nobody was close for decades at a time. Those days are kind of over. The proliferation of knowledge has made it impossible to keep secrets like that once they're out of the bag. So we need to be in a world where we can move quickly. Other technologies that are going to be a really big deal: I think there's a lot of stuff that can be done to deny humans access to areas uniquely. This is a little boring, but there's a lot of old stuff on the chemical warfare and biological warfare side that is not being used simply because everyone's agreed it's a bad idea. It works when everyone is a rational actor and afraid of biological weapons in a mutually assured destruction sense. But how does that work when you have irrational actors who believe that the highest glory in life is to die a martyr? You've shifted the equation where maybe they don't even mind overwhelming retaliation. I also worry about people believing that they can use biological weapons in a way that accomplishes their ideological aims. For example, imagine Hamas realizes it's probably a bad idea to use biological weapons because they'll end up killing a bunch of their own people.

I guess one thing we have in our favor in that regard is that Americans' DNA is so diverse. There is some robustness there, thanks to US immigration and over so many hundreds of years. The bio thing is interesting. I'm always worried about the next virus, the next HIV. That's one I'm really frightened about, just because it takes so long to even realize that you're sick. I was talking to a guy who works for the government. He mostly spends his time in airport bathrooms checking for viral shedding, because that's where most of the contamination comes from. They're just looking for evidence of new things. They don't even care what it is necessarily; they're just looking to see if something is proliferating. They don't even worry about whether it's worth understanding, because anything new is bad. You can't have anything you don't understand suddenly proliferating. That one is a little bit by the way. He's probably been thinking from the perspective of viruses that have long incubation periods or take a while for you to know you have them and then become lethal. But if you look at things that have been developed on the therapeutic side, like optogenetics, where they're figuring out how to build viruses that bind your cells and make them responsive to certain frequencies of light, you start to realize you can actually build pathogens that do not have an internal trigger. You could, for example, hypothetically build something that infects you, has absolutely no symptoms at all, and then when the summer comes and you're exposed to UV light of a certain concentration, all of a sudden you become extremely sick and die. You could deploy that so an entire country could become infected during one period and then, boom, without even knowing it, everyone's triggered all at once. The old rules of pathogens don't really apply to engineered viruses. You can have viruses that would never survive in the wild turning into something much scarier.

We've talked about this, but I don't want to be doing Anduril necessarily. I would actually be having much more fun building the world's fastest race cars or floating cities. I want to be doing skying. I want to be planning my dream ship RV that's going to take me to the moons of Saturn. That's really what I want to be working on. I'm a fan of uplifting species, which leads to a multi-polar world that's very dangerous, which is a very reasonable fear, but nonetheless I want to do it. Where's my super intelligent parent? That's what I want to be working on. But I'm working on national security because I think it's very important. You do have to deal with the seriousness of it. I don't want to say I'm cut differently than other people, because I've got thousands of people working at Anduril who think the same way and are dealing with those same pressures. But if you work on this every day, it's not quite so horrible being exposed to new threats. You're like, 'Oh, it's another threat we need to consider.' There are all these threats we're thinking about every day. When you spend all day thinking about how you're going to counter Russian weapon systems that are hugely outclassing our ability to defend, when you're thinking about how screwed we are versus China in a mass-on-mass conflict, you throw in the viruses, energy weapons, cyber attacks. There's a whole bunch of this stuff.

A lot of what China is building right now, I actually think is not the right thing for them to be building. They're investing in a lot of prestige projects: aircraft carriers, high-end fighter aircraft. I think a lot of it is just political. They want to be toe-to-toe with the United States. They want to show that they're a superpower. They want people in other countries who think that's what a superpower looks like to believe that China is equal to the US. So I would say those things are theater. When you hear people saying, 'Oh my God, China's going to build so many aircraft carriers, they're going to have all these fighter jets,' they're kind of missing the real game. The real game is the stuff that China is not going to reputationally ride on. They're not going to go to nearby nations and say, 'Hey, look, we can make tens of thousands of autonomous anti-ship missiles at a tenth of the price of the Americans,' because that's not what a superpower looks like in a marketing sense. But it is what a superpower looks like in terms of ability to project power, take over regions, and invade other countries with impunity, especially if they sink the entire US Navy in the first round of that fight and then just do whatever they want for the next 20 years as we remain unable to build any new ships. The real power of the Chinese military is the ability to manufacture. They have dozens to hundreds of times more military shipbuilding capacity than we do. People debate on how much, but it's at least dozens of times more capacity. And as someone who's made things in China, I know how cheap things can be done over there, not because the labor is cheap, but because the people are good, the regulations are good, the supply chain, logistics, and materials are very favorable. So when I talk about China kicking our ass in a mass-on-mass conflict, I'm talking about what happens when you have the things they're good at making versus the things we're good at making in an actual war, not a prestige war or marketing war. And unfortunately, right now they are very much on the right side of the equation. They're building weapons that are sometimes better than ours for a tenth of the price and a hundred times the scale.

It's a whole bunch of things. Anti-ship missiles are a huge concern, particularly ones that get into the high supersonic or low hypersonic regime where it's very hard to defend against them. And even if you can defend against one, can you defend against ten? Can you defend against a hundred? When an aircraft carrier costs so many billions of dollars, it doesn't take that many. You can justify throwing a lot of missiles at it, and it still makes a lot of sense from China's perspective. The scenario I'm talking about, I don't think China is going to go for a Pearl Harbor. I don't think they're going to aggressively seek out the US Navy wherever they are on the planet and try to wipe it out. That's not in their strategic interests, not today, probably not in the near future. But what they probably could do right now is deny the US Navy the ability to operate in pretty much any area that they make a priority. If you make enough hypersonic anti-ship missiles, you're not going to have ships existing in range of those missiles. And if the United States knows that, what actually happens is not the destruction of the US Navy but the neutering of it. It's not that we lose those ships; it's that we realize we can't actually use them. So they just sit somewhere else, watching Taiwan get invaded, then watching the Philippines get invaded, then watching Northern Japan get invaded, then watching Southern Japan get invaded. The idea that we don't have the technology... It's not the technology; it's the manufacturing capacity. We have anti-ship missiles; they're great. We don't have enough. This is the Anduril mission problem: how do we do these things cheaper, better, faster? The United States has a lot of excellent missiles, not just on the anti-ship side, but anti-air, tactical missiles for ground targets. But we don't make enough of them. If China makes a hundred times more missiles than we do, it's just very hard to win.

We have put a lot of effort into this. We're not there yet. We're doing a lot better than basically anyone in the entire country, but we're not perfect. The thing is, there are two standards as I see it. One is the legal standard, which we obviously comply with. They say you can't make certain components in China, you need to remove dependence on China with these materials and components. Complying with that is easy. The hard part is not relying on China in practice. On paper it's easy; in practice it's hard. There are a lot of companies that make things compliant with these rules, but if we actually sanctioned China and couldn't get any components or materials from them, or if they couldn't get stuff from us into them and back into us, their whole supply chain would fall apart. They would be unable to make things. There are a lot of people making things that they say are made in the United States, compliant with NDAA regulations, that are going to be impossible to manufacture if we actually go to war with China. That's just the reality. It really bothers me when companies only fit the paper standard and not the practical standard, because the paper standard is supposed to be the backstop for the practical standard. People have figured out how to get around it too much, and China has figured out how to get around it too much. One of the crazy things I've seen is executives at major defense companies going out and saying it's impossible to decouple from China, that it's impossible for the US or the defense industrial base to decouple from China. At the same time, these same companies are getting billions of dollars in contracts for weapon systems that are only useful in a world where we are fighting China toe-to-toe, which implies a decoupling. They say we can't decouple, but also give us tons of money for things that are only useful after we decouple. It becomes clear that it's really just a near-term argument for why Congress should not force practical independence from China into your company. You asked for advice to other companies: take a look at this dynamic, realize how farcical it is, realize that it is going to get forced on you politically at one point or another. Maybe it's out of the blue, maybe it's after China invades Taiwan. But I tell new founders all the time: don't end up in the situation that Apple is in. Don't end up where you've built a company so dependent on China that you can't speak your mind about things like Uyghur genocide, organ harvesting, censorship in their media, or Chinese censorship of American media. Tim Cook could not even talk about the espionage problem. I remember the hearing with Cook, Zuckerberg, and maybe Dorsey and Bezos. They were asked if there was a problem of corporate spies. Tim Cook was just like, 'I've never heard of anything like that.' And they cut to Zuckerberg, and he laughs and says, 'Of course there's a problem with this. Everybody knows there's a problem with this. That's the name of the game.' And it's not just China, it's not just spies from China. Of course there's a problem. You might remember when Peter Thiel was getting slammed for this, where he said that Google is certainly compromised by Chinese intelligence. They used the old routine: 'Peter Thiel claims without evidence that Google is compromised by intelligence.' But even just the statistical argument is so obvious. They have hundreds of thousands of employees. Imagine if China was so incompetent that they couldn't get a single asset into one of the largest American companies. Flip it the other way: imagine how incompetent the United States would be if we didn't have a single asset or source anywhere in Alibaba or Tencent. Isn't that obviously unbelievable? But people freak out about it when you talk about it, because Tim Cook can't talk about it. I think a lot of these media companies aren't allowed to talk about it either. They have to pretend it ain't so. Part of it is the compromise. Part of it is that a lot of people in media are genuinely too dumb to realize that spies are a common thing. It's just culture war stuff. They don't get it. They're thinking, 'Oh, no, spies? That would be an act of war.' They don't realize that there are spies everywhere. That's what is happening. I don't know where you've been, but that is the landscape. Imagine a world where California politicians are sleeping with Chinese spies but they haven't managed to get any spies into Google. It's obviously unbelievable. But you're right: Tim Cook can't talk about these things, and it's not because he's an idiot. It's because he is acting in self-interest to preserve his company. So I tell founders: don't let yourself get into that situation. It will happen if you're not careful. If you're careful about it, you can remain untethered from China. You can even do it if you are using certain Chinese materials and components, as long as you always have a backup plan, a second path. That's my biggest advice to people, because imagine how stupid you're going to feel if you start a company, raise money against all odds, achieve the dream, build a company worth many billions of dollars, create financial independence for you and all your employees, and then China invades Taiwan just like everyone expects, and Congress passes sanctions on China, and your company completely collapses and fails overnight. Who's the idiot? Is it everyone else, or is it you?

We're not in the most influential position. It's largely a political problem. The United States didn't just ship all our manufacturing overseas out of the blue. It was the result of specific policy decisions around how we levy tariffs, what we were willing to have made by other countries, what we were willing to sell to those other countries. The China problem is one of our own creation. We created it through policy. If you continue to allow foreign adversarial powers to sell heavily subsidized goods to the United States... Remember, it's not a fair fight. People say, 'I'll just do better.' But you don't get it. The Chinese government is already subsidizing certain industries to ensure we don't build up a competitive one for strategic reasons. If you figure out how to do it even cheaper, they'll just undercut you even more, because they don't need to make money in everything all at once. If we could all do this and everyone got way cheaper across every industry, eventually China runs out of debt and can't compete. But I'm not optimistic about the ability to outcompete China to an extreme degree in every single industry for as long as we have the energy policies, labor policies, and tariff policies that we have. You can't have free trade in a world where the other country's government is guaranteeing an unfair competition. A very narrow example: China doesn't even use tariffs. They just straight up ban huge swaths of the US economy from competing in China. It would be like if we said, 'Chinese movies and Chinese phones? Nope, just not legal.' We're having this principle conversation right now about TikTok, even though it's an actual spy app for the CCP. Even if it wasn't a spy app and was just a very influential social media platform, the dominant one in the country, meanwhile every single one of our competing pieces of software is banned in China. That makes no sense. That's what I was saying during the Trump years. You need to look at it. I was kind of frustrated that people made TikTok into a cultural issue. I'm totally on the culture war side of it, but practically speaking, you should not make this a culture war issue. Don't talk about how it's ruining our youth's ideals. Just say strictly on a trade basis: we cannot allow them to sell this thing to us if we can't sell the same thing to them. That should be totally fair. My idea was to just say, 'Hey China, we're not banning TikTok. We're saying there must be the law of equivalent exchange. You've got to have the ability to go back and forth.' They would have to let us in. Of course, China would have said no, which would have been a de facto TikTok ban. But that's how I think we should have done it: a pure trade issue. It would have been easy for Congress, easy for the president to justify. All the youths would whine, 'You're banning TikTok.' Say, 'Listen, why? You want free trade, you want globalism? Okay, let's do it. What we can have is a one-sided system. You have to be all the way free or not.'

This is what the AI doomers don't recognize. They imagine that we live in this isolationist world where if they can just protest enough and get automation out of the factories, you're going to keep all these factory jobs. That's not what's going to happen. You're just going to have everything made elsewhere in the world, and they'll have incredible wealth and prosperity of everything they could ever want, and we're all going to be buying handwoven baskets at the farmers market for $300. That's actually the future that these people are signing up for. You're right. We've got to stop just tinkering with the edges. We need to fundamentally rethink the way. I think we can do this. You're probably familiar with the fact that China can't make semiconductors without American and European tools. There are things that we've proven the West can do that China hasn't been able to grab their heads around, despite trying really hard. I would love it if we could prove that automated manufacturing on a large scale is one of those things, because they have not figured that out. They are still very manpower intensive. What if that is an advantage that we could build? What if things became cheaper to build in America than in China? That would basically solve the problem right there. You would screw their economy, you would starve them of all their productivity, and we would get all that money instead. That's my dream. Where's my fully automated communist factory? Robots in America.

I'm actually really confident on chips independence. Not just using traditional techniques, which we invented here and mastered here. I think we could rebuild that. There are also a lot of new techniques, both on the lithography side for traditional semiconductors, but also people doing things in optical computing. If that plays out, it could make it way easier for us to do. I actually feel pretty good on the high-end semiconductor side. I think that's a place where the US can do it effectively because of the structure of our economy and labor force. It doesn't take a ton of people. It's not necessarily hugely environmentally intensive. People complain about water use; shut up. It's so easy. Just go to a place with water. It's totally solvable. I'm actually more worried about everything downstream of those semiconductors. Let's suppose the Chips Act is a huge success and we're making high-end computer processors, GPUs, you name it. We don't actually have PCBA capability, product manufacturing capability. If you actually want those to turn into laptop computers, you need to send those chips back to China, where they will be turned into laptops, which are then sold to us at a huge margin. Now we make $10 on the chip and they make $100 on the laptop. Who's the winner? It's not actually us. Then imagine that we end up in a place where there are sanctions on China and we can't work with them. You could end up with a chip factory cranking out tons of chips, and you have chips in a warehouse with no ability to do anything with them because we can't build the supporting semiconductors, the USB host controllers, the actual motherboard assemblies cost-effectively, the batteries, the displays. That's actually my biggest concern: we solve the chips problem and then we don't have the rest of the value chain. I would feel much better if we were investing in that as well, but it's currently not really the political priority. Maybe it's an idea whose time has not yet come.

Americans suck because they don't care. When people try, Motorola tried this. They were making a phone in Texas in the early 2010s, and it was only a little more expensive than the Chinese-made version. They were making the whole phone in Texas. And guess what? Nobody cared. Nobody was willing to spend an extra $20. I think that while a lot of COVID was a memory hole, I think the concept of 'buy American' as a silly thing is over. I do think the average person, despite politics, sees utility in things coming from America. There are all sorts of reasons: you have your labor people, socialists who really care about this, union guys who care, a right-wing component, a defense component. There is a reason. So maybe the time has come back around. I really hope so. I want to work on that. Give me a note of optimism to take us out. What is something that makes you feel hopeful right now?

I get asked a lot, and this comes back to the beginning: if I would ever sell weapons to China, North Korea, or Russia. Of course, it's a ridiculous gotcha question to fill time and put me on my back foot and allow them to show me waffling and saying, 'I'll do whatever the government tells me to do.' It's a very predictable, annoying question. But the answer I give actually reflects my optimism. I actually hope that I do get to sell weapons to China. I hope that I do get to sell weapons to North Korea. That sounds crazy until you remember the example of Japan. There was a time we were fighting a world war against these guys, and now we are the backbone of their defense strategy. They just doubled their defense budget, and most of that is going to come to the United States for our weapon systems. Germany is more into their own stuff, but they buy some of ours. I'm glad we're able to sell Germany weapons. I think they're screwing up their country in a thousand ways, but they definitely pass the bar of a country we should want to win versus Russia. So I actually feel pretty hopeful about a future where the reason I'm never going to draw a line and say I would never sell to China or Russia is born of my optimism that you're going to see a change there. Whether it's a violent revolution, a democratic revolution, a foreign war-driven revolution, I don't know how it's going to play out. I hope within my lifetime that China is able to get back to what they were pre-Cultural Revolution. I could imagine a world where we're out there doing counternarcotics operations and counter-bioweapon operations with the Chinese military. That would be a really good reflection of the state of the world. Same thing with Russia. I don't think it's impossible for Russia to become an ally. They've got a long way to go, but it's not impossible. Look at what we did with Japan and the way it turned around. The relationship we have today tells me that just about anything is possible.