Anduril 论题

In order to understand Anduril as a company, you genuinely have to understand 100 years of military history that it is being counter-positioned against. Since the end of the Cold War, we have been constantly tripping over ourselves into the hubris of thinking that we have officially become the only great power left in the world and that things will never change, when the reality is that was a very brief period of time.

Ukraine snuck a million dollars’ worth of drones into Russian territory and was basically responsible for $7 billion of damage. It destroyed about a third of Russia’s long-range bomber fleet. That’s a million-dollar cost to generate $7 billion of damage. That is modern conflict, and we as a nation are not good at that.

Palmer Luckey talks about this a lot. In their very first pitch deck, he said, “We are going to make billions of dollars by saving the government tens of billions of dollars.” They called their shot early and basically nailed it. Military superiority is front-run by industrial superiority. We did that in World War II. The problem is that China is doing that today.

Pacifism really is a privilege of the protected. There is somebody else who is willing to do those things and build those capabilities to protect our right to say, “I don’t want to be involved in that.” Because anybody who can do something and doesn’t do something is ultimately responsible for the outcome.

Kyle Harrison, welcome back to Sorcery.

Thank you for having me.

Thank you for writing a 300-page book on Anduril. We’ve been waiting for this.

I did it just for you.

Thank you. Okay, so what is going on here? When did you start writing this?

1. The Book Takes Shape

We started working on a deep dive for Anduril toward the end of 2023, so it has evolved over several years into what it looks like now. We got to appreciate Anduril first as an investor. Contrary has been an investor in Anduril since the early days, and then, as we started to build out the firm, Contrary Research basically grew out of that.

That’s our public-facing research arm. The goal, rather than writing blog posts about how smart we are—which is the normal sort of VC marketing fodder—is to actually try to contribute something valuable to the ecosystem. Contrary Research has grown to 100,000 subscribers. We’ve written about 500 different companies, and we’ve been working our way through really in-depth reports on some of the most influential companies. We’ve done OpenAI, Stripe, things like that.

Basically, we just worked our way toward Anduril. We’ve tried very hard to be objective, even though we’re obviously biased toward Anduril as an investor. We want all of our research to be fairly objective. But as we dug into it, we realized that in order to understand Anduril as a company, you genuinely have to understand 100 years of military history that it is being counter-positioned against.

In order to do that, it didn’t lend itself to our typical sort of long-form online writing. A book sort of morphed over the course of several years, and then we started collaborating with Anduril on it and it got even bigger. It’s been a very big lift for our team over the last couple of years.

Wow.

We’re very excited to have it out.

So who was involved, and what was the actual process?

We have a research team within Contrary. I run the research team. We have an editor, a research associate, and basically a small army of research fellows who are people working in the industry. They’re engineers, they’re working in investing, whatever it might be. In order to get exposure on the writing side, they’ll come and write with us about things they’re interested in.

So we have this massive army. Over the years, there have been a lot of contributors to that project. I have been the main driving force in trying to get it across the finish line.

In terms of process, it started out very much as me consuming everything there was: reading every Anduril blog post they’d ever put out, every press release, listening to dozens of interviews with each of the individual founding members, and reading all kinds of different reports, white papers, and stuff like that.

I read several books. Once we got into the deeper military history, it wasn’t enough to watch a YouTube video. You actually have to go back and read the literature. A lot of it was just immersing myself in a ton of the underlying resources.

Then the team collaboration really came down to figuring out how to take this mountain of context and structure it into this narrative. That became very much a collaborative process with me, our editor, and several people on the team. Over the course of several months, we got it to a place where we had all the context in the right place.

Then it was a matter of collaborating with the Anduril team and making sure that we were getting the details right—that we were understanding how things actually evolved, technical details, things like that.

Before we go too deep into this, what were the biggest surprises that you took away from either the military history, Anduril, or both?

2. The Military History Context

First, I think the history is the most important context. It’s one of the things missing from the conversation. Everybody talks about Anduril the same way that we talk about Stripe, OpenAI, Databricks, or something like that, as if it feels very similar. The reality is that while all of these other companies have a technological history, most of them don’t have the same deep cultural, industrial, military, and geopolitical history that they’re actively defined by today. Anduril does have that history.

When I think about the history, one of the things that is both the most frustrating and surprising is that it’s just been multiple decades of a self-own. Since the end of the Cold War, we have basically been constantly tripping over ourselves into the hubris of thinking that we have officially become the only great power left in the world and that things will never change, when the reality is that was a very brief period of time when that was true. It’s never been true in history.

Everything about military and industrial policy operated on that assumption, and that assumption was just deeply wrong. It’s honestly heartbreaking, especially when you find these nuggets throughout history. You think, “This guy was calling it in 1997.” He was saying, “This is going to be a problem, and this is going to catch us off guard,” and nobody listened.

On the Anduril side, I don’t know if it’s surprising, but Palmer Luckey talks about this a lot. In their very first pitch deck, he said, “We are going to make billions of dollars by saving the government tens of billions of dollars in costs.” That revenue-cost equation has always been the process that Anduril has pursued as a company.

What is probably most surprising is that they called their shot early and basically nailed it. We’ll talk more about the details, but in their first pitch deck there were basically 5 things. They had all these weird ways to describe things like “VR for warfighters” and “perimeter security on a stick.” All these things sounded insane to their early seed investors, but one by one, they’ve nailed every single one whenever they’ve tried to build it out.

That’s a pretty impressive example of calling your shot.

Going into the 100 years of military history, what were some of the sources that you pulled from?

There are probably 2 big buckets. The biggest bucket is white papers. The Department of Defense is prolific at doing studies, war games, postmortems, exercises, and evaluating different things.

And you can access this?

I mean, you can access all of these.

Really?

These are on the internet. There’s a massive amount of context on the internet. Organizations like CSIS and a lot of think tanks have put out reports on different things.

That was probably the largest corpus of material. You just drill in and say, “Why don’t we build ships anymore?” Then you dig in, and there’s a specific law that was passed 150 years ago and has basically screwed us ever since. There are all these big problems that have caused issues.

The second bucket is some pretty incredible books that have been written by people who were boots-on-the-ground involved in different aspects of these processes. Probably the 2 that we reference the most—actually, 1 is by someone who, when he wrote it, I think was in the process of joining Anduril, but he wasn’t at Anduril as he was writing it. That’s Chris Bros, who’s now the Chief Strategy Officer and Co-President, I think, of Anduril.

Yeah.

It’s called The Kill Chain.

You said he was mentioned—you looked back, and he’s the most-mentioned source.

By far, yeah. I think we reference The Kill Chain over 150 times.

Oh my God.

It is the best introduction to the problem set. The gap this book ideally fills is that there’s so much historical context, but also the need to map Anduril’s trajectory onto that. I would imagine that’s a function of both the fact that he wasn’t at Anduril yet and the fact that you don’t want it to feel biased. You want it to be able to stand on its own: This is true independent of Anduril.

Everything in The Kill Chain is true. What we do is say, “All of these things are true, and there’s additional context you need to know. Now let’s map Anduril against that and say, ‘Okay, this thing is broken. Here’s why Anduril operates this way. This thing sucks. Here’s why Anduril does it.’” Right? Back and forth.

The Kill Chain is by far the most-referenced piece in the book because it’s one-to-one exactly what we’re trying to describe in the first half of the book. The second book is Skunk Works, written by a guy named Ben Rich.

Mm-hmm.

It’s one of the best books in terms of high-performance teams building incredible things. It’s a lot about the journey of building the stealth bomber, but there’s also a lot about the U-2 spy plane. Kelly Johnson is the guy who originally founded Skunk Works.

Lockheed is the company, and Skunk Works is the lab. Kelly Johnson founded Skunk Works, and he is probably one of the most impressive people in industrial history. Those 2 books together, I would say, laid a lot of the foundation for how we got ourselves into such a mess when we were so capable. There are a handful of others that we list in the sources, but those are the 2 books I would say laid a lot of the foundation.

So I guess walk through some of the specifics there. What happened post–World War II?

3. Bureaucracy Strangles Defense

If you set the stage for what things looked like at the end of World War II, the US was more than 40% of global manufacturing capacity—a massive powerhouse. Even in World War II, they say that we basically just outbuilt the enemy. Take the tanks we were making: Our tanks were technologically inferior to the Germans’ tanks, but we built so many of them and threw so much industrial capacity at it that this is literally how we won. We were just a powerhouse industrially.

Then you look at the DoD specifically. At one point during the Cold War, in the ’50s and ’60s, the DoD alone represented 36% of global R&D. Of all the research and development being done the world over, 36% of it was coming from DoD funding different projects and keeping us going through the Cold War. There was this massive capability, and you look at all of the things that we accomplished. We were spitting out one bomber plane, basically, per day—just massive amounts. It was an industrial powerhouse.

Then we get into the Cold War, with conflicts in Korea and Vietnam. But in the ’60s, there was a guy named Robert McNamara, who became the Secretary of Defense, and he came in with the hypothesis that we were just spending too much on defense. While having this industrial powerhouse was important, in his perspective it was too much. We were spending too much, so we needed to control spending.

He introduced this system called PPBE, which is planning, programming, budgeting, and execution. Ben Rich, who wrote Skunk Works, describes it as a plague descending upon the Skunk Works—auditors and bean counters. It was this army of people who came in not just to say, “You need to build this, and you need to try to do it at the lowest cost,” but to literally count every single purchase and every single order. Was it absolutely the cheapest we could get it?

It introduced a disgusting level of bureaucracy, thinking that would control spending. If there was so much friction, it would control spending. It worked for a couple of years. Our spending did go down.

Yeah.

But then we rip into Korea and Vietnam, and we go back into a conflict phase. Basically, McNamara has strangled the golden goose we had of industrial capacity. He’s made it so high-friction to build anything that now every individual project is dramatically delayed and significantly over budget. It’s just so expensive to do anything, and that becomes the new paradigm for building things—

But what makes it so expensive?

It’s the process. When you think about bureaucracy, if I ask you to build something that you don’t know how to build, it requires a ton of experimentation. But if you have a bean counter along the way, every time you try to run an experiment, they’re introducing 3 to 5 to 12 weeks of process to say, “Well, you need to fill out these forms in triplicate to make sure that we’re evaluating these vendors appropriately, and you can’t look at these vendors. You have to look at these vendors because there might be a cheaper vendor,” or whatever.

It’s just so much friction. It’s not any one thing. It’s not any one tax. It is a million taxes of friction. That breaks the process, and now you go into these conflicts.

The problem is that we get used to everything being way more expensive. If we’re going to spend money, spend all this time, and have everything be over budget and behind deadline, then what we should do is just build the biggest things.

During the Cold War, that mentality kind of made sense because we were doing a lot of what you can effectively think of as day-one framing. What you want is the biggest, baddest stick. You want the biggest ships, the biggest Navy, Army, and Air Force, and you want to bring that to bear on day 1 so that conflict is just too expensive. Forget about even nukes; you want to have so much big stuff.

We retrained our ecosystem to build these large, exquisite assets and to pour all of our time and money into that stuff. Fast-forward to today, and that is largely what you have. We’re probably pretty good at building massive aircraft carriers, nuclear submarines, and things like that. But everything is ungodly expensive, way behind deadline, and in some cases literally decades behind deadline. You have projects that should have cost $13 billion or $15 billion that are upwards of a trillion dollars, like the F-35 program. There are dramatic overages. That’s the broken system that we largely have today.

It’s also completely outdated because of how much backlog there is. With the Pentagon asking automakers to start helping build missiles and more advanced drones, there was some citation—I’ll probably get this wrong—but we’re throwing $2 million Ferraris at $50,000 Frisbees.

$50,000, yeah—Frisbees.

$50,000 Frisbees.

That’s right.

There’s just a big disconnect there. How do you see that technological disconnect over time, in terms of how we’ve actually lapsed from what we, quote-unquote, “have been” in terms of superiority in the military-industrial complex?

It is actually even worse than just a cost imbalance. We lack the iterative cycle to stay anywhere close to the cutting edge. To your point, it’s not only that everything is super expensive and super delayed; it’s also not anywhere close to the cutting edge.

For example, if we’re really good at building big, expensive aircraft carriers, the Chinese have developed missiles that they affectionately refer to as “carrier killers.” These still might be fairly expensive missiles—$15 million or $20 million—but they can destroy a multibillion-dollar ship.

That’s kind of at the cutting edge, right? They have such destructive power that has been developed over time, whereas our munitions are dramatically delayed. It’s an imbalance not only because it’s more expensive but also because it’s more old-school.

The problem with that is that if we get into any conflict with China, there’s suddenly a huge swath of our naval assets that we don’t want to put anywhere near China because they could just demolish them. That invalidates even a lot of the exquisite assets that we have. They become invalidated.

We’re really stuck behind. We don’t have as many iterative feedback cycles in the assets, even in the stuff that we are building. Conflict is change.

You look at Ukraine, and I think somebody described it as the military-industrial complex in production. They’re live-testing stuff much more rapidly than even the US at its best, even China or whoever. In Ukraine, we try something, and we immediately field it, and you get a ton of experience—

Well, I think even Neros went directly to Ukraine to start testing.

Yeah.

Which is a good use case example.

4. Ukraine Changes Modern Warfare

I think Ukraine is a good illustrative case study of this new age of modern conflict, where it is increasingly becoming all about high-volume, low-cost assets—what people call attritable assets. It's just this thing you can burn through a lot of and make very quickly.

In Ukraine, there was this—I think it was called Operation Spider Web—where Ukraine snuck a million dollars' worth of drones into Russian territory under the false floorboards of a truck. The drones then took flight, were able to autonomously identify high-value targets, and engaged them. They were basically responsible for around $7 billion of damage and destroyed around a third of Russia's long-range bomber fleet.

That's a million-dollar cost to generate $7 billion of damage. That's modern conflict: you have to be capable of spitting out a high volume of these low-cost assets to be able to engage, and we are not good at that. We as a nation are not good at that.

Taking the technology point out of the equation, are we capable? Have you been seeing that with the new administration and the DoD, and how they're embracing this new wave? It's also part of the new administration. How are they embracing this, and how is policy changing?

The good news is that the defense bill every year is one of the most bipartisan passages each year. It is representative of the fact that it actually spans partisan politics; it's not that Republicans love defense and Democrats hate it or whatever.

In 2022, even President Biden openly acknowledged that China represented an existential threat, which, candidly, even 5, 6, or 7 years ago, was not openly talked about. It was kind of taboo. We didn't acknowledge that risk.

I think increasingly what has changed is that it's less about the current administration doing something unique. I actually think you go back to Ukraine as a wake-up call for a lot of people, or even Chinese posturing in the South China Sea with Taiwan. A lot of those actions are acting as wake-up calls for a lot of people.

I think it's more a function of people acknowledging that there is conflict, whether we like it or not, and we do have to be capable of responding to it. We continue to have a lot of the largest things or biggest things on the scorecard, but in practical terms, we are not the biggest dog at the party. We have to acknowledge that there are some strategic changes we have to make to stay relevant.

We put out this clip from interviewing Chris Power at Reindustrialize last year. He talks a lot about the Davidson window, which is the time in which China is going to attack Taiwan. He makes it part of his mission to make sure that we're able to fulfill manufacturing capabilities to meet that, and it seems like we're getting into a contested area.

Going back to Ukraine as the wake-up call, how does this then affect timelines and what's coming up next?

I do think that, when you look at increased conflict—even over the course of this year—one of the hardest parts about writing the book was that, from the end of 2023 through the end of last year, we were iterating on it. In that time alone, you had escalation in Ukraine. You had Israel and Palestine happen while we were writing; October 7 happened. You had Trump's and Israel's actions in Iran. You had things going on in Venezuela.

Even just over the course of writing the book, global conflict changed shape. So that is—

Whenever a new war breaks out, do you have to go back to the drawing board and edit? What was that like?

It was actually dueling forces of urgency. On the one hand, the global climate was changing frequently, and there were frequent things we had to change.

At the same time, Anduril is probably one of the most prolific shippers of products: new iterations of things and new versions of stuff, entirely new projects, things that they weren't talking about that we thought we had a sneak peek at, and then by the time we got around to putting it in print, they had already talked about it.

Anduril's pace of execution against the backdrop of a global environment that's changing super fast meant that we were on draft 35 by the time we said, “Okay, lock. We have to—at some point, we have to actually print the book.”

Mm-hmm.

It was a lot of iteration.

Okay, so let's get to it. The book is literally The Anduril Thesis. What is The Anduril Thesis?

5. Anduril Builds To Mission

The core crux of the thesis behind Anduril is understanding 100 years of military history and then systematically walking through every aspect of Anduril that is deliberately counter-positioned to the way that the Defense Department and the US military are set up today.

That includes how we acquire new capabilities. There are things like fixed-price contracting, which is what Anduril does, versus cost-plus. There's a rich history there of why we do that and why it made sense briefly, but why it doesn't make sense anymore, even though we continue to do it.

There's also why the DOD believes in a sort of built-to-spec approach: “We will tell you exactly what to build.” That grew out of the ’40s, ’50s, and ’60s. As I said, the DOD was 36% of global R&D. It had a pretty good handle on what the cutting edge was when it was controlling almost half of all the latest technology. Silicon Valley was literally born out of the defense industry.

At that time, that made sense. Today, the DOD is responsible for less than 1% of global R&D. So the DOD doesn't really know what the cutting edge is. Palihapitiya has this great line where he says, “We absolutely want to listen to our customers because they understand their problem. We don't necessarily listen to them about what the solution should be, because they don't have a good framework for what the possible solutions are.”

Where the military is trying to acquire things built to spec, the goal of Anduril is to acquire things built to mission. They fund their own R&D. They put their own money up rather than asking the government for R&D dollars to go spend and build some brand-new thing.

Anduril is saying, “We're going to put our capital at risk, and the goal is that we so effectively satisfy the mission that you will want to pay for it without you having funded the R&D yourself.”

There are so many aspects of Anduril as a business that you don't appreciate as important unless you go back and map them to the 100 years of history that have gotten us into this wormhole mess.

What did you learn about the actual roadblocks and bottlenecks within government and DOD contracts? Did it come down to, “They already did thousands of pages of paperwork with one supplier, so they're just going to stick with that one. They don't want to bring on new ones, and that's why they're not trying new, innovative things”? Or is it really cost? What is the problem there?

The shape of the innovation block in military contracting—defense contracting generally—has really flipped on its head.

In the early days, when we talk about McNamara introducing all this friction, you had people like Ben Rich who, over the course of their careers, might work on 20, 30, or 40 iterations of different types of planes and bombers and all kinds of different things. You get to this place where it becomes standard that you'll maybe work on one over the course of your entire career. Everything just slows down.

Now, what you have is not existing primes as an army of Ben Riches who wish they could be unfettered by this bureaucracy. It is people who were born and raised in the bureaucracy, so they like it. It's comfortable. They have reshaped their entire business to be purpose-built for that bureaucracy.

You go back to the fixed-price contracting that Anduril is trying to push and say, “We want to be rewarded for our outcomes.” The standard is cost-plus, where you say, “Whatever it costs you, we will guarantee a certain percentage.” It's usually high single digits, or 10% or 15%. We'll just tack on that margin.

Think about the incentives in that. If the incentive is, “No matter what it costs you, I will give you 10% on top of that; that's your profit,” then 10% of $3 billion is a much bigger number than 10% of $150 million.

As a traditional prime, you are deeply incentivized to reinvent the wheel from scratch every single time. You're not going to pull anything off the shelf. You're not going to use any commercial products. You're going to go to the government and say—what you're asking for is a multibillion-dollar project, and maybe it's not even possible.

So take Anduril’s first fielded product, the Sentry Tower. Originally, I’m trying to remember. I want to say it was Boeing, but that doesn’t sound right. Maybe it was Lockheed. There was a contractor that had a contract back in the early 2000s with Customs and Border Protection to build effectively these digital fences to observe and monitor the border.

They said, “We think this is going to cost $5–10 billion, take 15 years, and probably defy the laws of physics. We can’t do it.” They tried and tried and tried and spent a ton of money on it, and eventually they just said, “Sorry. The military is trying to get us to break the laws of physics, and we just can’t do it. No one can do it. It’s impossible.”

Their incentive is to take as long as you possibly can, make it as expensive as humanly possible, build everything from scratch, and get this fat fee on top of a much bigger number, right? Then the Anduril team comes along and says, “Well, look, you don’t need to reinvent sensors, cameras, and batteries from scratch. There are some very effective tools in the commercial market. Let’s get cameras, sensors, and batteries, build a pack, put it on a stick, and stick it out on the southern border.”

Their first customer, if I remember correctly, wasn’t even the government. It was a private ranch owner who wanted to monitor his land because drug traffickers were coming across it. Originally, it was a private citizen paying to put up these towers so they could demonstrate the product.

Eventually, the Department of Homeland Security and others became large customers of the Sentry Tower, and it was dramatically cheaper than anything the prime contractor was trying to do. It didn’t have to cost billions of dollars because Anduril was incentivized to make it as cheap as possible, as quickly as possible, and as effective as possible, then sell that prepackaged capability rather than trying to convince the government to pay them $1 billion to play around with the possibility of a product.

So the Sentry Tower was the first product that they went after. Can you talk about the early days of Anduril and how the team got together? I was really fortunate—I don’t know how this happened—but we went over to Anduril and spent the day there. I was able to interview 4 of the co-founders: Matt Grimm, Trae Stephens, Palmer Luckey, and Brian Schimpf. It was amazing. It was so much fun. We did literally everything, and I got a little piece of history from each and every one of them. To your point, and to your research, too, what is the outside lens on how things got together and the climate around that?

6. The Founders Build A Prime

One of the things that is most fascinating about walking through the history is that it really is a deep vein of rich mythology. The deeper you go, the more impressive it is. One of the things you find out is that Palmer has been thinking about defense since he was in high school. While he was in high school, he volunteered at a lab where they were doing research on how to use virtual reality to help soldiers deal with PTSD. There was this combination of his love of VR and military technology that goes back as far as the pre-Oculus days. There was some basis there.

Then there’s this other vein of history. Palmer is one of the younger of the 4, but the other 3, I think, were all in high school during 9/11. Trey Stevens talks about that the most. He saw what happened and said, “I need to do something to be of service to my country.” When he joined the intelligence community—he deliberately wanted to go to Georgetown and work as a spook—he did that because of that early patriotism.

Yeah, he said he had a James Bond dream, and then all of a sudden he was trapped in a cubicle in a basement.

That’s right. Each of the founders has this early exposure that sparked the passion. The real linchpin is that Trae was at Founders Fund. Founders Fund had invested in Oculus, and there was an event that Founders Fund threw where they invited Palmer to come. Palmer and Trae met, and I think the first idea they batted back and forth was, “How could we build a force field?” That was the first idea.

Naturally.

How could we? How would we do that? Eventually, that evolved into somebody needing to build a real-life Stark Industries. There are other inspirations, too. Palmer was inspired by an anime character whose name I can’t remember—a billionaire who builds video games with the money he made as an arms dealer. I think he has a dragon, or a jet, that he flies around in. There are all these little passion projects and threads throughout each of their histories.

So the reverse of Palmer.

He did it backwards. That’s right. He did it backwards. He was literally inspired by that guy. After Palmer gets fired from Facebook, he says, “Okay, I’m going to start another company.” There were a handful of things—and this is part of the deeper mythology—that Palmer thought about doing, but building a new defense prime, a Stark Industries, was at the top of his list.

He called up Trae and said, “We should do this.” You have Trey, who is the Palantir thread. He pulls in Matt, whom he sees as the operational-excellence genius, and Brian, who is the engineering neo-mad scientist. Then Palmer pulls in Joseph, the other co-founder, who was an engineer at Oculus and, I think, the only one of the 5 co-founders to have served in the military.

They all came together from their prior experiences, and the story is that they got some Chick-fil-A and made a bad Google Slide. They sat down with a handful of people they had met in the past—basically, a recruiting lunch—to lay out exactly what they were going to do. This, I think, is where they had this famous experience: they laid out 5 things. I’m not going to remember all of them, but there were 5 things they laid out and said, “These are some of the things that we could do.”

Two of them are good bookends for Anduril’s current product portfolio. One was “perimeter security on a stick,” as they described it. It was literally what became the Sentry Towers: a camera, sensors, and a battery, all put on a stick. That was their first commercial product.

Another one of the projects was, I think, VR for the warfighter. At the time, in 2017, Microsoft was competing for an IVAS contract, which stands for Integrated Visual Augmentation System—basically, augmented reality for warfighters. Palmer desperately wanted to compete for it because he had been doing Oculus. Defense and VR were perfectly suited together, and some of his early investors convinced him that he needed to bide his time.

They told him that it was going to be a big thing to chew on and would take everything they had to compete for, versus the rapid-iteration process they needed to engage in. So he let it go. Microsoft got the project and did pretty poorly for several years. Then, a year ago, Microsoft basically admitted defeat to some extent, and the government said, “We’re going to recompete the contract.”

Anduril first came in as a potential partner to Microsoft and then eventually took over a $22 billion IVAS contract, which has now become their EagleEye product. It’s a ballistic-shell helmet with augmented vision. There are crazy videos of use cases where, with a combination of augmented reality and a Ghost drone flying overhead, you can literally see through walls. If there’s somebody coming back through this thing, the drone knows where they are. It can display that visually as they’re moving from behind the wall.

That product portfolio has expanded dramatically, in large part because from day 1, it was not about 1 individual product or 1 individual platform. It was about this software-defined, hardware-enabled web of products that can all work together and focus on the capabilities they’re trying to enable.

Could you walk through the timeline of the products and how they developed? They have quite the list of products now. It was so much fun to go see all of them, too.

Yeah. Fun fact: do you know what their actual first product was? They never fielded it or commercialized it, but the first product they ever built and prototyped was called Sentry. It wasn’t the tower.

It was an autonomous firefighting tank that would roll into a forest fire and deploy fire suppression, so they could actually save people and stuff. That was the original product: trying to figure out what they could do with autonomy to be able to seal a contract.

7. Lattice Connects The Arsenal

What's funny about the way that their product evolved is you can see these early days of them throwing a lot of stuff at the wall. But Lattice was this consistent through line in everything that they did. And so the reason that they talk about it, even though everybody kind of knows the drone or the munitions or the submarines or whatever, the real core product is...

A couple of people talk about it in different ways. DARPA has this great phrase where they describe it as mosaic warfare, pulling all these things together. Chris Bros talks about it as a military Internet of Things, having all these edge devices we've woven together.

But basically, it very much is this hub-and-spoke model of Lattice as an AI, sensor-fusion-driven brain. And then the goal is, across every domain and every capability and every force level, how can you plug as many assets into that brain as possible? So you start with Lattice.

One of the advantages is you can start with just monitoring. We just need to get information into the system. And so Sentry becomes the first product, and it's a sensor that can pull everything in. But still, Lattice is there to be able to identify and say, "Hey, this looks like this kind of person," versus, "This is a known thing. This is a dangerous thing. This is just a fox." Whatever, right? So there's actual intelligence there.

From the sensors, you start to expand out into UAVs, so you're getting into actual flight. It's still a lot about data gathering. So they had a quadcopter that they built, I think you saw. And I remember in your video, you saw the early prototype pieces of it with Matt and stuff. Again, it's a very similar idea: it's just about gathering information, but to do it from different angles and with different flexibilities. So you have a quadcopter, you have a Ghost drone, just a larger, longer-distance drone.

Then eventually you start to get into counter-drone systems. And so you have systems where you not only want to go observe; you actually want to be able to intercept a drone. Going back to the throwing Ferraris at Frisbees, the goal is for that to ideally not be a multimillion-dollar asset taking out a $20,000 Walmart drone or whatever.

And so they start to get into some of these loitering munitions. Altius is a product that they bought, and it's basically a tube-launched drone that can circle, identify, and then engage with a drone to take it out. And then you quickly see them expand into, "Okay, we have capabilities. We have sensors. We have monitoring and stuff like that. We're aerial." But now you get into actual munitions.

So you have things like Roadrunner, which is a payload-carrying asset that can destroy different threats and stuff. Originally, they had a Sentry tower that was near the ocean. Eventually you deepen the Sentry product suite into the Sentry Seabed, which is a sensor on the bottom of the ocean.

And you can talk to whales with it.

And you can talk to whales, which is the ideal use case, I think—

Yeah.

—is number 1. Get them on our side.

You start to expand out into all these different... And then you get into things like Dive, their Dive line of autonomous undersea vehicles and stuff. And now increasingly, what is most impressive is that what started as very much a product suite of adjacencies—these are all things that will augment a lot of existing assets—we're now getting into potentially asset replacements. So there are things that we can do instead.

So Copperhead is a torpedo product, and it can replace existing torpedoes, but it is also autonomous. So you have this ability to effectively loiter undersea, identify targets and things, and then be able to engage.

But you also have, I'd say, probably their crowning product today. I think there's a lot of things that are on the come that are going to blow people away, but I think that Fury is this very impressive capability because it is effectively an autonomous fighter jet. And one of the things that is most important about that is that—

Also, by the way, most of America is completely removed from the military and the industrial complex around that, so they don't actually understand the training, the day-to-day, the actual combat that goes into being a fighter pilot.

Mm.

You're literally dodging missiles in active war zones.

Mm.

And that's true of the majority of technology-enabled roles in the military. So you look at Fury, as an example, it's not only replacing in some cases—because there are some missions that we no longer need to send pilots into danger; we can do it autonomously—but there are some cases where you still want human judgment and speed and perspective.

But you can extend a pilot's capability with a small set of Furies that can go out and go further and identify danger. They can observe things. They can engage in a way that doesn't put the pilot at risk. That's a capability.

But even things like just monitoring, right? We've had drones for a long time, but when you think about a drone's capabilities, if you were to try and do something like send up a drone, have it go a couple of miles away, observe a target, identify where weak points are, where there are things to be aware of, and come back and report on that information, that could take a room full of a dozen people having to switch between several dozen different screens and messaging people.

Chris Bros talks about this in The Kill Chain, where he's seen these distracted, stressed-out, exhausted individuals who have to constantly be trying to say, "Okay, well, this is the data point I need here. I need to send it to this person because that's the person who's going to control the messaging to the force that's engaging here," or whatever. That is not going to lead to ideal outcomes.

Yeah.

And so autonomy introduces this incredible capability where we are going to do the things we have to do. Whether we like it or not, we need to do those things. This way ensures that the outcome is very likely higher quality and puts significantly fewer people at risk in the process.

Sorcery is brought to you by Brex, the financial stack trusted by more than 30,000 companies, including one in three venture-backed startups in the US. Nearly 40% of startups fail because they run out of cash. Brex is literally built to help founders avoid that. Unlike traditional banks that let your money sit idle, chipping away at it with fees, Brex is designed to help you spend smarter and move faster. Their all-in-one solution combines checking, treasury, and FDIC protection into one powerful account. You can send and receive money globally at lightning speeds, get 20 times the standard FDIC coverage through their partner banks, and even high yield from day one. With same-day and even same-hour liquidity, access your funds anytime. Companies like Scale AI, DoorDash, ServiceTitan, HIMSS, Anthropic, Flexport, Robinhood, and Plaid trust and use Brex. Start today at brex.com/sorcery. That's B-R-E-X.com/sorcery. Turing is training the next generation of AI with tasks that require real expertise and real-world judgment. That's why companies like NVIDIA, Anthropic, Salesforce, and Gemini partner with Turing. Turing builds realistic reinforcement learning environments and data systems based on real operational traces, the kind of infrastructure frontier labs need to train super intelligence. Visit turing.com/sourcery. So I'm curious. I've had people wonder—and I don't know if you know this information or not—which parts of these products are actively ready to use? Which ones are actually ready to be deployed?

The biggest proof point of Anduril's capabilities is when you look at Ukraine. Palmer says that Anduril has had products on the ground in Ukraine basically since day 1. They were over there pretty quickly, deploying assets and stuff like that.

I think that one of the things that is most important to understand about Anduril's go-to-market is that what you're doing is trying to enable specific capabilities for both the U.S. forces and our allies, right?

And so you look at things like Ghost Sharks in Australia. Anduril went to Australia, set up a factory specifically for production of DIVE assets—these autonomous undersea vehicles. They specifically set up shop to produce those because Australia represents...

AUKUS is the collaborative group of Australia, the U.K., and the U.S.—that collaborative force. Australia is the tip of the spear for that effort because they are right there in the Pacific, close to China. Taiwan is nearby.

Their need for an undersea capability is the strongest. And so you go set up shop in Australia. That contract that they have—I think it's a multibillion-dollar contract at this point, if I remember correctly, but it's a massive contract with the Australian government, maybe a couple hundred million. The whole point is to be able to build things there where they are most needed, and there it is getting fielded significantly. It's a capability that they need today.

There are a handful of examples of that where there's a very specific need. Probably the Sentry Towers are the other example. It's not only now on borders, but also things like military bases throughout the globe. You have Sentry Towers protecting those military bases.

There is a spectrum, a curve—an adoption curve—of capabilities that Anduril has. What can they capably do? Then there is what Anduril can capably do that also matches a near-term need. Then there is what Anduril can do that matches a longer-term need, where it may not be a thing yet.

For example, on the munition side, they're rapidly spinning up Arsenal One. They're going to be building Furies in Ohio, right? And so it's like that—

They did that on time and under budget.

That's right.

Actually, ahead of schedule—

That's right.

—and under budget.

Same thing, I think, with the Ghost Sharks, right?

Yeah.

Anduril—I remember Palmer tweeted about it: When's the last time you saw somebody on time and under budget, right?

Yeah.

They've done it a couple of times. But a lot of those things are not necessarily things that you are actively deploying in live combat zones. They are things that you have to build arsenals for—

Mm-hmm.

—in order to, in the future, engage in specific conflicts. I think that's a thing that people misunderstand: These large contracts, these large dollar amounts, and these large builds that Anduril is doing are not always because we're going to build a batch of drones and send them immediately to Ukraine for deployment. It's that we're trying to build these depots.

I think the other thing to understand is that we are staring down the barrel of a potential large-scale conflict with China. When you look at the existing munitions that we have, for example, if we were to engage in a large-scale conflict, I think it's something like 11 days—we would run out of munitions in 11 days.

Yeah.

And we wouldn't have the capability to rebuild them, right? In war games, I mean, we have been effectively losing war games when they model the scenario of the U.S. versus China since, I think, 2015. The U.S. loses those war games.

A big part of that is because our naval assets are at risk. We could get, I think, half of our naval fleet destroyed within the first couple of months of a large-scale conflict with China, and it would take us 25 years to rebuild, just to get back to where we were. Our shipbuilding capacity doesn't exist. Our munition—

Meanwhile, every ship in China is required to also be dual-use.

That's right. And really, everything in China is regulated to be dual-use, right? Military-civil fusion means anything you build, we immediately get to weave into the military.

Yeah.

I think it's important to acknowledge that the U.S. maintains that freedom and ability. Trey talks about this a lot: I'm grateful to live in a country where, if a certain company or certain individuals don't want to work with the military or don't want to work on military projects, great. I'm glad that they have that freedom.

But pacifism really is a privilege of the protected. There is somebody else who is willing to do those things and build those capabilities to protect our right to say, "I don't want to be involved in that," and that's fine. China does not afford that right to its citizens and its companies.

Yeah. I think he said in our interview that pacifism couldn't exist without a monopoly on violence.

So, to dig back into Seabed Sentry a bit more—and even DIVE—what are the biggest misconceptions about undersea warfare? I know Palmer is really interested in, what is it? Under—

Subterranean.

—subterranean warfare. But let's just talk about the sea first. What's going on there, and why is that so critical?

8. The Undersea Battlefront

It is all about surface area, right? How much surface area you have represents the threat vector of a specific domain. When you think about the air, it has a certain amount of volume. When Palmer talks about subterranean warfare, it is actually probably—I think it's the largest domain outside of space. But it is largely untraversable today. Several people are working on that, including Anduril, much to the surprise—

I don't know if you have heard this story, but there was a story of Palmer talking about subterranean warfare, and the marketing person at Anduril was there and said to the journalist, "Just to be clear, we're not working on that." Palmer was like, "Oh, no, we are. We've been working on some stuff." So even Anduril is working on it.

But the ocean is the largest surface area that is addressable outside of space, and one of the reasons that it is so important is that it is one of the highest-friction domains to engage in. When you have a large land space, there are a million different ways that you can monitor it—satellites, sensors, radar, and everything. The ocean is really difficult to navigate and understand and make sense of, in terms of where everything is. It's just really freaking big, so it's really difficult to pinpoint things. Even if you can identify and articulate them, it's really difficult to identify them.

The second reason that the ocean is so important—and this is probably another aspect that most people don't appreciate—is that there are massive undersea data cables. Originally, a lot of these, if I remember correctly, were invented and created by Bell Labs. Go back to this massive industrial capacity in the '50s and '60s and so on: They literally invented this way to digitally connect different parts of the world.

Today, I mean, it's literally—I remember, the numbers feel so stupid that I always feel like I'm getting them wrong—but I think it's something like $10 trillion a day, maybe an hour. There are massive amounts of transactional data, volumes, information, and everything transacting across these cables that are just sitting at the bottom of the ocean. A lot of them run up through the Baltics. They're literally just exposed for Russia to do what they will with them.

If you wanted to interrupt the global economy, taking out some of those cables is a pretty effective way to do it, right? You could literally knock out transportation, energy grids, data transfer, satellite communications—there are so many aspects of the global economy you could just knock out.

You have to understand that it's not just individual assets where we want to be careful of enemies, or less-than-friendlies, who have submarines or whatever. We also want to understand those. But we have to map a massive volume of the ocean surface to be able to protect those assets.

A combination of Anduril's products in the ocean-faring realm comes together for that defense. Starting with the Sentry Towers that they had, which were in proximity to the ocean, now they have these Seabed Sentries that are just incredible.

When you think about difficult environments to build a product for, candidly, in some cases the ocean is actually harder to build for than space, because space has a lot of neutralizing effects. It's just there, and as long as it's maintained, it can be maintained. I mean, you have satellites that have been in orbit since the '60s, whereas the ocean is relentless in how hard it is to field an asset there.

These Seabed Sentries are incredible products. They also generate a significantly difficult-to-achieve level of insight, allowing you to understand what's going on.

In terms of what you do as a result of that information, the entire DIVE line of products includes some that are for monitoring. Increasingly, they have munition variants of their DIVE products, so they can carry weapons. They have things like Barracuda, which is a small autonomous underwater vehicle that you can maneuver and that can carry munitions. They also have Copperheads, which are their torpedo products. They have munition variants that they can do things with.

And then, just recently—this is what's awesome about Anduril, both as a cultural artifact and as a massive products company—the last couple of product announcements they've made, they've made with animatics, like—

Yeah.

Like prototype videos. I don't know if you knew this, but in the last 2 of their videos, they have teased future products at the end. I think, if I remember correctly, the first one was for Fury. At the end of the Fury video, the camera panned under the ocean, and you had these products moving, but you didn't know what they were. Then it cut away.

Six or 9 months later, they announced the Copperhead products, so that was what was sort of teased. The video for the Copperhead products literally picked up where that video ended. It shows these Copperhead munitions moving around. They can be navigated and deployed from a Dive-XL, and then they can loiter and stuff.

What happened in this video is that you saw all of these assets moving toward a target underwater. Then the camera panned up above the water, and you saw a slight hint of an above-water vehicle. Then it cut away.

Just this week, Anduril announced a partnership with HD Hyundai where they are officially starting to build autonomous surface vehicles. They're starting to compete with folks like Ocionic and stuff like that. It goes back to this idea of mosaic warfare, of a military Internet of Things: You want every domain, every surface, and every force level to have some relevant asset that you can weave into this network of potential responses that somebody could control and command.

Wow, okay. Key takeaways: Don't mess with the undersea cables.

I want to weave back into the history. You studied 100 years of history. That must have taken a really long time. One of the key parts that you took out of that was that throughout these generations and waves, these founders are often paired with mavericks. Can you walk through some of those and how that relates to the current day?

9. Military Mavericks Return

This is a concept that I would say isn't unique to us, but it is expanded in the book and sort of hinted at in The Kill Chain, Christian Brose's book. The example they give is Bernard Schriever, who was put in charge of inventing intercontinental ballistic missiles, or ICBMs, which was something that we absolutely did not know how to do.

This was a missile that needed to travel basically across the entire planet within a matter of minutes, carrying munitions in a targeted way. It was a very difficult technical lift to pull off. As they were building ICBMs, President Eisenhower was this military maverick.

If you think about Schriever as this founder-style builder, there are a lot of obstacles to working with the government and building such a complicated thing. It's expensive, it's complex, you have to experiment, and it's dangerous. Eisenhower ran a lot of interference to make sure that they had the assets and resources they needed.

This one-to-one pairing of people who are doing hard things and military mavericks who are protecting and enabling their ability to execute has been a very important capability throughout the history of military innovation in the US.

You look back at modernizing the Navy—William Moffett, I think, was his name—and he had President Hoover. You had Kelly Johnson at Skunk Works. He had Richard Bissell, I think was his name, working for the CIA, who basically made the U-2 spy plane possible.

You have a lot of these things that, if left to bureaucracy, probably would never have gotten off the drawing board. Because you have these military mavericks who are willing to enable people to innovate, that's where things actually get done.

Anduril, I think, introduces the founder DNA. You didn't have a lot of people who were building truly innovative products for defense use cases. You had a lot of traditional primes, but they were mired in this broken system of inefficiency, so you didn't really have founder-level energy. Anduril brings that founder-level energy.

What's awesome is that you had a generation of people who would have been willing military mavericks. You had these prophets of urgency. In the '80s, '90s, and early 2000s, you had different admirals and military leaders saying, “Warfare is going to change. It's going to start to look like this high-volume, low-cost style of conflict. It's going to be significantly more intelligent. We have a lot of hubris, and we're building these big assets that are not going to be able to carry their weight.”

You had people saying that, but nobody was listening. Nobody was doing anything about it. Those people would have been military mavericks if they had had either institutional buy-in or founder-level innovation solving the problems.

One of the problems with complaining and trying to send up a flare is that if you don't have a solution, it's difficult to rally people. You also didn't have a lot of urgency because you didn't have China rising as a great power, the potential for conflict, and stuff like that.

Today, you have Anduril bringing the founder energy, great-power conflict rearing its head, and, importantly, people in the military. There are countless people the Anduril team would point to as saying, “These are the individuals who have been really valuable compatriots, unlocking the things that we need to actually adopt these capabilities.”

A really good example of this that I have just read and watched a bunch of podcasts about was Shyam's book, Mobilize, on Hyman Rickover. The point was that Hyman Rickover was the key pioneer behind the nuclear submarine, and it took a lot of fighting, founder mentality, and pushing back against a system.

He was also a strong-willed person with a strong personality. It was the same kind of fight: working against a system and being innovative ahead of your time when everyone's telling you no, but doing it anyway.

He also made the point, similar to the same theme, that America is actually not the worry of inflicted homicide. I hope this doesn't get bleeped out on YouTube, but the US is more vulnerable to suicide because we keep shooting ourselves in the foot with these things. Again, I might have to take that out because YouTube kind of sucks.

I thought that Shyam's book, Mobilize, was quite good and had a really nice perspective on that.

There's actually—I don't know, this might go into the bleeped-out part—but that concept comes from an Abraham Lincoln quote.

Mm-hmm.

He says, “If even the greatest army in the world were to form against us, with a Bonaparte at its head and the treasures of the world funding it, none of them could take a drink of water from the Mississippi or whatever. None of them could cross the Alleghenies.”

None of them could have an impact on the US because the US has such a unique capability in history, geography, and among nation-states. The US has a lot of unfair advantages. What Lincoln says is, “If we are to die, we are to die by suicide,” where it will be our own undoing if we fail to generate the capabilities that we need to be successful.

Yeah. Bringing it back to today, we are in the throes of an amazingly progressive new administration, one of the changes being the Department of War. We've now rebranded, and we're the Department of War.

I've had the opportunity to interview the Under Secretary of Department of War for Research and Development, the CTO, lots of titles there—Emil Michael. One of the most amazing things about talking to him was how he is rethinking that organization entirely. He's collapsing unnecessary parts of it, streamlining it, and making sure there aren't 14 levels of bureaucracy, so we're focused on the main critical points.

He even put out critical areas of interest where he's going through different categories. He and Hegseth are actually going to these new sites of all these competitors. People say we're in a defense bubble. Well, it turns out it's kind of a good thing because we're at a point where we're looking for automakers and new people to make missiles and help supply the military complex.

I think that while there is a lot of doom around our positioning and how it all led up to this point, thank God we had this kind of renaissance in defense tech, because there are new entrants. There are people who have been working in the military for a while who are now taking a technological bent to it and working with VCs and different kinds of funding sources, using their expertise and on-field experience to create this new generation of companies.

There is a long tail of companies that have received these government contracts, whether it's Palantir, which has gotten a lot of them, Anduril, Sironic. I just did a series at Skydio. There is a long tail of these companies, and then there's also a long tail of these new-generation companies that are trying to build off an American supply chain rather than relying on adversarial nations, such as China, which touches a lot of things.

But to your point, you also research all of this. What kind of companies and what other broader themes are you seeing within the tech ecosystem?

I think that one of the most important things is that the DOD starts—the Department of War starts to put up real, legitimate adoption of specific solutions. I do think that the defense bubble that we're in is very timely and very important. There's a great book called Boom by Burn Hobart.

It's good.

It's a very good book, and it talks about this idea that bubbles are necessary to fund innovation, and that is true. There are also a lot of huckster speculators. There are a lot of wannabes and fakes and stuff like that. There is also a lot of capital destruction. Even if there are legitimately earnest people who want to build for defense, they will be unsuccessful, and there will be capital destruction in that.

I think one of the most important things is that you build this ecosystem that can actually put up solutions, put points on the board, and prove out some of those solutions. One of the obstacles is that you're talking about cutting out a lot of the bureaucracy. That is accurate; that is a problem. And so there's a lot of streamlining you can do.

But another thing to think about is that a lot of the large purchases and assets, things like aircraft carriers and stuff, are not necessarily streamlined. There's one specific example—I think it was the F-35 or something. It has components from 1,900 different vendors spread across 350 different congressional districts. That is not the most efficient way to build anything, right? There's no way that that is streamlined.

And so one of the things I think that Anduril is leading the charge on is that you have to build not just capable products; you have to build capable supply chains. And when I think about what people need to enable in order to fund this defense bubble, yes, it is great to have companies that want to be new defense primes. If you are ambitious and strong-willed and are effective at not only building products but also selling to the government, well, the bummer of any limited conversation about Anduril is that there are so many things you can't talk about. But people like Matt Steckman, who's their CRO, are among the most effective people alive today at selling to the government, right?

And people like that—if you have those people, yes, you can be a new defense prime. But there aren't going to be 30 new defense primes. There are probably going to be a very small handful of companies that can check all the boxes you need to be able to do that. Obviously, I think Anduril can be one of them, and there may be a few others, but the vast majority down the long tail are people who are building specific capabilities that can plug in.

Packy McCormick talks a lot about this: Anduril can be almost like an API into the Department of War, where they have this established ecosystem. There are a lot of examples of companies they've been able to acquire and then immediately plug into that system, where the company independently would never have been able to get the same type of contracts and success and things like that. So there are going to be some companies that are building specialized capabilities that plug very well into the Andurils and the neo-primes. But the even longer tail is going to be people who are building supply chains, manufacturing capabilities, industrialization in general, and that is not going to come from defense alone.

Yeah.

You have to build this dual-use sort of industrial base because when you look at the history of defense, every single time, military superiority is front-run by industrial superiority. We did that in World War II. The problem is that China is doing that today, and if we're going to have any semblance of a chance in any potential conflict, we have to have some commensurate industrial capability.

It doesn't have to be exactly the same. We don't have to build the exact way they have built. We don't have to spend the next 30 years replicating their way of building. You look at Anduril as one comp for how we can build things significantly more efficiently. We have to do a ton more of that.

Going back to World War II, when we were using different manufacturing facilities—automakers, cereal makers, literally everything—to help build up our capabilities, the thing is now, with the Pentagon asking for that same call-out, to your point about supply chains, those companies don't have the supply chains ready to fulfill any type of order of that capacity.

Yeah.

But I want to go back to something you mentioned. One of the parts of Anduril that I was exposed to was its talent density. It is so clear. And when I've been at many of these—I've been at so many defense forums and conferences at this point, whether it's Reindustrialize, which touches on it, the Reagan National Defense Forum, or Hill & Valley as well—between all of those conferences, there was one point that I never considered before because I'm not in the throes of this.

Ross Fubini, who I had on for Hot Take, is the founding partner of XYZ Venture Capital. He said at Hill & Valley that there are probably 40 people in the U.S. who are capable of selling to the government. It is extremely hard. That is a clear bottleneck for lots of these companies and the ecosystem in general. Matt Steckman is one of those 40 people who's capable of selling to the government, and he's very good at it. Anduril has built out this world-class team of talent. Is there anybody else at Anduril who caught your eye when you were doing this research, where you thought, “Wow, I can't believe they grabbed that person”?

I think that one of the things that Anduril does well is that there are probably 2 buckets of people that they pull from. One is just a mission-driven foundation of sharp engineers. Palmer talks about this: the way to recruit really sharp people who are working in tech is to go to them and say, “Hey, probably the thing that you are working on does not really matter. Maybe you're making a ton of money, but you're optimizing buttons and ad spend for massive platforms and stuff. Is that what you want to do with your life?”

There's another one of our portfolio companies, Base Power, in Austin, where they have a campaign that says, “Don't tell your grandkids that all you did was B2B SaaS,” right? There is this mission-driven inspiration that the foundation-layer people really resonate with.

Then there's a second layer where I would describe it as very strong capability. And so you look at—I can't remember his name—but Anduril, in setting up Arsenal, went and got the person who, at Tesla, built out the manufacturing. The head of manufacturing at Anduril came from Tesla, has seen this streamlined experience, and is applying the same logic to the way that they're building Arsenal.

You look at, from a strategic perspective, what projects do you want to field at what time? You look at Chris Brose and his experience working with the Senate Armed Services Committee and stuff. Those types of people. You're finding these real hotbeds of very specific types of knowledge and experience.

And I think the reason that the Anduril team resonates with that kind of background is because that is the insight that they had, right? Palmer's experience came from building hardware. He is very capable at that and can plug in that capability, but it's the same thing as having a team that is largely coming from Palantir.

I think even in your interview with Trey he said this: Anduril, on day 1, got to start where Palantir was in year 10. That level of experience and knowledge is the kind of specialty that they had and resonated with. And so when they go to find people to plug into specific capabilities, those are the types of people they build with.

And I actually think that informs a lot of their acquisitions as well. You look at the companies that they acquire. You're basically saying, “Are they an Anduril competitor? No. They are a specialist at a very specific type of capability or a very specific domain.” Acquiring that and plugging it into the ecosystem resonates with their ability to understand what quality, skill, and specialization look like.

One of the things that I noticed when I interviewed 4 of the co-founders was that the company itself, especially on Twitter, is known for hype, being cool, the vibes, and all that. But they take this really seriously, as they should. There is a huge responsibility to, 1, be helping serve the military and armed services, but also to be preventing lives from being lost, to make sure that any conflicts are actually very precise and very efficient.

What were some of the misconceptions that you started to see the more deeply you went into the research? I would say you might be one of the most well-researched investors of theirs at this point.

Well, I would say that—and it's a really good question, because I think there are 3 buckets, if I had to take everything out of my head that I've learned about Anduril over the last couple of years and put them into 3 buckets conceptually.

The history is a core piece, and we've talked a ton about that. The products, the process, and product philosophy are a big piece, and we've talked a ton about that. The piece that we haven't talked about as much, and that I think is one of the most important elements of understanding the company, is the philosophical and almost spiritual element of trying to understand why.

When I talk about the mission-driven engineers, what is the mission? Because if the mission is to build cool stuff and blow stuff up, that's not a mission, right? That's a neat trick. The mission revolves around conflict deterrence, and Trey talks a lot about just war theory, but there is this concept that the only way to effectively prevent conflict is to have the force capable of engaging in that conflict to begin with.

If you wait and say, “We're pacifist. We're not going to do anything. We're not going to build anything. We're not going to build any capability,” and then suddenly the enemy shows up on your door, you're done. You don't have the choice, the freedom, or the flexibility to be able to respond to that conflict on the day of. You have to be credibly building that deterrent effect very early on, and the reason that you do that is not because you love war. It's not because you love conflict, and it's cool, and you grew up playing Call of Duty or whatever.

It is acknowledging the fact that innocent lives are lost when people who can do something don't do something, right? The idea of pacifism is a privilege of the protected. There has to be a group of people who are willing to protect those pacifists, or else they no longer become pacifists; they become victims.

The most critical aspect of conflict deterrence is that the only way to prevent this—I think it's called the Thucydides Trap, named after a very old historian—is this idea that any time a rising power threatens an established power, it ends in large-scale, great-power conflict. When you look throughout history, there are 16 instances where that happened. Of those 16, 12 resulted in mass-scale conflict, and we don't want that to happen.

Anduril's entire ethos is this idea of presenting that capability not only to the US—to be able to say, “We are capable of deterring conflict. It will be expensive and painful for you if you engage in conflict”—but also to our allies. We want to make them what Palmer calls “prickly porcupines,” where we want Taiwan to be prickly enough that China doesn't think it's worth it to step on them.

A lot of people who get angry about this call people like the team at Anduril warmongers and think that they want this. One misunderstanding about the military-industrial complex, Palmer will say, comes from that famous speech: the idea that every dollar we spend on munitions is a dollar we're not spending on education, preventing poverty, and doing all these things.

Palmer doesn't say, as a person who is making money building those weapons, “Well, but I really want to make more money, so let's make more munitions.” He's saying, “I completely agree, and so let's stop spending tens or hundreds of billions of dollars on weapons. Let's improve the efficiency curve of everything that we're doing, actually achieve the means that we need, and actually achieve conflict deterrence in a profitable and efficient way so that we can reallocate a significant amount of our resources to things that make life worth protecting.”

Every time you listen to an interview from any of the founders, there is that fundamental sort of hopefulness: they want to prevent those conflicts so that we can enjoy a life worth living.

10. The Defense Investment Shift

VCX by Fundrise, the public ticker for private tech, allowing investors of all sizes to invest in venture capital. View the portfolio at getvcx.com. That's getvcx.com. Some of you may not have heard this yet, but our sponsor Public just launched something called Generated Assets, and it brings AI into investing in a way I've honestly never seen before. Here's how it works. You type in an idea like AI-powered supply chain companies with positive free cash flow or defense tech companies growing revenue over 25% year over year. Public's AI then dispatches a swarm of agents that scan every single US stock, evaluates them, and instantly builds a custom index around your thesis. What really stands out is how clearly it explains why each stock is included. And before you invest, you can even back test your idea against the S&P 500, so you're making decisions with real context, not just guessing. And beyond generated assets, Public lets you invest in stocks, bonds, options, crypto all in one place. They'll even give you an uncapped 1% match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com/sorcery, paid for by Public Investing. Full disclosures in the description. Enterprise AI runs on Merge, the AI infraplatform for integrations, agent tooling, and model orchestration, so your teams ship product, not plumbing. Mistral, Dropbox, and Drata already trust Merge in production. Start building at merge.dev. Founders ship faster on Deel. Set up payroll for any country in minutes, hire anyone anywhere, and get visas handled fast so you stay focused on scaling. Deel takes care of onboarding, HR, IT, ER, benefits, and compliance so your team can grow without borders. It's why more than 37,000 fast-growing companies trust Deel to move fast. Visit deel.com/sorcery. That's D-E-E-L.com/S-O-U-R-C-E-R-Y. A track we haven't gone down yet is the funding history behind Anduril, and there's a lot of deep lore there. They're rumored and reported to have doubled their valuation from the last round. It was around $30 billion, and it's rumored that it's going to be around $60 billion now. Who knows when that announcement is coming out, and if there is one?

Could you bring us through the funding history of Anduril? When did you—because I know your firm, Contrary, got in at a very interesting time. Could you walk us through that and why it wasn't obvious? It might look entirely hype-y and very obvious now, but it wasn't that obvious in hindsight.

I can speak at a very high level about what's publicly known and some of the details, but I think what's underappreciated is the story and evolution of the narrative. When Anduril was first getting started, even though Palmer had been a successful founder and sold Oculus for $2.5 billion, it was a very unpopular decision to work in defense in 2017.

There are stories about Palmer having a lot of engineers he'd worked with at Meta who said, “Hey, I think you were dealt unjustly, and whatever you build next, I'd love to follow you.” Then he says, “Well, I'm building in defense,” and everybody ghosts him, right? Nobody wants to interact with that.

I can't remember if it was the Series A or the Series B, but Eli Gill led an early round. I think there's a story that he got maybe not death threats, but very angry messages from people he considered friends saying, “How could you? This is immoral, what you're doing.”

There were also things like Palmer going to give a talk at a university campus and literally having to get security because people were threatening physical violence against him because they wanted to promote pacifism, right? There was all of this deeply anti-militarist thought early on.

A lot of the early investors, if I remember correctly, were primarily Founders Fund and General Catalyst. They were the main early investors in Anduril. I think Andreessen comes along a little bit later. We first invested, I believe, in Series C, and then we've invested at least a little bit in every round since.

Anduril as a company—the vibe shifted hard in 2022, around the time of the Russian invasion of Ukraine. But what's interesting is there was this sort of one-two-punch dip where the narrative was starting to shift, where people were acknowledging, “Man, this is a large-scale conflict. This is significant. This is not something we see very often, this type of conflict.”

People were starting to wake up to the idea of, “Well, wait, we don't want these types of conflicts to be happening.” At the same time, in 2022, the broader market is collapsing, right? You sort of have this post-ZERP hangover: stocks are plummeting, everybody's struggling, and everybody feels a lot of pain.

Anduril even had a round around 2022 where, despite this sort of Ukrainian conflict going on, it was difficult for them to raise during that period. You had some investors who had pulled out after promising to invest and things like that. So there was this sort of dip before a real takeoff.

Then Palmer calls the rest of 2022 his “I told you so” tour, where he literally gets to go around and say, “Listen, we have been saying this from the beginning.” What's crazy is, again, going back to my most surprising experience studying the history, it is true across the board, even until very recently.

Even back at the 2014 Russian invasion of Crimea, by no means was everybody saying, “Oh, yeah, no, we're good with no more conflict.” There was actively still conflict.

People chose not to see it because, in their hubris, they thought it was just a matter of time until countries like Russia or even China would eventually acknowledge, “Man, America is so great. This market, free-market economy and whatever—they love us. Don’t they want to be like us? Look at our culture and Hollywood and everything is great.” That hubris was very pervasive for a really long time, and the vibe shift didn’t happen for a long time. In 2017, you had Putin saying, “Whoever controls AI will control the world.” And it’s the same idea: in 2017, Anduril gets started, and they’re saying, “Listen, from day one—”

The acronym is AI.

That’s right. Anduril Industries. From day one, they acknowledged that autonomy and AI have to be front and center in any viable force solution. That was not the popular conclusion. Post-2022, that has largely become the conclusion: acknowledging that conflict is increasing, tensions in Taiwan are increasing, Iran is going to continue to be a struggle, things in Venezuela, whatever.

There’s going to be global conflict. That’s going to happen. I think the next wave now that we have to deal with is what Palmer says: “Listen, this is a point in time where people are acknowledging the value and importance of what we’re doing. There will be ebbs and flows in that shift,” right?

And there’s a lot of politicization of specific conflicts that makes people not in favor of particular positions or allies that the US may have or whatever. Increasingly, we’re going to have to deal with that, where there are going to be people who say, “We shouldn’t be in the business of conflict, of deterrence, or whatever. We should just let it go,” because they don’t like the politics of the underlying conflicts.

But the problem with that is that you don’t have the privilege to just say that, because anybody who can do something and doesn’t do something is ultimately responsible for the outcome if we don’t do something that we were capable of doing. And if we rest on our laurels and say, “We’re just going to stay inefficient,” that is the same as doing nothing, and we will be responsible for larger-scale conflicts that will not only hurt human rights, that will not only hurt individual lives, but also economically.

You look at a conflict over Taiwan; I think people say it could be a $10 trillion hit to the global economy—bigger than the financial crisis, COVID, anything, I think, in some cases combined. There are so many reasons why we need to have effective deterrence that it’s not a vibe that we can continue to ignore.

A really good point of that is the conflicts for investors. For some people, they might not know, but investors have LPs. Sometimes those LPs have certain provisions, clauses within their LPAs where you can’t invest into vice categories or war and that kind of thing. So back in the day, there were a handful of funds that could not invest into these companies because of those clauses. I think the vibes have shifted, for sure.

But not for everybody. Before Contrary, I was a partner at Index Ventures. And Index, as far as I know, still to this day has a clause where they can’t invest in anything that is a weapon or could be used as a weapon. I remember they had investments in a drone company where they could have contracted a provision when they funded it, saying, “You will never allow this to be armed with weapons and stuff.”

Wow.

So it’s not only that we’re going to avoid weapons, but we are going to exert influence over the companies that we invest in to make sure they also don’t engage in weapons and defense use cases. In this case, it was a policing use case.

So there are still definitely firms, even major firms, that are not comfortable funding weapons. And I think, candidly, that is good. I think there should be people who remove themselves from this environment, because I think either you are willing to do the work to weigh and measure the philosophical implications of what you’re doing and acknowledge the weight of the burden that it is to engage in this type of technology, or you’re not. I think more people would be better suited if they engaged with that philosophical debate. But candidly, I think that’s true of most technology, right?

Yeah.

I think about that as well with addictive social media or even AI girlfriends and stuff—having things that can be destructive to human psychology and stuff. Everybody should be having more philosophical debates about the things that they are investing in. We don’t have nearly enough philosophical debates about the implications.

We just say, “Hey, this is a hot technology, and it looks like it’s ripping, and people really like it,” or, “It has a big market for this,” or whatever, and it’s just blinders on: big market, let’s do it. I think more people should have to embrace the philosophical stick that they’re picking up, because there are consequences at the other end of that stick.

Philosophy is back. Do you know one of the largest and hottest philosophers right now?

I don’t.

It’s Alex Karp.

Attaboy.

So there’s a really good quote from Chris Pike I think at Pace Capital, where he says, “People want to not just be right; they also want to be clever.” And so you look at these things that feel obvious, and you say, “I don’t want to do the same thing,” right?

Whereas, a couple years ago, buying Google would’ve been a pretty good decision to make, right? People were thinking that it was just slow and steady wins the race—a big hyperscaler or whatever. It’s like there were actually a lot of these underappreciated tailwinds that would advantage the company in the long run.

I think it is the same, but some of those are forced through negative narrative. So again, in 2022, there’s a great story: I think it was Sequoia and Brookfield that started a secondaries fund in the heat of 2022’s collapse. They went around just buying, saying, “Hey, there’s a bunch of these companies that we think are generational companies, but they’re going through a hard time. People want to sell. They’re freaking out about the market. The public markets are getting demolished and whatever.”

And so they went around and bought a bunch of positions in these companies, and it turns out these momentary blips throughout a company’s life are often not indicative of the long-term success of a good business. And so I think that it’s also one of the advantages of paying very close attention to a specific moment in time.

Our good friend Leopold.

Yeah. Making a ton of money off of just paying attention to the underlying thing, and you have memory chips, right? That is a logical second- or third-order effect of the different things that we’re focused on.

But because you’re paying attention to the loudest and proudest, spiciest name, current thing, you get distracted from the broader ecosystem of stuff that exists. And I think that’s the same thing: people are underappreciating a lot of the second- and third-order effects from a supply-chain perspective around defense because they’re so focused on, “Well, who’s going to be the next Anduril?” You can make a lot of money investing in things...

Or even, I saw a story about—I don’t remember what the specific stock was—but somebody had paid really close attention to the vendor of the ballistic shell for Anduril’s EagleEye. It was a small, publicly traded manufacturing company, and they were like, “Hey, they’ve got this contract with Anduril. This is likely going to rip as this contract...” I mean, it’s a $22 billion contract. As they start to go into production, they’re going to start ripping. They bought a ton of stock in that random little manufacturer. I think it was literally maybe a $100 million or $200 million market cap stock.

Mm.

Like, a very small company. And that stock has ripped because it’s paying attention to what those second- and third-order effects are rather than getting caught up in what the current thing is. That’s what bugs me about so much VC fodder: the idea that the internet is real life and all that matters is the current thing. You pay attention to this thing, and that’s all. I think you miss a lot because you’re missing out on the compounding nature of context—gathering lots of context over time and being able to make decisions based on the ecosystem instead of what’s the hottest new AI lab for X or whatever. It leads a lot of people astray.

Alpha. You heard it here first. So, after reading this book and writing this book, what are the key takeaways that you learned from Anduril’s leadership and their ability to perform over and over again?

When I reflect back on the last couple years, I was just spending a lot of time with the Anduril founders indirectly, hearing what they have to say, how they think about things, and how they articulate things in different contexts. Throughout the book, we had, I think, over 500 individual sources that we’ve referenced 1,000-plus times. Specifically, if you look at the dozen or so interviews with the founders, I went back and looked at transcripts of those things. There are... The word hope appears 44 different times, and there is this fundamental hopefulness in everything that Anduril—the company—builds and in the mission behind the company.

It’s this idea that if we thought all hope was lost, we wouldn’t be doing this. We’re doing this because we continue to have hope, and I think that speaks fundamentally to what a real mission is. Trey Stevens talks in an incredible essay about choosing good quests; he talks about it all the time. I think this speaks to what a good quest is, and it is not a specific product. It is not a specific solution.

Even for Anduril, it is not defending a specific nation, right? It is not just making the US capable of protecting itself. It is protecting a specific subset of ideals and values that represent this idea of Western democracy: human rights, checks and balances, strong rule of law, those fundamental elements. You believe in those things, and so you want to defend those things. You anchor your mission to that specific worldview and belief.

The founders of Anduril believe that those ideals and values are worth defending, even if Anduril didn’t exist. Even if they weren’t making a dollar from it, they would continue to believe fundamentally in the hopefulness of protecting those values. What is awesome is that speaks directly to the name of the company, Anduril, the Flame of the West from The Lord of the Rings.

There is a quote from The Two Towers that’s said by, I think, Faramir, whom Tolkien even said he most resonates with—the character with whom he most feels a kinship. I’m paraphrasing, but what Faramir says is, “I don’t love the blade for its brightness or sharpness. I don’t love the arrow for its swiftness or the warrior for his ferocity. I love only that which they defend.”

That is fundamentally true of Anduril as a company, of Anduril’s founders, and of the mission that they engage in. Every product decision, every hiring decision, every funding decision, every marketing decision, every aspect of the company comes back to this idea: What do we think will enable the protection of democracy most capably? That’s what we want to do. That informs the product decision.

People who go and try to chase a mission and say, “My quest is good because wouldn’t product ABC be awesome? Wouldn’t it be sick if this product existed?”—that misses the fundamental worldview. What you want is this underlying worldview that informs everything you do, and Anduril has that.

Wow. That’s a really good answer. One of my favorite quotes of all time is, “Hope is a good thing, maybe the best of things,” from The Shawshank Redemption. It’s great. Okay, so as we close out, this is just a simple question. It’s not too difficult. What is your hottest take right now?

My hottest take is that fewer people should be raising venture capital.

Wow. Why?

Which is deeply unself-serving of me as a venture capitalist.

Are you seeing too many bad pitches? Is that why?

I think I’m seeing too many companies and too many people who want to build things. Going back to the mission, venture capital is a very specific wedge. It’s a very specific funding mechanism. And because it’s so legible, as our friend Packy McCormick would say, it’s so sexy and appealing, people overstretch to try and fit their passion into the square hole of venture capital, and it’s problematic.

The problem is that there is a large swath of companies that could be very good companies that are very bad venture-backed businesses. They would be very good businesses, full stop, but they are very bad venture-backed businesses. More people should strive to deeply understand the mechanism of venture capital and what that requires in terms of the size and scale of outcome, scale of growth, requirements of burn and velocity and ambition, and be willing to say, “I have to want to build a $100 billion company, because if I don’t, it doesn’t justify the VC appetite that I’m trying to stir up.”

There are a lot of better ways to build businesses. That’s fundamentally one of the reasons why I write my blog, Investing 101: to articulate both the virtues and the vices of venture capital, why there are goods and bads, and to help people be more informed.

When I look at this long-tail ecosystem of reindustrialization more broadly, there are a lot of great companies that can be really good venture-backed businesses. Even if they’re not going to be neo-primes, they can still be really great venture-backed businesses. But there is a huge swath of businesses.

I have a friend who builds industrial-scale metal shelves. That’s their business: they build the shelves you see in Costco. They build some of those, but they also build things like racks for Tesla to hold different components of the cars and stuff like that. That guy is doing $17 million of revenue and $3 million of profit, and he owns the whole business.

He lives a great life, and we need more people like him to be able to do that. That would be a terrible venture-backed business because he’s grown 3% a year, give or take, for the last 5 or 6 years. It’s not growing super fast, but he has a deep client base that he does very well with, and there are a lot of really great opportunities. More people should seek alternative ways of building great businesses, because venture capital does not have a monopoly on great businesses.

Interesting. Sounds like he might have a good supply chain.

He indeed does.

There could be some growth opportunities there. Well, Kyle, thank you so much for coming here on short notice, giving me an advance copy of the book, and letting me read it in due time. I really appreciate all the time. Seriously, so much work has gone into this book and the research, and I’m definitely going to look up some of the sources, because the books that you mentioned throughout the conversation were really interesting. And I hope that writing this got you extra allocation. Thank you so much.

Yeah. We just do it for the love of the game.

Amazing. Well, thank you so much.

Thank you.

Hey, it's Molly. If you enjoy our interviews, check out our newsletter Sorcery.vc, where we deliver a once a week top deals and tech headlines email, and also go deeper on our podcast interviews. Subscribe to Sorcery today. And don't forget to subscribe to the podcast on YouTube, Spotify, Apple, or wherever you listen. Link in description to sign up.