E40: Venture, AI, and the Defense Industry with Joe Lonsdale of 8VC, Josh Wolfe of Lux Capital, and Matt Steckman of Anduril

0:03Today on Turpentine VC, we're airing a live panel discussion from the Hill and Valley Forum, an event that brought together technologists and lawmakers in Washington, D.C. the first week of May. This discussion centers on the future of military technology and AI with speakers Joe Lonsdale, managing partner at 8BC, Josh Wolfe, co-founder of Lux Capital, Matt Steckman, the CRO of Andrel, and Senator Mike Rounds. They discuss the cultural shift occurring in the DoD, where startups have a chance to beat defense primes. The first voice you'll hear is Senator Rounds, followed by Joe, then Josh and Matt afterwards.

0:35For more from the Hill and Valley, Follow the link in the description.

0:47Good afternoon. Everybody up? You got your caffeine? You're ready to rock and roll yet? Still doing it? We have less than 30 minutes, but what we wanted to do was to take some time and talk about our national defense, about AI, and about the connection with new technologies and what's happening. God, my background on this is very simple. Coming out of the Armed Services Committee and the Intel Committee, the one thing that we find out time and again is that these new technologies are making a difference in the way that we fight wars. And it's not just us, but it's small countries as well that can afford to buy the technologies.

1:24So today we're sitting here and we just decided that we're just going to kind of have a little bit of fun. But I'm going to offer some opening questions to the crew here. And what we'd like to be able to do with this, if I could, But I want to just, as I get my stuff ready to go, if each of the guys with me here today would introduce themselves and give just a brief thing about if they had one thing they wanted to get across to the team that's out here right now, what would it be? Let's take 30 seconds and just run across and do this. Fire away. All right. I'm Joe Lonsdale. I'm proud of having recruited Alex Karp and Sean here.

1:57So you got a good team I'm learning from there. And I've built a bunch of other companies, but a few more into defense recently. I think it's really key right now to let the new things in with having primes like Palantir, like Yandrel, and others. We're building things with them now, and there's a lot of ways we're applying AI, applying direct-to-energy to completely change the battlefield, and maybe spending more of our money on the new stuff. Josh Wolfe, co-founder of Lux Capital, we're a$5.5 billion fund. About a third of investments that we're making now are aerospace and defense, air, land, sea, space, subsea, cyber, and beyond.

2:31It's actually an amazing thing in venture because we're seeing an enormous amount of collaboration between venture investors, people that normally are competing, are forming syndicates, starting companies de novo, taking entrepreneurs that are vested out of existing defense companies and starting new companies with them. So it's actually a rebuilding of almost the capital formation of the industrial base, which is a beautiful thing. Number one message I would say is let the competition that is happening now, the capital formation that's happening now, the flow of talent that's happening now, just absolutely flourish.

2:56This is what made defense industry in America great. It is the very tenets of free markets and capitalism, the flow of talent. We have foreign adversaries. They're both in the information space and the physical mass and kinetic space. And let us do our thing to back the brilliant people that are starting companies and let the capital follow. And the more that we can break down the bureaucracy and return to the roots of what Silicon Valley was, which was really in electronic warfare and defense and not just the orchards of people working in garages. And to re-remember that history, I think is critical.

3:26Matthew Stackman, Chief Revenue Officer, Annual Industries. we're a defense product company. You know, on a day-to-day basis, I negotiate hundreds of contracts on a year-to-year basis with the government. Things are changing. It is very uneven. It is very spiky. There are pockets of excellence. And when it happens, it happens big and it happens fast. In the last two weeks, we announced our successful down select for the CCA Low Wingman Program for the Air Force and our successful delivery on the Australian GoShark Program. ahead of schedule and under budget. It was kind of the headline, not like - It's par for the course in the defense.

4:04Yeah. And that's exactly where I wanted to start because on the CCA and where you're going, what do you think about it? Does it suggest a shift in the way that we're going to be doing business at the DOD? There's a lot of thoughts going through my head on CCA, both personal and professional. I'll bet they're pretty good thoughts. Yeah. Well, on the personal level, not a lot of folks at Anderle have ever been the popular person at probably growing up. And now we're like the popular kid on the Hill and in DC, and it's very uncomfortable for most of us, but obviously quite exciting. Professionally, yeah, like the answer is yes, it's a watershed moment.

4:44I think that the thing that is being maybe lost in all the headlines is like, yes, it's a watershed moment for non-traditional defense. Yes, it's a watershed moment for the defense industrial base. Yes, it's exciting for how the Air Force and Secretary Kendall pursued this in a new way. But what should not be lost is we won the competition because we designed the best plane. Okay. Now let's go back a little bit. I know we've got a professional group on here and you probably all know exactly what we're talking about, but let's give some real basics on this. Just exactly what you did design when we talk about CCA.

5:17Sure. Yeah. So CCA is the Loyal Wingman concept. So think about a manned fighter aircraft F-35 with a bunch of friends around it that are helping it see and sense an effect at greater distances, at higher velocity, and greater lethality. It requires a good sense of how autonomy and AI can drive those aircraft because, of course, a quarterback in an F-35 can't be literally flying multiple aircraft in the same airspace. You have to rely on a high amount of software and automation and autonomy and ai which i know these guys will want to talk about um and the program started years and years ago actually in formation um uh through sort of the first five including some of the household defense primes uh involved uh now down to the final two of which of which were one of them um so we think we designed the best plane we're very excited about it we beat at this point boeing northrop and um and lockheed martin in the competition, companies that have been around for, I mean, some over a hundred years, if you really trace back their lineage.

6:24And we're ready to win the final select and we're ready to go fly some really, really cool planes. And here's the bottom line on this. Now, when we're talking about the top end of technology with stealth and communications and the ability to coordinate, but you're also talking the integration of AI into this process. I want to go back a little bit and get everybody else involved in this a little bit. One of the areas where, as a member of the U.S. Senate that I started learning about some things was when we started talking about the Nagorno-Karabakh War that occurred between Azerbaijan and Armenia.

6:55Here's two countries that you would never think being highly technical in nature, and yet they used AI-enabled devices that they picked up and purchased from other countries. In August of 2023, Deputy Secretary of Defense Kathleen Hicks unveiled the replicator initiative to counteract what China could do by fielding thousands of attributable autonomous systems in the next two years. This, in real sense, a lot of that came from what we learned, not just in Ukraine, but also in the Nagorno-Karabakh war between Azerbaijan and Armenia. Your thoughts about the future of warfare and what you're seeing in terms of the ideas that are out there as war changes its face?

7:39Well, I mean, attributable systems are obviously the future. AI is obviously the future. The problem right now is that we're oftentimes spending a million or$3 million on something to stop something that's very cheap. And you can't do that as skill. Out of prolonged warfare, you run out of money. A lot of what war is, it's an engineering problem at the end of the day. And you need to be able to produce more on your enemy and get them. So we just sold a bunch of radiation-sensitive missiles to anti-radiation missiles to Poland. And that's great. But these things are$3 or$4 million each to take out the Russian jammers that you can produce for tens of thousands of dollars.

8:08You can't keep doing this skill. It doesn't make any sense. And so it turns out for everything we're doing, if it turns out you shoot down swarms of drones with really, really high-powered microwave, that costs dollars per shot. That's better than using millions of dollars per shot because you're just not going to be able to do it at scale. So there's things that actually work now that we need to be investing in and that we know is the right answer. So it's replicators, the right high-level thing, but are we going to get the money to the things that actually are efficient at scale? And I think it's going in the right direction, but it's a little bit slower.

8:32But you're not just starting out on this. This has been something that the companies have been looking at. We've been working on this for a long time. I mean, eight years ago, we started EPRIS because we realized the drones and drone swarms and interoperable systems were going to be a big problem. We need to be able to turn them off. And, you know, within the U.S. Navy right now, it knows it's going to be able to fortunately triple its effectiveness with swarms of atomist vessels with AI partnering with weapons systems from Riannville and others. And there's things like that you could do with these AI things at scale that we need to do, frankly, because we're not going to be able to build enough ships otherwise to stay ahead of China.

9:00So it's not a it's not a case of AI being the weapon. It is a case of AI being incorporated into a weapon system that makes it modern day. You're going to have hundreds or thousands or tens of thousands of cheap things. The only way they can intelligently coordinate and work at scale in any scenario, it's going to be various types of AI to coordinate them. Yes. Josh, long running topic of conversation at events like these is how the U.S. can translate the dynamism of our robust tech sector and venture capital markets into a battlefield advantage. This is one of the real advantages that we do have over the PRC.

9:38In your view, how is the department doing in terms of leveraging our unique VC markets and turning that advantage into concrete battlefield capabilities? On the one hand, better than ever. And on the other hand, not good enough, but improving. If you think about battle space, as you mentioned in PRC, you think about Taiwan. As much as we are future forward and all of us are funding cutting edge systems, small and large scale, and increasingly you're seeing that it isn't just small scale drones, but whether it's undersea like Ghost or Fury, they're very large systems. It still is about mass. It's about mass on target.

10:17It's still it's about fuel. You've got diesel that's got 37 megajoules per kilogram. You can't get past that with software to figure out how far you're going to project a ship and then the rocket fuel that's going to project a missile and the PBX inside of that that's going to actually explode on a target. You look at the threat in Taiwan and you've got something on the order of 40 amphibious vessels. You've got 50 subs. You've got 200 surface combatants. You've got 2 ,000 fighters. You've got 3 ,000 UAS. You've got 750 satellites. These are all a combination of just off the top of your head. Yeah, I know.

10:53But this is the battlefield as we know it that's not hidden. And then you've got 200 ,000 mechanized infantry. So each one of those things has to be dealt with in a coordination that's almost like jujitsu. And the biggest source for the autonomous systems and the drones is not just the attributable system, but preventing the enemy from having that battle space and having lots of these things that can just, you know, pile up almost like Sierra Madre style throughout these things. So the Defense Department overall is doing well in improving our ability to develop new technologies to face these kinds of battle spaces.

11:24And you've got highly motivated people like us that are willing to put capital at risk. We're willing to wait the long periods of time. It was totally against the venture creed that we were going to do dual use things and you're just going to take too long with the bureaucracy of the government. That's changed. We're willing to fund companies like Anderil that are solely focused on U.S. and allies and solely focused on defense and the main departments of the government that are going to be the customer. And so that is a step change. And I think you keep feeding that. Let people win these programs.

11:53Capital will continue to flow, like I said earlier. The town will continue to flow, and you will have a beautiful flywheel that we look back in a generation and say, my God, we did it. We've talked about the fact that we have this advantage over the PRC, the availability of private sector with capital moving in. And we've got these great ideas that are out there. So let's come back in again. How is the department doing on integrating AI into these capabilities and accepting it in and trying to promote it in their systems? I think it's mixed. I think there's a general acceptance that this is inevitable, which is good.

12:26And I think you have pockets of the government that are leading and some that are following. You know, I think we generally get caught up maybe in the wrong parts of the conversation where the technology is so new, we sort of take it to the limit and we say, oh, there's going to be, you know, a Terminator killing machine being built. Like we're not even close to that. Like most CB can't even identify a cat at 100 meters, right? Like let's just kind of take a pause here. We should be figuring out the integration smartly and pragmatically of mature versions of artificial intelligence into our weapon system and not be taken aback or slowed down by the 50-year future, but actually focus on now and what's happening and countering our adversaries.

13:08One of the things that you watch what Anderil is doing, to me, there's almost an analogy of like in the finance markets before Bloomberg, it was like our chaos, the amount of information flow that you had. You think about the modern battle space and how much speed of information and chaos that there is and that we can anticipate there will be. being able to synthesize that so that a single operator can control multiple vessels, process that information, prioritize what the targets that they're going to focus on. It's akin to having like a Bloomberg system for somebody that's a trader, maybe with a shorter term frequency to be able to figure out what are the priority things and how do I integrate this and make high priority decisions.

13:41And to sort of bring it back to sort of like the principles of how government should go about this, like we didn't win the CCA downslide because of AI or any sort of novel application technology. We built it just because we're very good at systems design, right? Just hardcore, old school systems engineering that is gross and hard and you're spending late nights and you're chugging Red Bull and you're trying to figure out what the best possible answer is. None of that has to do with AI, right? And the other big companies that lost, none of that had to do with AI, right? It just had to do with letting new ideas, new people into the environment, breaking down all of the barriers to entry that exist for a lot of the folks actually in the room and giving us a chance and a shot on a fair playing field.

14:28And once that happened, we were able to excel. That was the big news. It wasn't that any of us were surprised that Andrel had built the best technology we knew they had. The question was, was the DoD ready to allow them to play and for the competition to win? And many times we'll build something that's clearly the best, but the primes still get the giant contracts. And they say, oh, you just wait a few more years, you'll eventually get there even after you crush them in a contest. And that's what we need to fix. Hey, we'll continue our interview in a moment after a word from our sponsors. There's another interesting thing, which maybe if I can ask him a question, because it's still fascinating to me, because of OSUK, they ended up going not to the US Navy, but they went to Australia and they did something in record time, under record budget.

15:07People applauded before on this, but it is really significant. Talk about that. Just why that is so significant. I mean, it caught your attention. Yeah. Yeah. There's always an alignment of a lot of stars. when these things were. I think the Australian government was ready from a technology perspective. There was a criticality to it based on their, literally just where they actually are on the map with perspective China. Like they're right there. We're not. They knew they couldn't rely on the rest of the alliance to field the capability. The US has had a program trying to do this for over a decade now.

15:47Over budget. Hasn't really delivered on the promise. So the Australians were basically got to a point where they're like, we got to do this by ourselves. This is a strategic capability. We can't not have this. We had the tech, we had the vision. I think Josh mentioned private capital. We were also able to take a lot of the risk on our own book. We took half the risk of the program. The Australian government dollar for dollar met us. So they were sort of pot committed in a way, but the risk was split evenly. That is very, very unusual in U.S. defense contracting. Normally, the U.S. government is bearing all of that risk, all of those dollars.

16:24And the primes can make mistakes and they can go slow and they can still turn a profit. In this case, if we made a mistake, we would have lost a lot of money and our investors probably would have made some pretty nasty calls to us, right, over the last couple of years. But it worked. and it drove us to move fast because it was our capital. And I think the program is going to succeed and I think it's going to be repatriated back to the US. This is a big advantage, I think, that the new companies you're seeing have an innovation against the big legacy companies is the old legacy companies, you don't do something unless you get them to pay you for it.

16:58And that makes innovation really hard. And so it's a lot riskier, but what Josh and I are funding and what Andrew is doing and our other companies are doing is we'll spend$50 million,$100 million on something we really believe is the best and you don't even pay us for it yet. Just like let us get in, let us compete, let us win. We're willing to take that kind of risk. That's one of the reasons, one of the many reasons why you're seeing all the new innovation companies are new companies because the primes just very rarely do that. Interesting that what you're also seeing now is a coordination. You're seeing the technical aspects of what each of them has to offer and then you're finding a venture capitalist or at least capital coming into this process and actually believing that they can win a contract with the Department of Defense.

17:39And what just happened may very well be one of the biggest eye-openers that we've had in quite some time. But there's also something else going on here. We have to move more quickly in terms of adopting the most capable and newest technologies that are available to us today. And the reason is because other countries can take what we could buy and they can buy it as well. And the shortcut in many cases is with a new technology being able to take over or to become better than an existing technology that's been incorporated into defense systems. I'm going to come back to the Nagorno-Karabakh war between Azerbaijan and Armenia.

18:24This is a case of where Azerbaijan, in a very short period of time, about 40 days, actually went back in using, in some cases, some, I'm going to call it for, please, let me say, primitive AI capabilities at that time. and they were able to identify kill boxes and literally go into an area and take out significant amounts of Armenian heavy equipment, whether we're talking anti-aircraft equipment, we're talking armored personnel, tanks, artillery pieces, and so forth. And they did it using drones that they could buy on the market. Now, if the United States doesn't recognize that we have to be better than any of our adversaries at adopting and then implementing that, then we're going to find ourselves at a significant disadvantage more quickly than what we otherwise might over a decade ago.

19:23What you bring to the table when we talk about the investors and the entrepreneurs who are developing these systems is something that we can do better than what we believe the PRC can do. But talk to me about the advantages that we have under our system if we take advantage of the private sector's capability to work with the DOD. Well, this is absolutely critical. Palantir is not just a top AI company in the DOD. It has billions of dollars of revenue from the private sector. It's competing and winning in the private sector and growing very quickly there. And that's something you don't really see going on in China for multiple reasons.

20:06We talk about the advantages versus China. Xi Jinping basically just assassinated a big part of this tech sector. People are terrified to build a second company. The majority of our top new companies right now are also second time building companies right now and and you have a lot of those people disappearing running away not not there and then and then we have still the very very best talent there in the private sector and ai so being able to take it being able to have things that are winning in both of those areas and letting them work between their palantir obviously is unique and being able to work in the dod we did a lot of we made a lot of mistakes over a very long period of time we had some very difficult things to deal with and kind of kind of broke through and eventually made it work making it you know i think the u.s has a lot at the edge of these companies.

20:42It does have to make it easier for other ones to be able to take things that are winning. I'll give you one example. All sorts of AI and logistics completely changed the commercial world, how logistics and supply chain works. None of that technology right now is the DOD. The DOD is very protectionist of a logistics area, does not want outsiders coming into that and working with it. It's very difficult. So there's things like that we could be taking advantage of if we were to change it up and open it up to outsiders. I'll just do two things, one on competition, one on culture. On competition, you go back to the Cold War, there was not one consumer product that Russia had that was really competitive with anything.

21:13They had MiG fighter jets and they had AK-47s because those had to fight on a global stage. Everything else, you wouldn't buy a toaster, you wouldn't buy a nuclear reactor. I mean, just everything was just inferior to a system that we had where people were able to compete and fail and the market could choose. So that system helps to select technology and talent, and we should continue to feed that. On the culture piece, for a generation and a half, we were warned about the military-industrial complex. And you had a chip on people's, you know, the gorilla on people's back saying, you know, we're not going to go into that or that's a bad thing and there's too much bureaucracy.

21:42You saw primes go from 70 down to five. China has military civil fusion. They're overt about it. It's an admonishment. If you're working on technology, it's part of the CCP. And so we don't need that, but we need people that are really motivated. And I think what Jacob and Christian and Delian and everybody that's gathered here is doing is recreating a culture that says, it is great to be an American patriot. It is great to fund technology that's specifically focused on American defense, that we do have adversaries that have malicious aims, and that we should fund these technologies and recruit the talent.

22:13And you're seeing that at Andruld, watching people line up that, I remember the first time, you know, when they were at the Reagan National Defense Forum. And at first, people were sort of laughing. It was like the classic Gandhi thing, you know, Palmer's in his Hawaiian shirt and his cargo pants. And then people are like, who the hell is this guy? And then everybody's lining up because they all want to be part of it now. And so, you know, first they laugh at you, and then they join you, and then you win kind of thing. And that's happening. And that's a cultural shift. And it's a beautiful thing to see.

22:38Yeah, I think we get caught in this like do loop talking about you reforming the system and policy changes and like all this stuff. And like, it's all important. We should do it. It's table stakes. There are actually, but I literally see you in the room. There are government champions in this room who actually have put the machine on their back and made it work and made it work quickly. CCA is an example of this. there's other examples that are represented in the room. So it's not a problem of, can the system do it? It's a problem of will and it's champion dependent right now. I think like if there's one message maybe I want to leave is the new entrants, we're ready to take on the big programs.

23:22We're done fighting for table scraps, right? Like let us to the table, give us a fair fight and we're going to execute and we're going to win. I think this will make the folks who we identify as being some of the largest contractors, it's going to make them more competitive. And I think there's very, very smart people there that are going to see what's going on. I also think that the Department of Defense, they recognize that in many cases, their process of acquiring and picking up information is broken. And the best example of that is, is we do rapid acquisitions, right? So if you're going to have something that's really important and you got to get it in, we don't use the regular acquisitions process.