0:00:06 Today on the A16Z podcast, we’re sharing Catherine Boyle’s recent interview on TBPN.
0:00:12 Catherine, a general partner at A16Z and a driving force behind the American Dynamism thesis,
0:00:15 joins the show to discuss the state of the movement.
0:00:20 From the Department of Defense’s latest reform efforts to the growing role of startups in national security and beyond,
0:00:25 Catherine breaks down how far American Dynamism has come and why the work is just getting started.
0:00:31 As a reminder, the content here is for informational purposes only.
0:00:34 Should not be taken as legal, business, tax, or investment advice,
0:00:36 or be used to evaluate any investment or security,
0:00:41 and is not directed at any investors or potential investors in any A16Z fund.
0:00:46 Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast.
0:00:53 For more details, including a link to our investments, please see a16z.com forward slash disclosures.
0:01:03 And next up, we have Catherine Boyle from Andreessen Horowitz, the pioneer of American Dynamism.
0:01:07 One of the top coinages of the last few years.
0:01:07 Yeah, for sure.
0:01:09 And more relevant than ever.
0:01:11 In some ways, mission accomplished.
0:01:16 I mean, last week in D.C., it was American Dynamism on display, right?
0:01:17 That’s right.
0:01:17 Everyone is a believer.
0:01:19 And the question is, where do we go from here?
0:01:22 And that’s what I’m excited to dig in with her today.
0:01:23 Catherine, welcome to the stream.
0:01:24 How are you doing?
0:01:25 Thanks for having me.
0:01:26 It’s about time.
0:01:27 I’m so happy to be here.
0:01:28 Good to have you.
0:01:30 Long-time listener, first-time caller.
0:01:31 Great to have you.
0:01:33 With the flag in the background.
0:01:34 Yes, fantastic background.
0:01:36 Where should we start?
0:01:40 I’m curious about state of affairs with the American Dynamism movement.
0:01:45 The project feels like it’s maybe time to rest on our laurels.
0:01:46 What do you think?
0:01:48 We’ve achieved American Dynamism.
0:01:49 It feels like it.
0:01:50 Yeah.
0:01:51 No, I mean, seriously.
0:01:53 It feels like it’s broken through.
0:01:54 It’s mainstream.
0:01:57 Cultural victory may be coming before actual victory.
0:01:58 Sure, sure, sure.
0:02:03 And that American Dynamism is almost mainstream now, at least in the venture world.
0:02:03 Yes.
0:02:05 But job’s not finished.
0:02:05 Job’s not finished.
0:02:10 So what are the key asks in D.C. from Silicon Valley right now?
0:02:11 What are the top projects?
0:02:17 Where should tech be focused in terms of the American Dynamism project broadly right now?
0:02:17 Totally.
0:02:19 So it’s nowhere near finished.
0:02:23 I mean, this is like three or four years into a 30-year project, which is always good when
0:02:24 you have those sorts of aims.
0:02:25 And I’d say it’s even longer than that.
0:02:29 When you think of Defense 1.0, it started around 2015, 2016.
0:02:32 It sort of has become this very large movement.
0:02:37 But I’ll tell you, like, last week was a huge, huge week for American Dynamism inside of
0:02:38 the DOD.
0:02:41 And the news sort of got buried in tech land.
0:02:44 But it was just, it’s probably one of the biggest things to happen in the first 100 days
0:02:45 of the Trump administration.
0:02:49 The Army announced what they’re calling their Army Transformation Initiative.
0:02:52 It was with Secretary Driscoll, General George.
0:02:56 They actually went on Fox and Friends, which was like a huge deal that they actually went public
0:02:56 with it.
0:02:59 And they said, it’s been way too long.
0:03:01 We have so many platforms we want to modernize.
0:03:05 We want to divest from technologies that are no longer useful.
0:03:07 We want to modernize the force.
0:03:11 We want to make sure that we get rid of civilian jobs that are not important anymore.
0:03:14 We want to make sure we are not having wasteful spending.
0:03:18 It was sort of like, you know, what I had read in NBC after it came out, they said, like,
0:03:19 the Army is doging itself.
0:03:20 Yeah.
0:03:25 I think that the real story of what the Army is doing, and kudos to them because they truly
0:03:29 are the first mover, is there are people inside of the DOD who have been saying these
0:03:34 things for years, pounding their head against the wall, saying they want acquisition reform,
0:03:38 saying they want to work with startups, saying they have to have new platforms that come
0:03:40 in and actually support the needs of the warfighter.
0:03:43 And they’ve been pounding their head against the wall with little results.
0:03:48 And so when you have a DOGE effort going on in Washington and an administration that really
0:03:52 wants to see the waste disappear, it allows for those people who are really forward thinking,
0:03:57 like General George and like Secretary Driscoll, to come forward and say, hey, we’re going to do this
0:04:00 ourselves. We’re going to pick out the new technologies that we need. We’re going to get
0:04:04 rid of things like Humvees that we haven’t needed in 20 years. We’re going to figure out what is
0:04:08 actually useful for the Army. And we’re going to do it ourselves. So it was a huge week. I think it
0:04:13 was probably one of the most, it was reported, but it didn’t get sort of the praise from technology
0:04:18 that it should have. Like this is an extraordinary movement that I think has really been a long time
0:04:22 coming. And it’s something that a lot of the early American dynamism companies have been pushing
0:04:25 for for a long, long time. So congratulations to the Army.
0:04:26 Yeah. So can you give me a little bit of a…
0:04:29 Yeah. It’s better to DOGE yourself than get DOGED.
0:04:30 Always DOGE yourself.
0:04:30 DOGE yourself.
0:04:35 Yeah. Can you give me a little bit of a tour of the market map of the beneficiaries of this
0:04:40 transformation? Obviously, everyone knows the Palantirs and the Anderals, but I imagine that
0:04:46 there are tons of pockets of value and projects that need to be overhauled. Is it mostly drones,
0:04:52 weapons systems, vehicles first? Or are there other areas that companies that you talk to are focused
0:04:54 on in this transformation process?
0:04:59 In the early days, that’s what’s been called out. So it’s early UAVs that were developed 20 years ago.
0:05:04 They’re not relevant post-Ukraine war. It’s actually sort of, I don’t want to say comical because it’s
0:05:09 not funny. But when you think about the fact that the Humvee was developed in 1980, it went into
0:05:14 production in 1985, and that the Army said in 2004, this actually isn’t useful for us anymore because
0:05:18 there’s this new type of warfare called IEDs and we’re not going to use it. And these are still in
0:05:24 production in 2025. So that’s a perfect example. And I think they’re showing certain programs that
0:05:29 are so long overdue that they’re going to be changed. But I think it’s also smaller things like,
0:05:33 you know, the program of record was developed when you had to build out these very, very large
0:05:38 platforms and you had to plan years and decades in advance. And when someone won a program, it was
0:05:43 understood that they were going to run that program for decades. And now the Army has ways to acquire
0:05:48 things where technology is changing at a pace and at a speed that really needs to have a genuine
0:05:53 competition every year, every couple of years. And so that really benefits all startups. That benefits
0:05:57 all, you know, incoming emerging technologies that are going to serve sort of the fight of the future.
0:06:03 They have specific call-outs that they’re pointing to now, but definitely this is great news for
0:06:07 startups because what it’s showing is that there is actually the will inside the DOD.
0:06:13 Is there any movement on procurement reform? I remember I watched this hilarious movie, Pentagon Wars,
0:06:18 all about the development of the Bradley fighting vehicle. And it’s a very funny movie. You know, everyone
0:06:24 has a different requirement. They all get put together and becomes this kind of platypus of a vehicle that is
0:06:30 part tank, part troop transport, all these different problems. Part of the benefit of modern technology is that
0:06:37 we do develop platforms and things like andurals, ghosts can do ISR and also do munitions and a whole bunch of
0:06:43 things. There are projects that do need flexibility, but is there a cultural shift around moving away
0:06:51 from exquisite systems or just when folks in defense tech say we need procurement reform, what are they
0:06:55 really talking about in 2025? Yeah, well, I think they’re talking about different things because I
0:07:00 think what this initiative is going to do is it’s going to allow the army, and I think there’ll be a lot
0:07:03 of, you know, replicas of this as well. I think other branches will look at this and say, this is a great
0:07:10 idea. Instead of being locked into a program for decades, they’re going to be able to say, actually, we would love to use
0:07:15 that capital for something new. We would love to recompete that program. We would like to be able to be better
0:07:20 capital allocators because now their hands are tied. And I think when you talk to people who are just in
0:07:26 normal business, not in defense world, and you say, hey, if you had to make a decision about a purchase
0:07:32 that’s going to last for 10 years and get no updates and you would not be allowed to change it, what would you
0:07:36 do? We would say that’s insane. Like, how is a CEO going to say they’re going to acquire technology for
0:07:40 their company that they’re going to use for 10 years and there’s going to be no software and updates, no
0:07:45 nothing. And if it’s not working, you can’t get rid of it. You’re told you can’t get rid of it. I mean,
0:07:49 that is literally what the DOD has to deal with. And so I think what’s great about this initiative,
0:07:54 again, it’s one, the fact that the army is going public says that they mean business and that they
0:07:59 have air cover to do this. But I think that the meta story that we’re going to tell ourselves is
0:08:03 Doge has been very public in the last week of what they’re doing. There’s been some pushback on,
0:08:08 you know, why are you working on IT systems? You know, everyone has sort of their favorite Doge
0:08:12 meme of why it’s not working. But the story of Doge, I think when we look back even in a year,
0:08:16 is going to be that it gave extraordinary air cover to reform in every department.
0:08:21 And the first example inside the DOD, this is the biggest example in the last hundred days,
0:08:25 to see General George out there saying, like, this is what we need to do and we are committed
0:08:29 and we’re going public because we are so committed, which doesn’t usually happen.
0:08:34 I just think it speaks volumes and tech should be celebrating. This is a big, big day for everyone
0:08:38 in the American Dynamism ecosystem, for every defense company that’s been fighting for this
0:08:38 for a long time.
0:08:39 Yeah.
0:08:44 Yeah. And for all of this to actually achieve those sort of 30 year goals or execute against
0:08:49 that 30 year plan, things need to be bipartisan. People need to realize we want efficiency and
0:08:51 innovation across every branch.
0:08:56 I’m curious on the investing side, I’m sure you have this painful experience all the time
0:09:01 where you meet companies that probably are going to be great businesses, are good for America,
0:09:08 but maybe aren’t a fit for venture. What’s your sort of updated thinking on understanding if
0:09:13 something can be a great, important business versus something that can truly be, you know,
0:09:14 a generational outcome?
0:09:20 Yeah. One of the biggest mistakes I see investors make is trying to predict TAM. So early, early
0:09:25 days of Andrel, a lot of people, you know, didn’t want to look at Andrel because of ethical reasons
0:09:29 or because they were worried about being involved in defense. But there was another meme that was
0:09:33 going around, which is almost comical now, which is, well, it’s kind of a small TAM, right? Like
0:09:39 a border security company, like, oh, they’re selling to DHS, Department of Homeland Security doesn’t
0:09:44 really have that big of a button. Like, these were real things that people said that are hilarious
0:09:50 now, you guys can imagine. So I think it is very difficult to predict a growing market, eventually what
0:09:54 some of these incredibly important technologies are going to be worth. But I agree with you, there are some
0:09:59 examples of companies that might not be, you know, standalone businesses, but will ultimately, you
0:10:03 know, Andrel’s done a very good job of acquiring businesses that aren’t going to be these venture
0:10:08 outcomes, but work, you know, very well within their platform. But I think in some ways, there’s
0:10:13 always surprises with companies that were initially passed on or people were very skeptical of their
0:10:17 TAM in the early days. And then you look back and you see just how much they’ve grown or how much the
0:10:20 product has shifted or how important the platform actually is.
0:10:25 Yeah. How have you been, you know, ignoring the politics of it all, but reacting to, you know,
0:10:31 the trade war in many ways, like when you have these like big geopolitical, you know, events playing
0:10:36 out, that doesn’t necessarily mean start to make a lot of venture investments because venture investments
0:10:41 take a long time to play out and, and it’s very hard to predict the future. Are you seeing new opportunities
0:10:46 related, you know, to the events of the last month? Are you still just, you know, continuing?
0:10:52 I imagine when you guys invested in Hadrian, you weren’t like betting on a trade war in
0:10:57 two years or something like that, right? But how do you think about timelines and is the benefit of
0:11:03 sort of thinking in that 30 year timeline that you’re kind of able to broadly, you know, ignore or not place
0:11:08 too much focus on the headlines of today and just think about what America needs in the long run?
0:11:14 I would say my bias as a very early stage investor is to not think about the immediate timeframes.
0:11:19 These are very long cycles. You can sort of see trend lines, but it’s hard to know what actual
0:11:23 events are going to happen, obviously. So I think even when we made the investment in Hadrian, as you
0:11:27 called out, there was a movement towards re-industrialization and towards investing in
0:11:31 manufacturing that was early and nascent. But if you were hearing the signs or spending a lot of time in
0:11:37 D.C., both sides were very focused in Washington on how do we think about investing in America,
0:11:41 re-industrialization, how do we bring back manufacturing? So, you know, it felt like it
0:11:45 was a message that was being heard then. It’s just, of course, been accelerated. And I think truly,
0:11:50 if you think about the next 10, 20 years, re-industrialization is going to be a very important
0:11:55 theme. So, you know, it can feel like everything is hot right now or feel like we’re in the middle of
0:12:00 something. But ultimately, I think we’re, again, in this like very, very early, you know,
0:12:04 three or four years into a 30-year journey, you know, it took decades for globalization to really
0:12:08 hit its peak. And now we’re sort of seeing the pendulum swing again. And so you’re going to see
0:12:12 a lot of companies that are built in the next few years that become generational companies.
0:12:19 Yeah. How do you think about the kind of broader market map of American dynamism? Obviously, Anderill is
0:12:25 like just a great case study in the American dynamism thesis. But at the same time, as you go through
0:12:29 the American dynamism website, you can go back to like the moon landing and the development of the
0:12:35 iPhone as like examples. At the same time, there’s this question about like the Anderill of X is Anderill
0:12:41 potentially. But then that doesn’t always come true if you’re talking about something that’s truly
0:12:46 outside of their purview in consumer or in, you know, Flock Safety or Hadrian. These companies are not
0:12:51 competitive, but maybe fit in the thesis. How are you seeing the investing landscape of American
0:12:59 dynamism kind of evolve as more people come into the to the category, but then think outside the box
0:13:04 and address different issues? I mean, I’ve seen even like some education stuff kind of fit the broader
0:13:06 thesis. So how has that evolved over the last couple of years?
0:13:09 Yeah, we define it as companies that are actively supporting the national interest.
0:13:14 So it is a very simple definition and founders have, you know, different interpretations of
0:13:19 what it means. But there’s common themes. And actually, this goes into why we decided to have
0:13:23 a separate fund, why we decided to build out the platform is because these companies need something
0:13:28 entirely different than a true enterprise or a true consumer company. And when we looked back at
0:13:34 our early portfolio of Shield AI, Anderill, Astronus, these companies that were sort of what I would call
0:13:39 Space and Defense 1.0, we’d sort of like put them in the enterprise category as though they’re like
0:13:44 no different than a company that’s selling business software to the Fortune 500, right? It doesn’t make
0:13:48 any sense. They have totally different needs. You know, Anderill famously said that they had a lobbyist
0:13:54 on staff on week one. There’s things that companies need that our view is that we could build a platform
0:13:59 to help support these companies, namely in Washington, understanding who their buyers are on the BD side,
0:14:04 which is a very difficult kind of role to hire for inside of early stage startups. But then also
0:14:08 understanding the Washington game, which is very important for companies to understand if they’re
0:14:12 going to be selling directly to the federal government. Now, you mentioned education, and
0:14:15 there’s a lot of companies in our portfolio, too, that are selling to state and local. And that is
0:14:21 a totally different sales motion. You know, that is something where, you know, a company like Flock Safety
0:14:27 has sort of rewritten the rules of how you sell directly to a police force or how you even follow what I would
0:14:31 call kind of like a second city strategy of not going to the biggest cities, but going to these
0:14:36 smaller municipalities and getting a lot of, you know, almost circling a big city with the suburbs
0:14:40 around it and kind of getting a lot of momentum from the citizens. But all these companies have very
0:14:46 similar needs and sort of things that they have to think about early rather than later. And we’ve now
0:14:51 seen enough of sort of the early success stories and public safety and, you know, aerospace defense,
0:14:55 like sort of these generational companies that came up in the last several years that the boom that’s
0:15:00 happening in these categories, many of them want to replicate those playbooks and have,
0:15:02 I think, with a lot of success.
0:15:07 Yeah. Can you talk a little bit about almost like lobbying as value add for venture capital?
0:15:13 I remember I was running an Andreessen-backed company a decade ago, and I met the CEO of McDonald’s
0:15:20 through Andreessen at some happy hour. And there were trainings on B2B sales and PR and all this stuff,
0:15:25 but there was no concept of regulatory or lobbying. But I imagine that’s a piece of it,
0:15:30 but it’s at the same time, you need to eventually staff your own government affairs team. How are you
0:15:34 working with early stage founders to get them up and running in Washington?
0:15:39 Yeah, I would say a lot of the founders that we backed are very, I would say, sophisticated in their
0:15:43 knowledge of who they need to be meeting with or the types of people they should be meeting with in
0:15:48 the DOD. But the thing that I think we’re actually, I would say, even more successful in doing that’s
0:15:53 really important is making all that knowledge public. You know, we make our playbooks public.
0:15:59 My partner, Layla, who runs our go-to-market in DC, she wrote this incredible glossary of things you
0:16:03 need to know if you’re even going to approach a venture capital firm, you know, about a defense
0:16:07 tech company. Like, these are the acronyms. These are all the acronyms you could possibly hear in a
0:16:11 conversation with the DOD. And it’s things like that where we do want to make that public, and we want
0:16:16 to help educate the ecosystem. And I can tell you, like, you know, five or six years ago, the number of
0:16:21 venture capitalists who understood the different types of contracting vehicles, that understood the names of,
0:16:25 you know, of different people on the Appropriations Committee. I mean, these things that are now sort of,
0:16:29 I’d say, taken for granted, were not well known. And so I think that’s a huge part of it, too,
0:16:34 is really helping the ecosystem get up to speed, helping companies sort of speed run that early
0:16:37 stage process of, like, you can ask any dumb question, and we’re going to help you with it.
0:16:42 But then there is also something to be said of, it is much easier to get a meeting with certain people
0:16:47 if you are at a dinner that’s sponsored by a group of people who are always in Washington. I mean,
0:16:51 we have a Washington office now. We are fully staffed in terms of both Republicans
0:16:55 and Democrats and people who work on both sides of the aisle, people who specialize in
0:16:59 DODs, people who specialize in certain types of the DOD. And I think that is like a very important
0:17:04 thing to be able to say, okay, you need to meet with X, Y, and Z people, or you need to understand
0:17:06 the glossary before you can even begin to have those conversations.
0:17:12 Do you think defense tech is now mature? It’s oversaturated? I was joking with Jordy that I
0:17:17 think world peace is like maybe six months away. And then I’m going to start poaching top defense
0:17:22 tech talent to build the next generation of advertising optimization.
0:17:26 Because I think that we just got to get them back in the ad optimization game.
0:17:26 Get them back.
0:17:28 Get them back in the next company. You know it.
0:17:31 Yeah, yeah, yeah. But I mean, seriously, like it does seem like…
0:17:35 That would be the most, you know, the open AI, you know, AGI is always six months away.
0:17:41 You know, defense tech founders need to just go like, yeah, just 2 billion more and like world peace.
0:17:44 World peace, yeah. But I mean, there is a serious question here. Like,
0:17:48 I know some people who are like, just so excited that they’re jumping into things, but you know,
0:17:51 I’m not even in the industry, but I’m a little bit more tapped in. And I’m like,
0:17:55 there are already seven companies working on that exact thing. I don’t know if this is the best time.
0:18:00 Is it worthwhile to steer, you know, these incredible hackers, these great entrepreneurs,
0:18:06 like maybe towards the more tangential hard tech problems? Like what I see with like,
0:18:11 what base power is doing is like, it’s hard tech, it has defense roots, but it’s not directly
0:18:16 something that’s on Andrew Roll’s roadmap. What advice are you giving to kind of the entrepreneurs
0:18:19 that are like in between things, thinking about serving the national interest, but not
0:18:24 necessarily putting themselves on a collision course with a, you know, multi-billion dollar
0:18:29 founder mode company? So I’ll say deterrence is the constant project, right? So like your whole,
0:18:33 like the meme of maybe six months away from world peace. Yeah, of course.
0:18:36 I actually think that was part of the problem in the nineties, right? Like, like, seriously,
0:18:38 that was the big problem. Yeah, people thought it would be over. Yeah, democracy will-
0:18:39 The end of history. End of history.
0:18:43 Yeah, end of history. We’ve flourished and we don’t need to be working on these things. So
0:18:47 it is very important that we’ve gone back almost to the roots of the DOD saying like,
0:18:51 hey, actually we remember what it’s like to be a country at war. And we need to be constantly
0:18:55 focused on the next technologies. We need to be focused on deterrence, thinking of it as deterrence,
0:19:00 because we want to prevent war, but we have to be continuously building. So from that perspective,
0:19:05 I think, again, we’re only a few years into this real movement of Silicon Valley caring about
0:19:09 working with the DOD. And I hope that it’s a 30 year project. I think that’s what we all really
0:19:13 should be focused on is making sure it’s a 30 year project and even longer than that. But to your
0:19:17 point, what I think is so interesting about companies that, you know, are founded out of
0:19:21 Anduril or out of SpaceX, you know, we’ve done an analysis where we looked at all of the founders
0:19:26 who’ve left SpaceX in the last say 10 years. And there’s hundreds of companies that have been formed
0:19:32 in just wildly different sectors, whether it’s, you know, Radiant Nuclear, working on nuclear energy,
0:19:35 you know, Castellion, which is in our portfolio and they’re building hypersonic weapons. I mean,
0:19:39 some of the best founders are trained. I always say they go to the school of Elon Musk,
0:19:43 they learn manufacturing, they learn production, and then they want to take that to something that,
0:19:46 you know, is pretty low hanging fruit. Like they want to make sure that they’re competing against
0:19:51 the incumbents of yesterday who have not modernized their production, who’ve not modernized a lot of
0:19:56 the technology that they’re working on. And so, you know, I think you see that with a lot of the,
0:20:01 you know, yes, there are some extremely crowded fields, but then there are also areas of defense that
0:20:04 that are really just boring and completely untouched. And you’re seeing founders realize
0:20:09 that too, that it’s, it’s not something that that’s, you know, interesting to any of the existing
0:20:12 companies and it’s low hanging fruit. It’d be interesting to work on that. Or they’re interested
0:20:16 in being a tier one supplier. We have a number of companies that are really focused on the supply
0:20:21 chain aspect of defense and their partners to Andrel and their partners to SpaceX and other
0:20:26 companies in the ecosystem. So you really are seeing founders like understand that question in a very
0:20:30 sophisticated way and say, okay, we’re going to go after the parts of the supply chain or the things
0:20:34 that the DOD needs that no one is focusing on. And that’s been exciting to see too.
0:20:42 Can you talk about M&A in defense tech broadly? Andrel’s done this very well. Saronic announced a deal
0:20:48 last week acquiring Gulfcraft. That feels super significant. I’m curious, you know,
0:20:53 how you advise founders kind of broadly when thinking about that. We actually had Augustus
0:20:58 on from Rainmaker earlier who had acquired a company in his space, but when’s the right time to be,
0:21:03 you know, thinking about that as somebody in defense tech and yeah, what kind of opportunities
0:21:08 do you think make the most sense? Totally. Well, I think, I mean, both Andrel and Saronic,
0:21:13 they have incredibly unique stories in terms of where they’re operating and sort of what they need
0:21:17 to do in order to grow and scale. And they’ve done it at a speed that is just incredible, right? Like
0:21:22 they have very sophisticated teams that know a lot about acquisition. I’d say for earlier stage companies,
0:21:27 we’re seeing more companies that are potentially interested in doing that. It can speed up innovation.
0:21:30 It can speed up being able to work with certain customers. That’s for sure. If you’re acquiring a
0:21:35 certain capability so that you can sell to a major prime, that’s something we’ve seen more of too,
0:21:38 which is interesting and exciting. Like I don’t think we were seeing that several years ago and
0:21:42 now we’re certainly seeing companies experiment with that. But when you said actually M&A, I actually
0:21:46 thought you were going towards something that I think is actually more likely to happen in the future
0:21:51 that hasn’t happened in a long time. When you look at these existing prime companies, the big five say,
0:21:55 they’ve really only acquired companies that have not raised any venture dollars, right? Like they don’t
0:22:00 acquire companies that are kind of seen as these bleeding edge companies to shore up their capabilities.
0:22:04 And my instinct, you know, we’re talking about army transformation initiative, we’re talking about a
0:22:09 government that’s becoming far more sophisticated and a DOD that’s becoming far more competitive,
0:22:13 right? It hasn’t been competitive for decades. And now you’re seeing all of these startups come in.
0:22:18 My prediction, if we’re looking five, 10 years out, is that the companies that have not been
0:22:23 acquisitive for the best engineers and the best technologists and these capabilities that they need
0:22:28 are going to find that as their only solution. And I think we could potentially even see another
0:22:32 Last Supper situation, which, of course, in the 90s was the famous case where the government came to
0:22:37 all these primes and said, you have to merge, you have to have kind of forced mergers and acquisitions
0:22:41 because the budget’s going to decrease. And of course, that was probably the wrong strategy,
0:22:45 given sort of the results that came out of that. But I do think it is something that I would not be
0:22:50 surprised if in five or 10 years, you’re seeing the existing primes that have been around in many
0:22:55 cases for a hundred years saying, we have to work with these startups in a much more tangible way.
0:23:00 And you could see a highly acquisitive ecosystem that don’t necessarily kind of write into their
0:23:05 kind of thesis today. How do you think about leadership at the individual primes? And,
0:23:10 you know, people over the last few years, I mean, Boeing has been dragged through the dirt
0:23:15 by pretty much everyone. But I think of it as a great, in the fullness of time, it’s a great company.
0:23:19 I ain’t going. Yeah. John is John so loyal. He’ll never fly out of excitement.
0:23:20 He’ll never fly out of excitement.
0:23:25 As a white collar worker, you know, you don’t risk your life very often. When I go on a business trip
0:23:28 and I step on a 737 max, I’m locked in.
0:23:34 No. And I mean, I just look at it as China would love to have a company that was actually competitive
0:23:40 with Boeing, right? It’s a hugely strategic asset. But I’m curious, do you think that, you know,
0:23:44 any of the primes, you know, and every now and then you’ll see a prime release a video that’s like,
0:23:49 clearly they hired a marketing agency and said, like, make us like an Andrewle movie, you know,
0:23:53 and then they put it out. But do you see that the leadership at the primes?
0:23:56 Well, Lockheed Martin invented artificial intelligence. Remember?
0:23:59 Yeah, they came out last week and claimed that they invented artificial intelligence.
0:24:00 They basically just said, you’re welcome.
0:24:03 You’re welcome. Yeah. By the way, you’re welcome.
0:24:08 But I’m curious, do you have conversations with them or is it, is it?
0:24:13 Because, I mean, even though it’s not an opportunity for, you know, venture capitalists
0:24:16 necessarily, like, it would be great if they were highly functioning in the American interest,
0:24:17 like as Americans.
0:24:21 And then you have the, you have the program that was spun out of, you know,
0:24:23 was it Microsoft or Microsoft to Andrewle?
0:24:25 Oh, yeah, yeah. The HoloLens. Yeah, sorry.
0:24:29 So I think there’s probably more kind of even spin out opportunities where new companies
0:24:33 can create, you know, value on top of existing programs.
0:24:37 Yeah, I think, you know, Palmer and actually Brian Schimpf has done an incredible podcast on this,
0:24:41 where he talks about sort of what happened at these primes and why things sort of went by the
0:24:45 wayside. And it’s partially because they really stopped focusing on research and development.
0:24:46 Yeah.
0:24:48 They started, they didn’t really need to. There was no real competition.
0:24:53 And they kind of recognized that, that, you know, they would always get paid by the government to do
0:24:57 new things. You know, again, like, it’s sort of this confluence of factors that led us to be,
0:25:02 I don’t really complacent. But I went back actually last night and was reading the first few pages of
0:25:06 The Kill Chain by Christian Brose, which again, reading it, it was written, I believe in 2019,
0:25:11 things changed so dramatically in terms of the conversation. But it’s like going back in a time
0:25:15 warp and saying, wow, like in 2019, people really didn’t care that Boeing was collapsing or that
0:25:19 there were these private or these public companies that were doing no research and development because
0:25:23 it didn’t matter. Right. That was pre-war in Ukraine. It was sort of, you know, in some ways it was
0:25:27 security theater. Right. Like we don’t actually have to remain secure. We just have to pretend we’re secure.
0:25:32 And so I think there is this new sort of wake up call where a lot of these companies are going to
0:25:36 say, one, if we, if we can’t recruit the engineers and do the research and development in house, we’re
0:25:40 going to have to acquire it. So again, that’s why I think you’re going to see a lot more acquisitions
0:25:44 over the next several years, because I think a lot of these companies are really going to have to
0:25:49 change. But two, like these initiatives inside the DoD that are now getting real steam, that is going
0:25:54 to force incredible competition that has not existed, even in the last 10 years when we’ve all been
0:25:59 investing in American dynamism. So I’m, I’m actually much more like hopeful and excited about where I
0:26:02 think the world is going, because I genuinely believe that a lot of these players have sort of
0:26:05 woken up and are looking for solutions because now they know they have to.
0:26:10 A while ago, I was talking to Trey about just the lack of the deeper supply chain,
0:26:15 specifically in drone motors. Like there are no small drone motor manufacturers in the United States.
0:26:19 They’re almost all made in China. And that feels like, oh, there’s almost a startup idea there,
0:26:25 but I don’t know if it’s a venture idea. There’s actually a drone motor company in Washington.
0:26:30 They outsourced some of their supply chain recently. That feels like almost like we need an American
0:26:36 dynamism, private equity fund to just turn those companies around. They’re not going to be these
0:26:41 power law, a hundred billion dollar companies, but they might produce 20% returns more reliably. And
0:26:47 there’s maybe no venture style, zero loss of capital risk. Do you think we need a American dynamism for
0:26:52 private equity? Is that something Andreessen would do at some point? I mean, you’re kind of in every asset
0:26:58 class now. So anything’s possible, but is there a flip side to the venture model within investing in
0:27:01 the national interest? Well, I certainly think we’ve invested in some companies that are focused on
0:27:06 component parts. You know, we’re invested in AMCA. I know that Jay was on recently. So like there are
0:27:10 more and more companies that are figuring out to do this. And again, those are the examples of companies
0:27:15 that are, that are, you know, much more focused on how do we acquire companies? How do we make them,
0:27:19 I would say tech forward, but also think about like how quickly we can get into the supply chain and some
0:27:23 of these larger primes. But I think you’re seeing a lot of innovation around the edges on this and
0:27:28 you’re probably going to see more and more founders who recognize that if that’s where the real problem
0:27:31 is, they’re going to build there and they’re going to build in the best way that suits them.
0:27:36 So yeah, it does seem like there’s almost like a way to turn something like MP materials. We were
0:27:41 talking about like, you wouldn’t think like, oh yeah, venture is suitable for like mining at all.
0:27:44 But like now there’s a couple of mining companies that are figuring out how to inject enough
0:27:49 technology to make it potentially a venture scale opportunity, which is interesting. Do you have
0:27:54 anything else? I have a couple more. I got a totally switching gears, but you had a post recently
0:27:58 that I liked. It was, I’m committed to doing whatever the opposite of gentle parenting is.
0:28:06 And I wanted to ask you if you found any Lindy books on parenting, anything that’s sort of resonated
0:28:12 that you’re implementing. John and I both have similar aged children. And I always have this,
0:28:17 you know, sort of concern around, you know, you want to experiment, you know, with parenting and
0:28:23 try new things and maybe not just take exactly what the mainstream media says is the right way to do
0:28:27 parenting. But then, you know, your children have one life. You’re not trying to run A/B tests,
0:28:28 you know, on there.
0:28:33 I have three boys. So I employ what I call the snake pit strategy, which is you lock them all in
0:28:37 a room and then they, it’s just a snake pit and they just like wrestle. And you know, if there’s
0:28:39 damage, they’ll heal. And that’s fine.
0:28:44 That’s the right way to do it. Well, I followed up that tweet with the tried and true Irish strategy,
0:28:50 which is the hay method. You just shout hay and preach loud. Yeah. Hey, hey, hey, you know,
0:28:55 it works. Like there’s something about the word hay where your sons actually turn around and listen to
0:29:00 you. But sadly, I, you know, there, there aren’t like any books, like old timey books that I found
0:29:05 that actually teach, I would say the best way to train children or to child rear. But you know,
0:29:09 it’s, it’s interesting. I always think that grandmothers kind of know best. So there’s a
0:29:14 grandmother in your life anywhere. They remember how it used to be done and how effective it was.
0:29:18 And it was, you know, probably harder in the olden days too. So it’s like, basically just ask
0:29:22 grandma, like grandma. I would plug free range kids all about, like, our society has moved
0:29:25 towards like, don’t let the kids just run around in the neighborhood. They could get kidnapped.
0:29:28 There’s so many bad things that could happen. There’s been a lot of fear mongering from the
0:29:35 media. And so that’s kind of led to kids turning in, in, inside becoming inside kids, staying on the
0:29:40 iPads or whatever. But there’s this movement in the free range kids to just be like, yeah, actually,
0:29:44 like you’re six, you can ride a bike, like ride your bike to the park. Like, and that will enforce the
0:29:50 society to maintain safety. I need to find the repeat of parenting. Yeah. That’s the next alpha.
0:29:56 Yeah. My problem with the grandma method is that my mother and mother-in-law just want to let the
0:30:00 kids do exactly what they want to do. You wanted two cookies. You want three cookies. So maybe they’re,
0:30:02 maybe they’re right. Maybe that is Lindy. Who knows? Maybe it is Lindy.
0:30:06 I’m the great grandmother, right? Like the, the one who remembers how tough it was.
0:30:10 Yeah, that’s right. I want to get your reaction to Warren Buffett. Obviously he stepped down
0:30:17 or announced his transition at Berkshire Hathaway this weekend. What do you take away from Buffett’s
0:30:22 legacy as an investor? It’s obviously a very different type of investing, but there’s so many
0:30:28 interesting lessons there from company building to investing to everything else. What was your reaction?
0:30:31 Yeah. You know, I’ll take a little bit of a different take because I was watching, you know,
0:30:36 the annual meeting last year and there was this moment that happened and I actually wrote about it
0:30:41 and a piece on friendship and founder friendship where he was doing his usual, you know, going
0:30:45 through company analysis. And then he just kind of forgets where he is and says, “Charlie.” And
0:30:50 everyone stopped. It was like, you know, I think I, I think I tear up seeing it because he was so in his
0:30:55 zone after so many years of working together, he had forgotten that Charlie had passed and almost
0:30:59 embarrassed about it. But I thought it was the most beautiful moment because one of the things I don’t
0:31:04 think we talk enough about in Venture World is founder friendship. And I mean like deep,
0:31:07 deep, deep friendship, not like, “Oh, we went to college together and we were friends or whatever.
0:31:11 We’re going to start a startup together.” I mean, those people who like work together decades and
0:31:16 decades out, I actually think this is why family businesses often work better, where even if you
0:31:20 look like the Collison brothers, it’s like they’ve been sharing resources, you know, since childhood,
0:31:25 since they can remember. And like, there’s something about going through life with someone,
0:31:30 suffering with someone, understanding how to like, you know, end someone’s sentences that leads to these
0:31:34 just incredibly rich and beautiful companies. And I think if, you know, if we did an analysis
0:31:39 in Andreessen Horowitz and just looked at the companies that were true outliers, I think there
0:31:43 would be stories of these people are like brothers and sisters. And Andrel’s certainly this, right?
0:31:47 Like it’s, you know, the founders there were DARPA challenged together, like their first day of
0:31:52 college, right? In some ways, there’s something about just having these deep relationships that span the
0:31:58 test of time where you’re on a journey with someone and it’s real like Aristotelian friendship, not like
0:32:03 faux friendship, but true love. And clearly you saw that with them. It’s just a remarkable thing how they
0:32:08 were able to kind of be true brothers and kind of, you know, each other’s better half throughout their
0:32:09 business career for as long as they were.
0:32:12 It’s amazing. Last question. What should we do with Alcatraz?
0:32:18 Oh, you know, I, I love all, I love all the ideas of turning it into a casino, but I haven’t seen that
0:32:24 one. I like, I was saying, I was saying tax haven and, and no, no general solicitation rules. So you
0:32:30 can like go out there and sell your angel. Lock up periods, just unfettered libertarian capitalism out
0:32:35 there. That sounds good. But there is something about bringing it back in its original form. You know,
0:32:40 it’s like, there is something about these buildings that, that the president likes to restore into their
0:32:44 former glory. And so if Alcatraz is the case, like to keep the historical details accurate,
0:32:48 you can kind of see where it’s coming from. He’s, he’s definitely a historicist in that, in that
0:32:54 regard. Okay. Well, thank you so much for joining us. This was great. Come back on again. Thanks for
0:32:56 having me. Have a good one. We’ll talk soon.
0:33:03 Thanks for listening to the A16Z podcast. If you enjoyed the episode, let us know by leaving
0:33:09 a review at ratethispodcast.com/a16z. We’ve got more great conversations coming your way.
0:33:10 See you next time.

0:00:05 ICP is the central nervous system of the entire customer journey.
0:00:10 Conversion rates, expansion rates, the length of your sales cycles.
0:00:15 If you look at those, they’re almost all telling you something about your ICP at all times.
0:00:19 If you’re like, “We sell to everyone, we build to everyone for everyone,”
0:00:22 maybe you’re not so on to anybody yet.
0:00:28 It not only shows who you need to target and what you need to target the customer with,
0:00:33 but also why the customer would need your product.
0:00:37 Your ideal customer profile, or ICP, is the lodestar of your company.
0:00:41 It defines who you’re building, marketing, and selling your products to.
0:00:45 And most growth stage founders think they know who their ICP is
0:00:47 because they found product market fit after all.
0:00:50 I don’t know that you necessarily have found product market fit.
0:00:55 Sometimes you just get really lucky and I call that the curse of early success.
0:01:02 And here’s the thing: very few can define and refine their ICP well enough to keep the company focused on it as they grow.
0:01:06 This lack of clarity can open a Pandora’s box of problems across the org.
0:01:09 Pipeline not getting filled? Chances are it’s an ICP problem.
0:01:12 Product roadmap stalling out? ICP problem.
0:01:15 Marketing spend through the roof? ICP problem.
0:01:18 You could be missing a huge market opportunity if you misidentify your ICP.
0:01:24 In this first episode of A16Z Growth’s New Company Scaling Series, The A16Z Guide to Growth,
0:01:29 we take a step back and explain why understanding your ICP should be a company-wide effort
0:01:33 and why getting this right is even more important in the AI era.
0:01:38 As soon as you have a successful product, there will absolutely be competition in the market.
0:01:45 A16Z Growth Partner Emma Janoski sits down with Growth General Partner and former CRO of Segment, Joe Morrissey,
0:01:52 but also A16Z Partners Michael King, who was at Gartner before building full-stack marketing teams at companies like GitHub and VMware,
0:01:57 and Mark Reagan, who was most recently the VP of RevOps at Segment.
0:02:03 Together, they dive into what truly makes a great ICP, including what it is, but also what it isn’t.
0:02:08 Is it meaningfully different than the way you might talk about segmentation or a psychographic or firmographic,
0:02:12 or is it a constellation of all of those things put together?
0:02:20 They also tackle how you know if you’ve outgrown your existing ICP and how and when to define but also redefine it as you scale.
0:02:25 They touch on how to make some hard decisions when you’re implementing a new ICP, like saying no to customers,
0:02:28 and how it shows up in the business when you get it right.
0:02:33 The first voice is Emma’s, then we have Michael, Mark, and then finally, Joe.
0:02:35 Let’s get started.
0:02:44 As a reminder, the content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice,
0:02:50 or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund.
0:02:56 Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast.
0:03:03 For more details, including a link to our investments, please see A16Z.com/disclosures.
0:03:08 Why are we talking about ICP?
0:03:13 If I am a growth stage founder, I found product market fit.
0:03:15 I probably know my ICP, right?
0:03:16 I’m growing, I’m scaling.
0:03:17 Why do I need to care about this?
0:03:21 I don’t know that you necessarily have found product market fit.
0:03:26 Sometimes you just get really lucky, and I call that the curse of early success.
0:03:33 Sometimes you just have such a compelling product, or you sometimes just have such a compelling story around the product that people gravitate towards you.
0:03:35 And you don’t necessarily have to have an ICP.
0:03:38 You’re just selling to anybody who will come through the front door.
0:03:40 And if that’s successful, you’re like, “Great.
0:03:41 I’ve nailed my ICP.
0:03:42 It’s everybody.”
0:03:50 But when you start to refine your sales process, you start to bring in more people to sell the product, it starts being not led by a founder.
0:03:58 When a conversation is led by a salesperson, or perhaps an SDR, or perhaps someone else like that, all of a sudden, that ICP becomes a lot more important.
0:04:00 You’ve got to scale a marketing program.
0:04:01 You’ve got to scale a sales team.
0:04:03 You’ve got to scale a bunch of other things like that.
0:04:05 All of a sudden, you’ve really got to know who you’re talking to.
0:04:06 Why are they buying?
0:04:07 What are they buying?
0:04:09 How are they selling it internally?
0:04:24 Before I came into this role, when I was an operator in the companies I had been in before in revenue operations, I don’t think I had enough respect for just to what extent the ICP is the central nervous system of the entire customer journey.
0:04:39 Almost every KPI that everybody is familiar with today, when you think of conversion rates, expansion rates, the length of your sales cycles, any of these things that normally would signal to you, “Oh, we have a pipeline issue,” or “We need more SDRs,” and things like this.
0:04:44 The truth is, if you look at those, they’re almost all telling you something about your ICP at all times.
0:04:53 So, for example, there are some companies I’ve been working with over the past couple of years where, over time, they start to see declining pipeline conversion rates.
0:04:59 And you start to circle around that and look for things you can do better tactically in that, “Well, we need a little bit better messaging here.
0:05:04 We have this laggy process where there’s too much time where leads are in this or that queue.”
0:05:07 And those things still could be true.
0:05:16 But in almost every one of those cases, what you usually find is that you have your sales force often talking to the wrong people with the wrong message.
0:05:30 And it goes all the way back to that continuous alignment and always being relevant and targeting the best possible customers and the best personas within those customers with the best messaging and relevancy around how your product is going to solve pain.
0:05:35 And then when you get that right, all of those indicators will tell you whether you have it right or not.
0:05:36 They’re not going to lie to you.
0:05:41 If they’re hitting the targets you want, you’re doing a pretty good job of operationalizing your ICP.
0:05:49 But I think that focus on the customer journey is the reason why it’s so hard to diagnose an ICP problem because you’re looking at, “Oh, I’ve got an onboarding problem.
0:05:51 I’ve got a marketing problem.
0:05:52 I’ve got a sales problem.”
0:05:59 I think to your point, Mark, you’ve got to look at the entire customer journey from first touch through when they’re a million dollar customer.
0:06:06 And if you start to see issues anywhere along the way, first make sure you’ve got your ICP correct.
0:06:10 And then start to look at, “Okay, maybe I do have a customer support issue.”
0:06:16 Because if your ICP is missing, you can throw a lot more dollars at marketing or change your marketing mix up or do whatever you want.
0:06:20 And if you still got your ICP wrong, those are still going to be dollars you’ll spend.
0:06:26 If you’re like, “We sell to everyone, we bill to everyone for everyone,” maybe you’re not selling to anybody yet.
0:06:34 And I think the biggest downstream negative consequence is actually one where it’s missed opportunity.
0:06:37 So the biggest risk often is opportunity cost.
0:06:45 So when you’re chasing the wrong profile, your ideal customers might be adopting a competitor’s product or simply unaware of your solution.
0:06:49 And so you could be missing a huge market opportunity if you misidentify your ICP.
0:06:51 Let’s just even take a step back.
0:06:53 What is an ICP?
0:06:54 What counts as an ICP?
0:06:57 What does a good ICP look like?
0:06:59 What does a bad ICP look like?
0:07:09 An ICP should tell you not only who you should target, but also why those customers should need you.
0:07:17 So broadly defined, an ICP is a detailed description of the ideal set of customers for your product.
0:07:27 And it typically should include firmographic details, behavioral traits of those companies, and it should be as narrowly defined as you can get it.
0:07:43 A good example of that would be global B2C multinational corporations that are multi-brand, multi-product, and want to use data to go direct to consumer.
0:07:44 Yeah.
0:07:58 So you’re not only getting a sense of the type of company and the vertical and the industry that they’re playing in, but why they need your product, what the unique value for your product is.
0:08:01 It’s actually got to include defensible differentiation.
0:08:05 And so you think about what is defensible differentiation.
0:08:06 It comes in three forms.
0:08:08 One, there’s unique differentiation.
0:08:15 There’s the things that your product does and the pains that your product solves that your competitors can’t.
0:08:23 And then there’s holistic differentiation, there’s the things that your product does and your competitors’ product does, but maybe you do them better.
0:08:25 And then there’s holistic differentiation.
0:08:29 There’s the things about you as a company or your solution that are very different.
0:08:36 You might be the best funded or you may have the most experience in this particular domain.
0:08:52 And I think that is so critical in determining what your ICP is because you’ve not just got to look and identify those customers that have the biggest pain points that you can solve, but also the ones where you have the most defensible differentiation versus your competitors.
0:08:53 I agree with you 100%.
0:08:58 I think the more narrow you can define it at first, the better off you are.
0:09:02 The ICP should be, I’ll call them searchable metrics.
0:09:15 In other words, like if you define your ICP as companies that have these specific characteristics around how they think about customers and their buying patterns and things like that, how are you going to find them?
0:09:23 But if you say it is a customer with international capabilities of a certain size and a certain dollar percentage, then you can target them from a market.
0:09:35 Now, marketing data has gotten a lot better and customer data has gotten a lot better, but you still can’t describe in flowery language what you think this company might be feeling and thinking because you’re not going to be able to market to that, right?
0:09:40 You’re not going to be able to find them online or at conferences or any other places, right?
0:09:41 You need those objective qualities.
0:09:42 You need those objective qualities.
0:09:44 Those objective measurable qualities.
0:09:49 I think Gong was a great example of a very, very narrow customer focus.
0:09:56 Like they focused on B2B software companies selling, you know, at a certain amount of dollars with a certain amount of employees selling over Zoom.
0:10:01 And it narrowed it down to, I think the number was 5,000 total companies in their ICP.
0:10:03 And then they built from that base up.
0:10:09 I think the customer success platform I was talking about has done a really, really good job at narrowing down.
0:10:10 This is Pylon.
0:10:13 And what they’ve done is they’ve done a great job at narrowing down.
0:10:23 We sell to B2B software companies that are supporting enterprise companies with highly complex support flows and customer success teams and multiple teams involved.
0:10:30 So I think that narrow focus early on, particularly in competitive spaces, is really, really good.
0:10:38 I mean, I’m hearing all of this and it sounds like, yeah, you want to get pretty granular, but I think there are a lot of different ways to slice the question of who is your customer.
0:10:46 And so does the ICP, is it meaningfully different than the way you might talk about segmentation or a psychographic or firmographic?
0:10:50 Or is it a constellation of all of those things put together?
0:11:04 The ICP gives you a place to focus, whereas the personas give you a who to focus on and who to build value statements for and who to market to directly and sell to directly.
0:11:08 One is more about focusing, one’s a little bit more about building an audience.
0:11:09 You need them both, though.
0:11:18 So segmentation is really about identifying the groups of customers that would be interested in your product.
0:11:23 So you could say this is SaaS companies or telco companies.
0:11:37 ICP is much more narrow, and so you’re explicitly defining the type of company, the size of the company, the pain points that that company has, why they need your product.
0:11:40 And then personas are something different, right?
0:11:46 So personas are the individuals within those companies that own the pains that your product solves.
0:11:52 That can, in many cases, influence the choice to buy your product.
0:11:59 And very often are the folks who are actually going to make the decision to buy the product in the end.
0:12:02 Yeah, I mean, that encapsulates a lot of how I think about it.
0:12:06 I think your ICP has everything to do with the type of company that you’re selling to.
0:12:14 I think your persona does have to be very, very different because the persona is oftentimes encompassing a buying circle, not a single buyer.
0:12:17 Sometimes you have a single buyer, and that’s fine.
0:12:22 But oftentimes you have multiple buyers or multiple stakeholders in the buying process.
0:12:23 You have a security team.
0:12:24 You have an IT team.
0:12:25 You have a champion.
0:12:26 You have users.
0:12:28 You have influencers within that.
0:12:31 And I think you do need to separate those two out.
0:12:34 Is there a gut test that we could give to founders?
0:12:37 Like, a couple questions they should ask themselves.
0:12:39 Do you know who your ICP is?
0:12:40 I have five questions I ask.
0:12:41 Love it.
0:12:42 Let’s hear them.
0:12:47 So the first one is, which of your current customers makes the most out of your products and services?
0:12:48 Who uses it the most?
0:12:50 Who are your best users, your biggest users?
0:12:53 What traits do those customers have in common?
0:12:59 What reoccurring objections do you see when you lose an opportunity or when people churn?
0:13:02 Which customers are the easiest to upsell and why?
0:13:07 And what do the customers of your closest competitors have in common?
0:13:12 If they can answer all of those questions, then they typically will know their ICP very, very well.
0:13:17 Now, again, I have a list if they can answer all the questions, company size, industries,
0:13:22 have problems, company specifics, unique buying behaviors, type of business, all those types of things.
0:13:24 They can answer all those, then they’ve got that.
0:13:29 If they can’t, they’ll use those first five questions to find out what the ICP is.
0:13:30 That’s a great framework.
0:13:38 And I think it also alludes to one of the challenges you have when you’re working with startup companies or when you’re in a startup,
0:13:45 when a lot of what you’re trying to do there is you’re taking your best educated guesses at the answers to all those questions, right?
0:13:50 And it is something at first where you just don’t have a lot of feedback in the market.
0:13:52 You still don’t have a lot of customers yet.
0:14:12 You don’t have a lot of signals from all the different segments, the geographies, the individual personas, especially the earlier stage that we run into because you want to go as broad as possible because you’re trying to actually explore that product market fit that you have and start to express your growth in a bunch of different directions through that.
0:14:18 And you don’t want to bargain against yourself by becoming too precise too early.
0:14:24 However, I think if we’re all being honest with ourselves, usually focus is the problem, right?
0:14:43 It is usually the case that if you went and arbitrarily picked five different companies and looked at the way they are defining their ICP and going to market and operationalizing that ICP, they’re nearly always not as focused as they could be and thus not nearly as effective as they could be exploiting that.
0:15:12 And what happens over time is you have the ability to gather those and harness those and start to create a feedback loop that allows you to answer Michael’s questions, both in terms of the things you know and the things you believe as hopefully a leader in your space, but also due to what’s coming back from the interactions your sales force is having with prospects, what your customer success managers are having with your customers, which customers are expanding and why.
0:15:19 Why you have all of these signals out there that you can start to harness and bring into the answers to the questions in that framework.
0:15:25 Yeah, I think this is actually a pitch why you need a strong RevOps practice in your organization early.
0:15:33 And again, this is maybe a pitch for you, Mark, but I think you do need to have it because otherwise you’re taking the few customer conversations that you’ve had and that’s providing bias.
0:15:44 You need a standardized methodology of looking at these interrogating these and applying the right amount of recency bias to the organizations that you are interacting with most regularly.
0:15:49 I love, Mark, you were saying this is kind of like the central nervous system of your whole org.
0:15:53 The word that occurred to me was root cause of a lot of issues, right?
0:15:56 But if it’s so important, who’s responsible for defining it?
0:15:59 I think it’s really a company-wide responsibility, right?
0:16:04 I think in the early stages, for sure, it’s got to come from the founders, right?
0:16:08 And so there are kind of different stages of finding ICP.
0:16:17 So I would say pre-product market fit, you’ve actually got to be pretty open-minded about what ICP will end up looking for because you haven’t found it yet.
0:16:26 But once you have hit product market fit, I think it’s really incumbent on the founders to pay close attention to how they think this is going to evolve.
0:16:30 And then throughout the company’s journey, there are a couple of inflection points.
0:16:40 Certainly when you move from founder-led sales to a more repeatable motion, typically like around the series A, then you’re bringing in a sales team and it’s scaling, right?
0:16:48 And so you start to see sales leaders become very closely involved in defining and refining the ICP.
0:16:54 As you’re moving up market or down market, that has a massive impact on how your ICP changes.
0:17:04 Just ultimately, I think you very often see the responsibility for ICP, at least the growth stage, being shared at the executive team level.
0:17:05 I’d agree with that.
0:17:11 I think everybody has to own it, if you will, and everybody should have input to it.
0:17:14 Your sales team, they’re looking at six to 12 months.
0:17:17 The marketing team, they’re looking at 18 to 24 months.
0:17:18 Your product team, they’re looking at…
0:17:22 So I think the different kind of viewpoints are important to bring to the table.
0:17:27 But each of them are going to have input to that ICP.
0:17:33 And your customer success team is going to have all historical data of which customers have been successful, which have not, which have churned off, all those things like that.
0:17:40 And so again, that refinement process is going to occur if you continue to ask questions.
0:17:42 Well, why are these customers successful?
0:17:44 Well, how does the product service these?
0:17:45 To steal Joe’s line.
0:17:48 You keep asking why until you get to the root of the problem.
0:17:58 The other thing I would look at, and I think one of the rough sketches that I’ll do, is when I’ll put together an ICP, I’ll take a look at the TAM of that ICP.
0:18:03 And I’ll understand what is my fair percentage, what is my unfair percentage.
0:18:08 And if I’m capturing an unfair percentage of that TAM, then I know that’s probably a really good ICP for me.
0:18:16 If I’m unable to capture even my fair share, then maybe there’s not really good alignment between my use cases and that particular ICP.
0:18:19 Wait, and how would you be able to tell what percentage you can capture?
0:18:23 Is it because of the alignment of use cases with their pain points?
0:18:26 Correct. Use cases with their pain points and what the total TAM is, right?
0:18:30 And if you’re like, hey, if I’m alone in this market, then I should be able to win 80% of my deals on it.
0:18:34 If I’m one of four players, then, you know, maybe I should get 25% of that marketplace.
0:18:41 I’m wondering about the sort of product intuition, vision and mission here, which is kind of an X factor, right?
0:18:46 Where I can imagine founders getting a bunch of data saying like, this should be our ICP.
0:18:50 And founders thinking, that’s not really what I’m building or like, that’s not really what my mission is.
0:18:59 And so I’m curious if any of you have seen companies that have maybe gotten some data back and thought, actually, that’s not what I really want to build.
0:19:02 And I’ve gone on to do something else and been successful.
0:19:07 I’ve seen the opposite thing. I worked with a founder who was from the security world.
0:19:10 And they were building what they thought was a security product.
0:19:16 And they built it and they talked to a number of security buyers and none of the security buyers bit on it.
0:19:21 But what they noticed is that every single time they had a conversation with a security buyer, they brought in a platform ops people.
0:19:25 They brought in basically the platform operations people to either validate or have the conversation.
0:19:31 And eventually, after a couple of these conversations, the founder and I were talking and he said, you know,
0:19:34 I don’t think I’m building a security product. I think I’m building a DevOps product.
0:19:37 And we went through it and we said, well, who’s going to benefit from it? Who’s going to use it?
0:19:42 So I’ve seen that opposite problem where the product intuition and their history took them one direction.
0:19:47 But in reality, the customer feedback took them to the right ICP eventually.
0:19:52 What’s the outcome from an ICP exercise? Does your ICP fit on one page of a Google Doc?
0:19:58 They go through, Michael, they ask some of your questions, they work backwards from existing customer data.
0:20:00 What’s the thing everybody creates?
0:20:15 In my opinion, the ICP is a list of qualities and differentiators and firmographic and typographic information that then your marketing team, your sales team, your product team can all action on.
0:20:18 Should be probably less than a page in my opinion.
0:20:31 I think that’s right. It not only shows you who you need to target and what you need to target the customer with, but also why the customer would need your product specifically.
0:20:41 And who owns the pain within those organizations, as Michael said before, who are the champions, who are the influencers, who are the economic buyers.
0:20:48 And as the RevOps guy, I’m going to say that it is having the data and technology to support that, right?
0:21:03 There are a lot of very good companies that will work with people like Michael or have people like Michael and they put together this pristine, amazing, elegant ICP on slides and it’s great, right?
0:21:05 And that’s definitely part of it.
0:21:23 But how are you pulling information back into the process that the people in your marketing organization and your product organization and your sales enablement organization, et cetera, how are you getting all that information and then making that part of the way you’re defining your ICP?
0:21:28 There are a lot of companies that aren’t doing that very well and they’re still operating from theory.
0:21:39 And what they’ll find is they will eventually fall out of alignment with the market and they will see all these leading indicators that start to tell them that’s the issue.
0:21:46 And it’s because they’re still operating from a position of a little too much hubris and a little too much of an echo chamber.
0:21:48 I kind of want to throw out some examples.
0:21:53 I’m thinking of ICPs in the context of a company like OpenAI.
0:21:54 Do they have an ICP?
0:21:56 It’s ChatGPT all the way down.
0:21:57 It’s ChatGPT and you’ve got some APIs.
0:22:03 It’s the same product regardless if you’re a consumer or you’re working in the enterprise.
0:22:11 And I’m wondering if with this sort of new generation of products coming out, is the idea of an ICP still useful?
0:22:16 Make no mistake, those companies are also doing what we’re talking about here internally.
0:22:26 Obviously, if you are in the AI space right now and you are selling generative AI platforms and large language models and things like that, it’s a good time, right?
0:22:42 But you still need some way to be able to distribute and go after your market in a way that is prioritized like that in order to practically have all these great things like great conversion rates and expansion rates.
0:22:54 All these kind of lagging indicators that tell you you’re selling it and expand the market well, but are also leading indicators that are truly indicative of how well you’ve set up your ICP and operationalized it.
0:23:03 Yeah, I think if you have a product that has no competitors and is a brand new product and is an incredibly effective product.
0:23:06 And I mean, that was the early open AI days, right?
0:23:07 I think absolutely.
0:23:09 I mean, do you have to do it?
0:23:12 No, you’re kind of like, look, the product trumps all.
0:23:19 But the problem with that is that as soon as you have a successful product, there will absolutely be competition in the market.
0:23:32 And so you will have to eventually, even if you don’t do it at the very, very beginning, like you will have to segment the market, you will have to find your use cases, you will have to find the ICPs for those use cases, all of those pieces like that.
0:23:42 So you may see early success in a pure, like we’ve got the very best product out there and everybody’s just going to use this product, but that will not last.
0:23:45 It never has, at least in the 28 years that I’ve been doing this.
0:23:49 Like, as soon as you have a rockin’ product that’s selling well, guess what?
0:23:50 You have competition.
0:23:51 Totally.
0:23:53 The rubber meets the road eventually, right?
0:23:55 And you’ve got to figure out ways to sustainable growth.
0:23:58 No, I think that sustainable growth is the piece right there.
0:24:04 Like, you can get to one place, but in order to grow from that one place, you’ve got to double down on these best practices.
0:24:17 I think when you’re in the growth stages, maybe you’ve got the elusive product market fit for one kind of customer, but you need to, for platform companies, serve multiple personas with an org, or maybe you want to go a little more vertical.
0:24:24 And so how do you balance that question as you’re scaling, needing to find more customers?
0:24:25 Do you want to go more vertical?
0:24:27 Do you want to go more horizontal?
0:24:37 Eleven Labs is a great example of this, where they are selling an excellent platform, which can appeal to a lot of different potential buyer groups.
0:24:47 They’ve defined over a dozen different cuts of their ICP, but they are very clear on which ones they are most focused on right now.
0:24:48 And what that really means, right?
0:24:49 And what that really means, right?
0:25:01 How they are thinking about product innovation, building content, target accounts that they are assigning out to their sellers, where they’re putting salespeople geographically.
0:25:19 Even if they have ICPs that they know really well and they’re reporting on that data, they might have those, comparatively speaking, deprioritized in order to remain focused, but still have an approach to those companies that are a little bit more out on the fringe of their capabilities.
0:25:31 And the other point about this is, they’re also a very good example of a company that is clearly listening to customers on how mature their product is in certain areas.
0:25:33 And they’re not overselling past that.
0:25:56 They’re being really careful about making sure that the ICPs that they’ve targeted are where they know their product is going to be an absolute grand slam versus the areas where they plan to go to and they have some capabilities, but they know that they still need to develop it out a little further there before they really go hard at that particular area.
0:26:01 So they have a sort of stack ranked ICPs, it sounds like.
0:26:07 The group of companies that they know, this is a slam dunk, I can like really go long on solving this use case.
0:26:15 But then it sounds like maybe it’s opportunistic ICPs, places where they can expand and that they know, yeah, we can build into that.
0:26:20 And I would even add really good focus is more of an exercise in stratification.
0:26:28 It’s the way you’re segmenting the market and the best possible companies is this small circle.
0:26:37 That is where we’re going to be hyper focused because we know we get tremendous yield, expansion, just great things happen there in the center of that.
0:26:52 Then it’s these concentric circles out of that, where you have others that are still very good fits, but they are maybe not as quick to expand or a little bit more of a grind to actually convert those into customers, they expand a little more slowly.
0:26:56 And then you eventually fall out into lesser and lesser fit.
0:26:59 And it’s like proceed with caution on some of these over here, right?
0:27:00 Right.
0:27:05 So it’s a very prudent way to go about it, very mature way to go about it, where you know you’re going to go there.
0:27:10 But at the moment, you’re being careful because you want to be responsible in the market and the way you’re growing.
0:27:14 You want to be able to make sure your product is doing the things you say it’s going to do and that you’re meeting commitments.
0:27:22 Because if you do that really well, then those ICPs will be absolutely open and ready for you as your product evolves towards that.
0:27:25 Michael, I feel like this is entirely in your wheelhouse.
0:27:28 I feel like this is what you do day in and day out.
0:27:29 That’s exactly it.
0:27:35 I mean, I think no matter how general the tool is, you have to narrow your marketing and sales efforts.
0:27:47 And I think focusing down on those use cases and those ICPs gives you an ability to spend marketing dollars in the right place, spend sales efforts in the right places, supporting the right customers.
0:27:52 And sometimes at the same time, like not support, not market to, not sell to other customers.
0:27:58 It’s really hard when someone waves a handful of cash and says, I want to buy your product, but you’re not my ICP.
0:28:00 It’s really hard to say no to that.
0:28:05 But sometimes I think you need to say no to that because you can get pulled off track with customer requirements.
0:28:14 You can get pulled off track with a customer sales effort that’s going to lead you down what might be a cash rich place, but not necessarily a strategy rich place.
0:28:21 In other words, it won’t necessarily net you the next customer and the customer after that and the customer after that, which is way more important than the one customer you just landed.
0:28:27 I think we’ve talked a lot about use cases and focusing on the pain point or problem that you solve.
0:28:36 But what I want to figure out and we started talking about this when Michael was like, yeah, this is why you got to have a great RevOps program is let’s say you get this ICP.
0:28:42 How do you know whether it’s working and how often do you need to continue refining it?
0:28:46 ICP is definitely not a set and forget thing, right?
0:28:47 It does evolve over time.
0:28:53 It certainly evolves when you go from founder led sales into repeatable into scaling up.
0:28:58 And very often that happens when you’re moving market segment.
0:29:04 So many companies start out in SMB, they move to mid market, they then move to enterprise.
0:29:06 The ICP needs to change.
0:29:08 They expand geographically.
0:29:11 The ICP may be different in different markets.
0:29:14 Pricing and packaging changes happen.
0:29:19 Competitive pressures and external events in the market impact ICP.
0:29:21 So it does evolve over time.
0:29:27 For me, at least the critical thing is you’ve got to create a very tight feedback loop with the market.
0:29:32 So there needs to be a tight feedback loop between sales and product.
0:29:44 And I think it’s also super critical that founders, CEOs, executives at every level, customer success, product management, engineering, that they’re also out in front of customers.
0:29:53 That they’re listening to customers, they’re engaging with customers, they’re understanding how they’re using the product, where the gaps are, where they want to go next.
0:29:58 Because that’s super important as you sort of evolve the product roadmap over time.
0:30:00 And then everything else follows from that.
0:30:04 So therein lies the problem with ICPs in the real world.
0:30:09 There is a constant evolution out there in terms of buyer interests and the nature of the problems that they have.
0:30:11 Your competitors are constantly changing.
0:30:21 Everybody you compete with is trying to do the same thing you’re doing, which is get better and better at all of these things like their conversion rates, expansion rates, and all these other good signals that tell you’re doing a good job.
0:30:30 Joe and I went through this experience when we first joined Segment, for instance, where you go and you have this big project of setting the market.
0:30:36 three, six months later, that has started to lose precision, right?
0:30:39 It’s not practical to run these projects constantly, though.
0:30:45 And so the challenge has been, how do you do it where you have a practical feedback loop and practical revision of that?
0:30:54 You’re not continuously just thrashing your sales force with new pivots in the way that they’re supposed to be selling and talking to the things.
0:30:57 And hey, you know what, these were the accounts we assigned you in your territory.
0:31:01 We’re going to continuously change that every couple of weeks because of the ICP.
0:31:04 No one’s going to like that, right?
0:31:08 I do think there are indicators when you nail an ICP.
0:31:12 I mean, I think you start to see the cost of your CPLs go down.
0:31:15 I think you start to see your sales efficiency go up.
0:31:17 You start to see higher renewals.
0:31:22 You start to see those indications that you’ve got it right.
0:31:24 I agree with Mark, though, 100 percent on this.
0:31:25 You can’t flip flop it.
0:31:31 What I tell folks all the time is with messaging and with targeting, like the minute you’re getting sick of it, you’ve got to double down on it, right?
0:31:34 Oh, I was going to say, what’s the threshold? Yeah.
0:31:35 Yeah, yeah.
0:31:46 So when you’re getting sick of hearing your same messaging or you’re getting sick of being focused on the same customers, that means it’s time to double down on it because it takes the customers a lot longer to input the messaging.
0:31:49 It takes the customers a lot longer to understand the use cases.
0:31:52 It takes them a lot longer than it takes you because you’re living in it day and day.
0:31:57 Tell us to founders all the time, like you think about your product 24 out of the 24 hours of the day.
0:32:01 Your customer thinks about your product maybe 10, 15 minutes a quarter.
0:32:04 You just got to keep hammering it and hammering it.
0:32:09 I think what actually is really interesting to me, though, is how is this going to change now in the era of AI?
0:32:14 Fundamentally, the question of what is our ideal customer is a perpetual strategic one.
0:32:23 But AI, I think, actually offers huge promise to be able to continually evolve that and really micro segment ICPs.
0:32:36 So you think about ICP evolving from a static document into a living data powered model that lives in your CRM where you can do continuous refinement on it.
0:32:43 It helps you discover better ICPs, but you still need to make decisions on where you steer the business.
0:32:52 And then ultimately, founders and executive teams are going to have to make decisions on where they invest and then how much they invest.
0:33:05 The companies that I’m working with right now are plotting these AI based roadmaps, instrumenting the central nervous system with all of this signal detection that gets pulled back in to both product and marketing,
0:33:20 which then gets almost immediately disseminated into the materials that are being created, the sales enablement that’s going on, the way that you can go and present the next accounts to call on as a rep, who to talk to, what to talk to them about.
0:33:30 And the revolution that’s coming with this is a matter of being able to do this very, very quickly and very, very precisely, continuously.
0:33:36 And we’re going to be wondering, well, why didn’t everybody do this? And the truth is, it’s available to all of them, but it’s a mindset.
0:33:43 It is definitely the culture of your organization that’s going to determine if you’re able to take advantage of it with those new capabilities there.
0:33:54 If you take AI out of the picture again and go back to the companies that have been doing this, the best companies that have done this best for years, they are tremendously well aligned across marketing and product and sales and customer success.
0:34:06 It’s an amazing, just cohesive approach to how to think about the ICP, how to innovate your product in that direction, how to go and prospect and sell and expand your customers usage.
0:34:11 If you have that mindset and you’re willing to put in the work, you could be really good at this.
0:34:17 The piece that people have been doing already is the actual collecting of this information, right?
0:34:41 So whether it’s like gong calls or the precision in which you can measure your marketing spend and your marketing effectiveness and everything else like that, or the customer success platforms that are recording all of the information, all of those information sources gives you the raw materials for an AI to come in or to an AI products to come in and really understand what is successful throughout the entire customer journey.
0:34:48 I think we’ve done a great job as an industry, collecting all this information and unfortunately it’s lived in silos or individual people.
0:34:53 But I think we now have the opportunity with AI, what does AI do really, really well? Large pattern recognition.
0:34:58 So across the entire customer journey, understanding where your ICPs are landing and where they’re not.
0:34:59 That’s exciting to me.
0:35:01 As a marketer, I get really fired up about that.
0:35:15 I love that because the sheer amount of data collection feels perhaps unique to the moment, but the decision to relentlessly collect, pursue, analyze and operationalize the data has been around for ages.
0:35:24 All right, that is all for today. If you did make it this far, first of all, thank you.
0:35:32 We put a lot of thought into each of these episodes, whether it’s guests, the calendar Tetris, the cycles with our amazing editor, Tommy, until the music is just right.
0:35:41 So if you like what we’ve put together, consider dropping us a line at ratethispodcast.com/a16z and let us know what your favorite episode is.
0:35:44 It’ll make my day and I’m sure Tommy’s too.
0:35:46 We’ll catch you on the flip side.

0:00:02 – We have a software crisis.
0:00:04 – There’s just no time to wait.
0:00:05 If you don’t have that sense of urgency,
0:00:07 we’re not gonna accomplish what we need to do.
0:00:09 – You can build the most advanced equipment,
0:00:12 you can produce it at the largest scale possible,
0:00:14 but if you can’t get it where it’s needed,
0:00:16 when it’s needed, it doesn’t exist.
0:00:18 – There’s always been collective defense,
0:00:20 but not necessarily collective logistics.
0:00:22 Any one nation can’t do it alone.
0:00:24 – Whatever we’re modernizing now,
0:00:25 this is not gonna be the last update.
0:00:26 Software’s never complete.
0:00:29 Software is moving incredibly fast.
0:00:30 The best type of board you fight
0:00:33 is one that you don’t have to fight at all.
0:00:35 – The largest companies in the world today
0:00:37 are software companies.
0:00:39 And many of these companies have become household names
0:00:42 for developing fun software.
0:00:44 Software that powers the games we play,
0:00:45 the apps we scroll,
0:00:48 software that helps run our lives and manage our work.
0:00:52 But the world also needs serious software.
0:00:54 The kind of software that runs the autonomous vehicles
0:00:55 in San Francisco,
0:00:58 the software that ensures planes take off and land safely,
0:01:00 the software that ensures critical supplies,
0:01:03 make it to our shores and equally the front lines
0:01:06 of our military, even in contested environments.
0:01:09 The challenge, across the commercial sector
0:01:10 and the public sector,
0:01:14 much of this serious software is built on legacy technology.
0:01:17 And with the world moving at the speed of software,
0:01:20 our infrastructure gets more brittle with each passing year.
0:01:23 When it breaks, it causes inconveniences at best,
0:01:26 and tragedies at worst.
0:01:27 But here is the good news.
0:01:29 Software continues to eat the world,
0:01:32 and the brightest minds are increasingly interested
0:01:35 in solving these serious problems.
0:01:36 So in today’s episode,
0:01:38 recorded live at our American Dynamism Summit
0:01:41 in the heart of Washington, DC,
0:01:43 we sit down with Philip Buchendorf,
0:01:45 and recently retired Lieutenant General,
0:01:47 Leonard J. Kaczynski.
0:01:50 Philip is the co-founder and CEO of Airspace Intelligence,
0:01:54 a company working to address the software crisis across some
0:01:58 of the country’s most critical public and private sector institutions,
0:02:00 from air traffic to defense.
0:02:04 Lieutenant General Kaczynski is ASI’s Chief Strategy Officer
0:02:08 and the former Director of Logistics for the Joint Staff,
0:02:12 spending over three decades of leadership in air mobility and logistics,
0:02:16 seeing firsthand what consequences we face if logistics are overlooked.
0:02:17 It’s just like oxygen.
0:02:21 It’s fine up until you don’t have it and then it becomes a concern.
0:02:23 Together with our very own Leila Hay,
0:02:27 A16Z’s go-to-market partner focused on American dynamism,
0:02:31 the group explores the challenge of hardening America’s logistical network,
0:02:36 but also how the public and private sectors can join forces via dual-use software,
0:02:41 and the modernization and risk posture that we need from governing agencies.
0:02:44 So what’s at stake if we don’t get this right?
0:02:47 Listen in to find out.
0:02:51 As a reminder, the content here is for informational purposes only,
0:02:54 should not be taken as legal, business, tax, or investment advice,
0:02:56 or be used to evaluate any investment or security,
0:03:01 and is not directed at any investors or potential investors in any A16Z fund.
0:03:06 Please note that A16Z and its affiliates may also maintain investments in the companies discussed
0:03:07 in this podcast.
0:03:09 For more details, including a link to our investments,
0:03:12 please see A16Z.com/disclosures.
0:03:20 Philip, I’d love to start things off with your story.
0:03:24 On your ASI company page, and I quote, it says,
0:03:29 “12 years ago you were frustrated by Germany’s stagnant approach to national security and its
0:03:33 economic degrowth mindset.” Tell us about that moment in your life and why you came to America.
0:03:38 Grew up in Germany, spent some time in the UK, and then pretty much after college.
0:03:43 So this is about 2011, 2012. I learned to basically figure out like,
0:03:44 “Hey, what am I going to do in my twenties?”
0:03:48 I knew I want to be in a fast-paced, demanding environment.
0:03:53 But what I found in Europe was most of my friends just wanted to party.
0:03:59 Germany just decided to pull out a nuclear energy to become fully dependent on cheap Russian gas.
0:04:05 And then the top 5% in my friends and social network glorified consulting to go into consulting.
0:04:11 At the same time, the government made it as hard as possible to build anything from starting a company
0:04:18 to venture capital. At the same time, while all that kind of happened and while I observed that and
0:04:23 internalized it, I heard about Silicon Valley. And I was like, that sounds like an interesting place.
0:04:29 It sounds like everyone is just obsessed with building and technology. And so a friend of mine and I,
0:04:34 we basically traveled to Palo Alto, stayed in a hacker house for the first three years.
0:04:41 I slept on bunk beds and ate frozen food from Trader Joe’s, but it was the most exciting environment
0:04:45 to be in. Everyone was building in that hacker house. Everyone was thinking startups and technology.
0:04:51 And fast forward, I now probably spend nearly as much time or even more time in the US than I did in
0:04:58 Germany. Our second daughter was just born here in DC last week. In many ways, I lived the American dream.
0:05:03 Everything that I did, everything that I learned, none of this could have been built, I would say,
0:05:09 somewhere else. It was the ecosystem, the ethos, the energy, the people that enabled that. And I’m very,
0:05:16 very grateful, but also determined that that was very much the right decision on where to go.
0:05:21 I love that. You actually started your career in the world of autonomy and autonomous vehicles.
0:05:24 Can you talk about your career journey there and what brought you to ASI?
0:05:30 Yeah. I mean, when KD, Lucas and I started ASI, I would say we were kind of the ultimate outsiders.
0:05:36 We weren’t pilots. We were not in defense before. We were not logisticians. We worked on autonomous
0:05:43 driving. And so this is around 2017, 2018. Autonomous driving was very popular, very hot. And while many
0:05:48 liked that, we didn’t. It felt overly crowded. And we started this company with a simple question,
0:05:54 which is what are other modes of transportation that require better software. And the first six
0:05:58 months of ASI, we went to operation centers and wanted to understand what’s the state of software,
0:06:03 what’s the state of technology, no matter if that is in maritime or in air operation centers.
0:06:08 And I would say to some extent we expected or hope to see science fiction. What we saw was like the most
0:06:15 ancient software possible. And that very much made it very, very obvious to start this company. It was a
0:06:22 mission to enable the kind of world’s most critical operations and optimize the most valuable assets and
0:06:25 infrastructure that we have as a country. And that’s how we got started.
0:06:32 Fascinating. Leo, I’d love to hear about you. I know that recently in the Joint Chiefs serving as
0:06:36 the Director of Logistics, Three Star General, you just retired. What brought you to ASI?
0:06:43 Yeah. So I spent over 30 years of my career in the military, starting with air mobility operations,
0:06:47 and then the last seven years, really logistics. So after I retired, I got the advice to kind of give
0:06:51 yourself about six months to figure out what’s next after doing something for that long. Although I knew
0:06:55 what I was passionate about, national security, logistics optimization type things. And I had a
0:07:00 chance over that six months to meet different companies, see different technologies, and see
0:07:06 a lot of things going on there. But when I met Philip and the team at ASI, something unique and something,
0:07:12 I guess, compelling to me. One was, while distribution platforms and planes and ships are all quite
0:07:17 important, it really comes down to the data. To be able to access that data, to be able to optimize and
0:07:20 figure out what we need to do. That was something I’d say struggled with in the Department of Defense,
0:07:25 to be able to just access and then to be able to understand and be able to really optimize the
0:07:30 things we need to do now in the future. So the combination of that piece being an exciting startup,
0:07:35 being exciting, not just support the military, but also the whole commercial sector, which is
0:07:40 hugely important for both. Absolutely. And can you tell us a little bit more about your career prior
0:07:45 to ASI? Sure. I started off actually was initially going to be an engineer. I went to graduate school
0:07:50 for industrial engineering, optimization type things, neural networks back 30 plus years ago,
0:07:56 before we really had processing power to do that. Then quickly went into pilot training and then flew
0:08:02 mobility planes for many years. But last really seven years or eight years ago, my first real foray into
0:08:07 like really broader logistics, I was a director of logistics for US Africa Command out of Stuttgart,
0:08:12 Germany. And that was also during the pandemic. But just working with Department of Defense logistics
0:08:16 and infrastructure, trying to move around Africa, just the size and scope. And then you add on the
0:08:21 pandemic and then went to Japan for command assignment, then came back to the joint staff as a director for
0:08:28 logistics with Ukraine going on, support for Israel, everything else, and just realizing challenges with
0:08:33 our defense industrial-based challenges with what we have. And really one of the main efforts there was really that data
0:08:39 software piece, which we just weren’t very good at trying hard, but had a lot of catching up to do in the
0:08:45 Department of Defense. Makes a lot of sense. So you’ve been all over the world, and we’re just seeing these common core
0:08:53 challenges across everywhere you were. A lot of it is around data and software. Well, I’d love to kick off and first talk about the air
0:08:59 domain. So obviously, ASI got its start in the aviation space. It’s an area where we’ve seen a lot of
0:09:05 challenges. On the one hand, we as consumers are told that flying is the safest mode of transport, and
0:09:11 statistically that’s correct. On the other hand, we’re seeing headlines every day. We’re seeing news about air
0:09:18 traffic control shortages, staffing challenges. Can you just help us paint a picture of the aviation industry and what this
0:09:22 data play is? First of all, it is definitely by far still the safest domain, right? And mode of
0:09:29 transportation. But we’re fundamentally looking at three different problems. First one is staffing. So
0:09:34 there’s a significant staffing shortage right now in the industry. Why is that? You’ve seen a lot of
0:09:41 retirements throughout the COVID pandemic. Training was not happening at the same speed during the COVID
0:09:47 pandemic. I would say in general, more broadly beyond air traffic control, the industry might have also lost
0:09:52 the ability to attract the very, very best talent. And we can talk more about what that also meant for
0:09:58 software. So we have a staffing shortage on one hand, then we equally have a software crisis. We got
0:10:04 legacy software that is filtering, that is falling apart. Whenever it happens, you have these massive
0:10:11 outages that are incredibly consequential to the entire industry. And then you got very outdated
0:10:17 infrastructure. I would say what is not talked about enough is how they are actually all interconnected.
0:10:23 Let me give you an example. Staffing on software go hand in hand. If you have better software that is
0:10:29 much more intuitive, you can train people much faster. Even more, if you have software that is
0:10:35 supporting the operator, that operator is a lot more productive. You’re minimizing workload. If you
0:10:41 don’t have to do 10,000 clicks, but just a few, or you have AI that is assisting you in your decision-making,
0:10:46 it can be much, much more productive. And then, even more, that is leading to second-order consequences.
0:10:49 If you’re increasing productivity, that means you can actually pay people more,
0:10:55 right? And that means more people want to operate or work in the field. And so I think what is not
0:11:02 considered enough is how staffing and infrastructure are fundamentally actually software problems. You can’t
0:11:08 separate these three areas and look at them in a kind of an isolated way. Software is eating the world,
0:11:14 and that is very much true for this domain. And I think internalizing that, and as there’s a mandate
0:11:20 to modernize now, and pull this sector and this industry forward, like looking at it through the
0:11:23 software lens is absolutely critical. And I think some of that is happening.
0:11:28 And it sounds like there are all of these challenges, but they’re sort of being looked at in silos. It’s like
0:11:32 people are trying to say there’s a people problem here, there’s an infrastructure problem here,
0:11:37 there’s a tech problem here, but actually these things are all connected. And if we can modernize
0:11:39 them, that’s how we’re going to be able to move faster.
0:11:44 Absolutely. And you’re training a new generation, right? If you’re training a 25-year-old
0:11:50 air traffic controller, like that individual grew up with iPad, Snapchat, etc., right? Google is using
0:11:55 Google Maps when they drive a car, right? The generation that has retired or is about to retire,
0:11:59 they grew up with IBM green screens. Like they are familiar on how to use that technology. They used
0:12:05 that technology for the last 30 years. But the new generation that is being trained now,
0:12:10 they’re not familiar with those legacy tools. And I think it’s absolutely essential that software
0:12:16 is being modernized and is being brought up to speed what people are used to use from a software
0:12:17 quality perspective.
0:12:22 And actually, I’d love to double-click on that. We’re here at the American Dynamism Summit in
0:12:28 Washington, D.C. We all are grappling with the fact that there was a tragic accident here in our
0:12:32 city just a couple months ago. A commercial airline collided with a military copter. Philip,
0:12:38 I believe you were at Reagan National Airport at the moment that that happened. People assume that good
0:12:42 safety records equate with good technology, but it sounds like we’re actually dealing with a lot of legacy
0:12:45 technology. Can you help us understand what that looks like today?
0:12:50 The way I would frame it is it’s very much a philosophy problem. The philosophy on
0:12:55 what software should look like might be a bit broken. And I think it boils down to three issues.
0:13:02 The first one is software and compute are very much connected. So what does that mean? If you have
0:13:08 software systems that are deeply coupled with the compute power, it’s very hard to modernize anything.
0:13:14 Specifically, if you’re dealing with an industry where you have facilities all over the country.
0:13:20 So you cannot just launch an update over the year and all the software is updated. That doesn’t work.
0:13:26 No, you would literally have to go from facility to update the software or patch things if that is needed.
0:13:30 And then every single time you’re running at risk, it’s like, okay, can compute actually handle that
0:13:35 new software update? So the first part that really has to happen is when we think about modernization,
0:13:40 it’s like, how do we separate software and compute? It’s essential. Because again, whatever we’re
0:13:44 modernizing now, this is not going to be the last update. Software is never complete. Software is
0:13:48 moving incredibly fast. So the separation of software and compute, absolutely essential.
0:13:55 The second part is, historically, software in this domain has been built as if it would be hardware,
0:13:59 while the entire world is moving towards a direction where even the hardware companies build
0:14:03 hardware as if it would be software. So it’s in many ways, the inverse, all right? So whenever there’s
0:14:07 a modernization effort and no matter if this is with the government or on the commercial side and
0:14:13 air operations, there’s usually a need, a program is started, it’s being funded. The first thing that
0:14:17 happens is like a thousand page documentation is written, already tens of millions of dollars are
0:14:22 spent on just writing the documentation, no software built yet, nothing shipped, nothing works yet.
0:14:27 It’s just a documentation, right? So now we’re already tens of millions of dollars into just documentation.
0:14:32 And then over the next 10 years, for hundreds of millions of dollars, like software is being written
0:14:37 from scratch for that particular problem area or against those requirements. And then like 10 years
0:14:43 later, all that is magically considered working and finished and it’s being rolled out. Obviously,
0:14:47 by that time, it’s already out of date and antiquated because the state of technology has changed a lot
0:14:51 over those 10 years. And then it’s somewhat maintained for the next 20 years, right? But I mean,
0:14:56 we all know this is not how software is built given how fast it’s changing, how fast it’s moving.
0:15:03 And then the third one is that the companies that historically participated in the space can no
0:15:09 longer attract the very best software engineers. Like the very best software engineers do not want to
0:15:16 work in that ecosystem and with those structures. They want to build rapidly, right? They want to build
0:15:21 close to the user. They don’t want to be handed a list of like 10,000 requirements and then just write
0:15:25 code against that. But the combination of what all these things meant is like, you are basically
0:15:31 in the setup that does not produce a software that should be produced. And then if you wrap all of that
0:15:36 into an acquisition framework that is incentivizing those philosophies, you’ve got a really big problem
0:15:40 at hand. And that is where we are right now. But I think things are about to change.
0:15:43 We may talk about this later, but when I just think about the Department of Defense,
0:15:48 the way you describe that ecosystem and our challenge is exactly the same, especially in the logistics business,
0:15:52 within each service, like the Army and Marine Corps, Air Force. Within each subunit,
0:15:58 as you go down, legacy systems are programmed by great people decades ago that weren’t designed to be
0:16:04 able to connect and to do that. And we spend so much time trying to ad hoc systems to put together where
0:16:08 we really should be thinking forward to have that clean sheet type of software that can be iterative.
0:16:13 There’s just no time to wait. I mean, you have to move fast. And even moving fast, it still takes
0:16:16 some time. But if you don’t have that sense of urgency, we’re not going to accomplish what we need to do.
0:16:23 On that point around urgency, we have to fix the status quo. It’s exciting that this administration
0:16:29 seems very committed to moving fast. Would love your perspective on how fast we can solve some of these
0:16:30 big challenges.
0:16:35 Yeah, I think things will only change if the momentum kind of stays what it is right now,
0:16:40 if there’s a real urgency for change. I think President Trump, Secretary Duffy set the direction
0:16:47 and the mandate. I think the next step is for Congress to fund the modernization efforts at the
0:16:52 FAA. I would argue that in many ways, it’s one of those areas where there’s strong bipartisan support for
0:16:57 this. It’s very hard to argue like why not to modernize air traffic control systems and why the U.S. should not
0:17:03 have the very best software in that field. I think everyone agrees that it’s the very best what we should
0:17:09 have. Right. But then I would say at the same time, it’s important that the guidelines are put in place
0:17:17 on how to spend that money and how to not repeat the same mistakes from the past. We don’t need to spend
0:17:23 10 years on custom development if we can actually purchase software that already works in the private
0:17:28 sector, that is already commercially deployed, that we can literally just purchase and use as is. Maybe
0:17:33 make a few modifications, but it’s already stuff that is available. That is a lot more efficient,
0:17:38 that allows for much faster modernization, and it’s also the safest because it’s already proven.
0:17:43 So I think the next two steps here is like making sure the resources are there. I think it’s fair to
0:17:47 say the FAA has historically not had, or most recently has not had the resources that they
0:17:52 need to modernize, but then making sure the structures are done the right way so we’re not
0:17:58 repeating the same mistakes over again. Absolutely. And when you say proven software,
0:18:04 we’d love to understand what exactly you mean by that. How has ASI worked with the commercial sector
0:18:09 and the DOD already to help provide some of these capabilities? Absolutely. I mean, to give you some
0:18:15 examples, right? So when it comes to some of the modernization efforts that the FAA will pursue around,
0:18:19 for example, air traffic management, like a lot of that software that is needed and a lot of the
0:18:24 capabilities, we already have commercially deployed with the airlines, right? In many ways, a lot of the
0:18:30 airlines are advocating they would love the FAA to use this type of software. So instead of building
0:18:35 that capability from scratch for hundreds of millions of dollars, and that has historically been the idea,
0:18:41 right? Why not use something that already is deployed, that already is used by some of the
0:18:45 largest airlines in the country? Like, why not use that? Because it already works. And the same has
0:18:49 been true when we started working with the U.S. Air Force. The reason why we were able to deploy within
0:18:54 months and have seen our software being used in live operations was simply because it already worked in
0:19:01 the commercial sector. It was already deployed there in kind of 24/7, 365 days a year type of fashion. I think
0:19:06 there are certain areas where dual use is a good idea and there are certain areas where dual use is
0:19:11 not a good idea. But when it comes to some of these industries where the private sector and the public
0:19:17 sector have to collaborate very closely together, dual use is a phenomenally good idea. Not just because
0:19:22 it’s more efficient, but it also enables more collaboration. And when it comes to the national
0:19:27 airspace system, it’s a system that is managed by the government, but it’s used by the private sector,
0:19:28 meaning Bay Airlines.
0:19:32 Yeah, at the risk of really bringing a Silicon Valley term to the table, like there’s a network
0:19:33 effect here, right?
0:19:34 A hundred percent.
0:19:39 You want everybody singing to the same tune. And if you have a platform where everyone has access to
0:19:41 the same data, then everything can be more efficient.
0:19:42 Yep.
0:19:46 Just adding from my experience in the Department of Defense aspect, and maybe people don’t realize,
0:19:53 but day-to-day that U.S. military uses commercial transport, whether it’s trucking, rail, air,
0:19:59 sea, and then any kind of contingent or disaster type of escalation, we would have to actually use
0:20:04 more. So it’s really the same resources, the same need for that collaboration and when we need to be
0:20:07 able to really work well together on the same system and platforms.
0:20:13 Absolutely. And that brings me to the next thing I wanted to double click on with how ASI has been
0:20:20 expanding into areas like logistics. Can you maybe just walk us through what exactly that looks like
0:20:25 with the DOD? It feels like a black box for many folks. And when we say contested logistics,
0:20:26 what does that exactly mean?
0:20:30 So contested logistics, that’s actually a military term. Every military term, we have to like
0:20:37 definitions and doctrine and even secret stuff that we discuss. But just to understand the concept of
0:20:41 it. And most of us order things online. I’m sure you may have done that within the last week. So when
0:20:45 you order it, you worry about the price, maybe when it might get there. But most people have no concern
0:20:49 about where it’s built or the supply chain when it gets there, unless it’s going to be delayed or you
0:20:54 realize that there’s some weather system in that. Any given day, you have contested logistics. It’s just
0:21:00 that the consumer and actually even the CEOs of some companies or even senior four-star generals may
0:21:06 not think about logistics day to day because it hasn’t been a problem in the past. The idea of contested
0:21:11 logistics is that whether it’s weather, whether it’s maintenance, whether it’s other situations, you’re
0:21:15 going to have challenges. And how do you understand that, predict, and optimize? And that’s what ASI, that type
0:21:20 software. But on the other side, from the military perspective, is that the contested logistics from
0:21:26 adversaries will look for your vulnerabilities, which would be in our supply chain and logistics
0:21:32 pieces. So that’s where you really to be able to outdo them and to maintain to be able to do what you need to do is to be able
0:21:38 to understand where your vulnerabilities might be, have that resiliency in there and predictive to go around.
0:21:44 Another one, we, in the military, move an aircraft carrier or a group of soldiers somewhere.
0:21:50 You make that decision, but people don’t often think about the whole supply chain, the tail, I guess, of that you’d call it. The food supply,
0:21:56 munitions, everything else to get them there and get them back. But that’s hugely important, because no matter what great weapon system you have,
0:22:02 if you can’t supply and sustain it and move it where you need to, it’s not effective. And I really think logistics and
0:22:07 experts do think of it as a weapon system itself. I mean, it’s your competitive advantage if you can leverage
0:22:13 it. If you can’t and you don’t see it, it becomes your greatest weakness and vulnerability. So we want
0:22:16 it to be your competitive advantage by providing this capability.
0:22:21 It’s almost like electricity or water. You don’t think about it working or not working until you flip
0:22:25 the switch and your light doesn’t come on or you turn the faucet and the water doesn’t run. So it’s
0:22:26 really the backbone of everything.
0:22:32 We’ve seen that from, I think, day to day, the pandemic when face masks or toilet paper, things like
0:22:37 that. But for the military, I saw that with support to Ukraine, where there’s just munitions and things
0:22:43 we are moving and just the resources to do that, the supply chain to be able to replenish those. And as you
0:22:47 start thinking through and potential crises as things escalate, you’d want to be able to
0:22:52 predict that and to understand that better. It’s just like oxygen. It’s fine up until you don’t have
0:22:53 it and then it becomes a concern.
0:22:55 Yeah. Philip, what’s your perspective?
0:22:59 To echo very much what Leo shared, we talked about
0:23:03 air traffic control and air operations before. Same there, right? You just assume things are working
0:23:08 until they don’t, right? That’s why I think sometimes those sectors are a little harder to
0:23:13 gather everyone’s interest for it or to make sure the funding is there, right? It’s very easy to
0:23:18 spend money on the fancy weapon system, the autonomous drone, the new kind of high tech
0:23:23 equipment, whatever it is, because it’s like physical, it’s like visible, etc. But when it comes to the,
0:23:27 I would say, quote unquote, the silent software that runs in the background that enables all of this to
0:23:33 work, that allows the most advanced equipment to go where it’s needed, when it’s needed. I think that’s
0:23:38 sometimes too much of an afterthought. And I think in many ways you could argue you can build the most
0:23:44 advanced equipment, you can produce it at the largest scale possible, but if you can’t get it where it’s
0:23:50 needed, when it’s needed, it doesn’t exist. Why has that been such an underserved topic? Because to
0:23:55 your exact point, when we think about defense and some of these new capabilities, we’re talking about
0:24:00 autonomous drones and counter UIS capabilities and electronic warfare. We haven’t been talking nearly
0:24:06 enough about logistics, but we have just been through the pandemic. Like we’ve seen the disruption
0:24:08 in our personal lives. Why do we still have this disconnect?
0:24:13 I think human nature wanting to go back to the status quo. I mean, you have all this stress and
0:24:16 you just want to take a deep breath and go back to what seems comfortable. But we’ve been able to be
0:24:23 comfortable for the last few decades because we haven’t had a global war, fortunately. And so it takes
0:24:27 pressure points like the pandemic to really see, here’s where our vulnerabilities. And that’s where
0:24:32 there was a lot of investment going, taking place. But then that human nature of just easing back,
0:24:37 same with air traffic control. You know, great people working on that, but they deserve a much better
0:24:42 system, software system, at least the 21st century to do that. And you hope it doesn’t have to come to
0:24:47 something where there’s almost a crisis or something. And the other is we’ve been over the last several
0:24:52 decades just accustomed to just in time. Logistics has just been a cost area. You’re worried
0:24:57 about cost, reducing that. It’s all well and good if there’s nothing out there threatening that supply
0:25:01 chain. But you realize that just in time isn’t in time at all if you can’t get that part or that
0:25:06 supply in need. Companies that I talk to and others are realizing this, that really their competitive
0:25:11 advantage can be in that resilient and that understanding that supply chain and logistics piece.
0:25:15 But instead of thinking of the cost, it should be something that’s their competitive advantage,
0:25:21 something that provides profit or in this for the military, more deterrence and capability.
0:25:22 We’ve been talking a lot about this.
0:25:27 I love it. Well, I want to go back to the dual use topic. That’s an area where we spend a lot of time.
0:25:33 We think about companies that are, have an existing commercial capability that is doing great work in
0:25:39 the commercial sector. They can take that exact same capability to the DOD. Walk us through how ASI fits
0:25:43 into that puzzle and some of the areas where you all are going to be able to lean in.
0:25:51 I think there are certain domains, certain capabilities where building something specific for one sector is
0:25:56 absolutely the right way to go, right? A new missile or an aircraft carrier, there’s not much commercial
0:26:03 applicability for that, right? Logistics is one of those domains where I would argue that it’s like the
0:26:09 flagship example where you want to have dual use. And why is that? Leo to some extent already alluded to it.
0:26:16 From a defense perspective, a lot of the capacity resides in the private sector on the commercial
0:26:21 side. And that goes way beyond the civil reserve fleet, right? It’s the same infrastructure that
0:26:27 are used, the same ports, right? And guess what? Our adversaries are actually trying to deploy their
0:26:34 software into allied ports. Like the Chinese are really good in making sure their software runs in
0:26:37 ports. They’re giving it away for free. And there’s a reason why they do that, right?
0:26:44 And then from the other side, let’s say the private sector, like if you’re providing mission-critical
0:26:49 infrastructure, like transportation, you want to make sure using military-grade software because
0:26:55 the stakes are just so high, right? So you want these two sectors to be very close. Yes,
0:26:59 there needs to be separation, but there’s nothing better than actually running very similar to the
0:27:05 the same software stack on both sides so that these two sectors can communicate, can collaborate,
0:27:12 can share data when it’s needed. And that it doesn’t feel in the moment of crisis, oh,
0:27:17 shoot, now we need to understand like what does a software stack, what does the data structures look
0:27:22 like on the other side so we can actually coordinate. You don’t want to figure that out when you’re in the
0:27:28 moment of a crisis. You want to have that done before, right? And that’s very much, I think,
0:27:32 what we’re trying to do at ASI. We want to make sure that the very best logistics software is the very
0:27:38 best software to operate mission-critical operations is deployed with the companies that are doing that
0:27:44 in the private sector as well as with the government, and then enable these two sectors to collaborate.
0:27:49 So in many ways, it’s like I’m always saying it’s like the truest form of dual use because we’re not
0:27:54 only sharing technology, but we’re also enabling collaboration and communication between these two
0:27:59 sectors. And I think then to some extent you can make the argument of you want to use similar ideas
0:28:04 or a similar mental model when it comes to the collaboration between the US and its allies,
0:28:10 specifically on the side of military logistics, Contessa logistics. It is always about the
0:28:15 integration with our allied partners. How can we tap their infrastructure? How can we use some of the
0:28:20 capacity that they have? How can that all be coordinated? And there are times where you want to have more
0:28:25 separation, and then there are times where you want to have tighter collaboration and more sharing,
0:28:27 and you need to have the infrastructure in place that allows for that.
0:28:32 So in my previous job, I was a US rep for the NATO Logistics Committee. Over the last few years in
0:28:37 NATO, there’s always been collective defense, but not necessarily collective logistics. Logistics was on
0:28:43 the nation to do things, but the realization is that any one nation can’t do it alone. So this idea of
0:28:48 collective logistics across NATO, which the US is a member of NATO, is all common sense, but they finally
0:28:53 put it into not only an understanding, common understanding, but planning and other things to
0:28:58 how best to utilize our collective logistics capabilities and plan for that in the future.
0:29:05 If we talk about modernizing logistics for the DOD, what exactly needs to change? Is this a policy issue?
0:29:10 Is it a culture issue? Is it an experimentation issue? What do we need to fix here?
0:29:16 I think it’s really thinking very differently about the software. And when I say software,
0:29:20 in this case, I don’t mean like just the legacy software, but I think also what we deployed in
0:29:26 recent years. I think the last five, 10 years, a lot of the modernization that happened was just like
0:29:30 putting up new dashboards that are running in these operation centers on like bigger TVs,
0:29:35 but we haven’t really deployed software to the warfighter, to the operator. David Yulevich is
0:29:40 putting it very well when he’s saying the world is getting a lot more spicy. It’s much spicier now.
0:29:46 And what that means is we actually need different software. When the world is stable, you can operate
0:29:50 off like near real-time displays. That means the human operator is seeing problems as they happen,
0:29:56 and then they react to it. But that’s not necessarily the world we’re living in. The world is a lot more
0:30:01 uncertain now. And I think that means we need to have software that is showing the operator what is
0:30:06 about to happen in the operating domain. How do they need to adjust? I would say anticipation in
0:30:11 many ways is a new high ground. When it comes to software, we’ve seen three evolution steps. The
0:30:17 first evolution was we have compute. This is in the 1970s, 1980s. We had workstations. They were not
0:30:23 connected. You input some data, and there’s some optimization process or some form of processing
0:30:27 that is happening, and then you have an output. The next evolution was when all those workstations
0:30:33 became connected, right? The internet. So now a lot more data became online, and the next step from
0:30:37 there was like, oh great, now we can extend that to the internet of syncs. A lot more sensors became
0:30:41 online. So the big challenge was like, how do we fuse all that data, right? How do we make sure
0:30:46 we have a great common operating picture? I think that was very much kind of the focus over the last
0:30:53 15 years or so. So data fusion and like displaying that data and making it accessible very often to these
0:30:59 critical industries and the military. I think now we’re at the very beginning of a new revolution,
0:31:05 which is prediction machines, right? How do we actually build interfaces that are predicting what
0:31:09 is about to happen? What is the state of the operating domain? What is the state of the supply chain? What
0:31:15 is the state of the assets we’re operating, not just right now, but over the next hours, over the next
0:31:21 days, over the next weeks? And how can we forward simulate that? And that is an enormous advantage for
0:31:26 an organization if you have that capability. And now we need to make sure we’re rolling that out,
0:31:31 and we’re embedding it in our operations. From an ASI perspective, we pioneered some of that work,
0:31:36 very specifically in the air domain. But we need to do that more broadly now, across all domains,
0:31:43 across both sectors, private and public. And then I think the second part is like we need to enable
0:31:47 much tighter collaboration between the two sectors. We talked about that already a little bit, but the
0:31:56 key issue is going to be how do we get more capacity ASAP, right? And yes, we can think about how do we
0:32:00 build more ships and all of that, and there are clear needs for that. But again, all those things take
0:32:05 time. But at the same time, there is already a lot of capacity within the Western Hemisphere, between
0:32:11 the US and its partners and allies, as well as between the private and the public sector. Like how do we
0:32:17 enable collaboration communication to make sure that capacity can be used, and can be used effectively or
0:32:24 efficiently. And then lastly, I think it’s important that these sectors are not just an afterthought. How do we
0:32:30 make sure that these sectors get the funding and the care and the attention, not just when
0:32:35 things fall apart? Because that means it’s too late, but how do we make sure we invest in these
0:32:41 sectors proactively before things fall apart? I think the last point about just the funding
0:32:44 and prioritization, I think from policy, that logistics is just fundamental to everything we
0:32:49 do and to fund it and prioritize it. The other is, and this came from your air traffic controller
0:32:55 explanation about just the skill sets that it took in the past to get proficient maybe with older
0:33:00 systems. And so I had a big challenge in the joint staff as a director for logistics in creating
0:33:06 joint logisticians. So in the military term, joint means you’ve got the Army, Air Force, Navy, Marines,
0:33:10 and each of those have perspectives. If Air Force wants to move something, they think about doing it by
0:33:15 air, Navy, by sea, for example, or maybe Army by ground. First, you have to be an expert in that kind
0:33:20 of logistics, the air logistics. That takes several years and not a decade. And then opportunities to
0:33:24 become a joint logistics expert, which means you have to understand all of those. That takes a lot of time,
0:33:29 and we’re challenged to do that. Not that we shouldn’t strive to do that, but when you have AI
0:33:36 decision-making tools that can enable individuals to make decisions, just to facilitate, instead of
0:33:40 taking 20 years to train someone to do this, you have software that you can be trained and still learn,
0:33:45 but can give you the option of multimodal, just send it by ship or air or what the best decision is.
0:33:50 And I think that’s really an accelerator from what we need to do, because not that we still
0:33:56 shouldn’t train and achieve that, but in today’s technology, we should be leveraging that technology
0:34:01 as opposed to struggling to try to provide this one person that can do everything. And if that person’s
0:34:07 not there, then you can’t succeed. Exactly. If I’m a developer, I don’t need to go out and buy
0:34:11 a bunch of servers and rack and stack them in my garage. I can swipe my credit card with one of the
0:34:16 cloud providers and focus on higher level efforts, actually build the software. So it sounds like
0:34:21 there’s a similar opportunity here for acceleration. Philip, I’d love to ask you one more question,
0:34:26 because you started to touch on the intersection of AI and logistics. If we’re doing all this right,
0:34:31 and we get in a time machine 10 years from now, what are some of the problems that we’re going to be able
0:34:39 to solve with getting that right? One is we will be able to do a lot more with the capacity that we
0:34:48 have available. So that’s one. I think two is we will be able to harden our logistic networks in an
0:34:54 uncertain world or uncertain state of the world. I think no matter how some of these crises are going
0:34:59 to pan out, I think the probability is very high that the next few decades are going to be a bit more
0:35:06 dynamic and uncertain than the last two decades were. And that means that every form of supply chain,
0:35:15 any piece of mission critical infrastructure will be in one form or shape be disrupted. And how do we
0:35:22 have software that allows us to very quickly reroute stuff so that the impact of that uncertainty doesn’t
0:35:30 impact the warfighter doesn’t impact us, the civilian infrastructure, because we have the software
0:35:36 systems in place that are able to kind of balance that out and reroute things. And again, I think the
0:35:41 uncertainty we have a obviously from geopolitical tensions, but equally from sanctions, also from a
0:35:47 climate perspective, like increasingly more volatile weather has a huge implication, for example, to the
0:35:53 national airspace system and travel. And so how can we have logistic systems that are able to anticipate
0:35:58 these challenges and then balance things out when needed because you have that predictive capability?
0:36:03 No one I know wants to fight a war, but we want to be able to prevent that. Philip already alluded to, no matter what
0:36:09 ships and aircraft and high tech weapons you have, if you can’t sustain them, you can move them where they need to be,
0:36:14 the adversary knows that. Or even if you can move them where they need to be, but you can’t keep them sustained for any given
0:36:20 amount of time, that doesn’t provide that deterrence. And that’s something fundamental that we’ve, I think, maybe not had to
0:36:26 think about as a nation for many decades, but having that ability to do that, to provide that deterrence
0:36:33 based on understanding and really fully taking advantage of the logistics capability allows us to be
0:36:37 stronger. And hopefully, fast forward 10 years from now, I can’t tell you everything that will happen that
0:36:42 time, but we will still be able to deter and make sure that it’s a free and open world because of that.
0:36:47 If we don’t have that capability, it makes anything we do from the national security perspective that much
0:36:51 harder. And I don’t want it to be that much harder for our military men and women out there.
0:36:56 What are the risks if we don’t get this right? If we don’t modernize, what’s at stake?
0:37:00 There should be a sense of urging because there’s no time. There’s a book out there called The 100-Year
0:37:03 Marathon by Michael Pillsbury. And some of the background and the premise of the book is that
0:37:09 adversaries like China have been looking at our vulnerabilities and supply chains for quite some
0:37:14 time. And in a marathon, even if Philip’s a much faster runner, but if I start running today,
0:37:19 he doesn’t start to tomorrow, I probably will win. And this is where the sense of urgency to catch up
0:37:26 with our risks. Our risks lie in our ports, lie within our supply chain. You’ll see news releases
0:37:32 about hacking into water supply systems in places in Texas. And you’re under, why is that all happening?
0:37:37 If you look at Sun Tzu ancient strategy, the best type of war you fight is one that you don’t have to
0:37:42 fight at all. We just, as a nation, we have been able to power project and do things from everywhere
0:37:46 around the world. And it’s been great for the United States to do that. And we have an amazing capability,
0:37:52 but in this age of contested logistics with hypersonic missiles and cyber and space threats,
0:37:57 our ability, not just to operate abroad, but just to be able to leave our own ports and to be able to
0:38:03 move rail and everything uncontested, no longer exists. It probably hasn’t existed for a few years
0:38:09 or more. And so that’s really where the risk is. So that’s why the risk of not really taking action,
0:38:14 not drastically updating your system to the way that cutting industry operates at the Department of
0:38:20 Defense with logistics has built a system 40 years ago and done minor updates to it. That’s not the way
0:38:24 we need to operate in the future. And also from a geography perspective, I mean, this stuff is so
0:38:29 important, right? I mean, in many ways, the greatest asset for the U.S. is you’ve got a massive ocean
0:38:34 to the west, you’ve got a massive ocean to the east, right? It’s actually very, very hard for any
0:38:41 adversary to attack the U.S. on its homeland, given you have the oceans. But at the same time,
0:38:46 from a global power projection, we need to be able to overcome these vast distances over the oceans.
0:38:52 And that requires logistics, right? And our adversaries know that. And like their strategies
0:38:56 are very much like, okay, how do we cripple key logistic infrastructure, no matter if it’s
0:39:05 in the homeland or with allies and partners, to make sure the U.S. is limited in its capability
0:39:10 to make sure the equipment can go where it’s needed, when it’s needed, right? And that requires software
0:39:16 as much as a physical infrastructure to allow that we maintain the ability to project power globally.
0:39:20 Just some tangible examples. If you look back to the colonial pipeline,
0:39:25 for those who are in the east coast, couldn’t get gas for quite some time. You look at the Suez Canal
0:39:30 back in 2021, you had a ship that was trying to parallel the park. It stuck there for six days,
0:39:36 just the billions almost of dollars of trade that were affected by that. Those are things that had
0:39:42 different reasons to do it, no nefarious actions in particular. But you could just imagine those
0:39:46 vulnerabilities that we have. And you just can’t stop if we’re in a crisis. You’re going to have to
0:39:51 be able to overcome that. And that’s why this predictive logistics capability, AI enabled,
0:39:56 is a way that we would have to be able to look and think to come up with solutions. I think that’s so key.
0:40:01 And then how do we understand the threat profile for every single node in the system, right? How do we
0:40:06 understand the threats that might impact key infrastructure? We see a lot how underwater sea
0:40:13 cables are being attacked, right? Like every key node in the system is a vulnerability. And like,
0:40:19 how do we make sure we have the technology in place to detect any threats, no matter in what form or
0:40:24 shape they are coming? And then how do we quickly counter that by relying on other nodes more than
0:40:31 maybe particular nodes that are impacted in their capacity and efficiency. Now, if you made it this
0:40:36 far, a reminder that this was recorded live at our third annual American Dynamism Summit in the heart of
0:40:41 Washington, D.C. And if you’d like to see more exclusive content from the summit, head on over
0:40:54 to a16z.com/american-dynamism-summit, or you can click the link in our description.

0:00:02 Is there a better way? I think there is.
0:00:05 Every single interface that I interact with,
0:00:08 every single problem space that I’m trying to solve
0:00:12 are going to be made easier by virtue of this new technology.
0:00:14 If you were starting from scratch today,
0:00:17 you probably wouldn’t build this app-centric world.
0:00:21 You can imagine a post-phone world.
0:00:24 The past 20 years of consumer technology
0:00:27 have been a story of apps, of touchscreens, and of smartphones.
0:00:31 These form factors seemingly appeared out of nowhere
0:00:34 and may be replaced just as quickly as they were ushered in,
0:00:37 perhaps by a new AI-enabled stack,
0:00:40 a new computing experience that is more agentic,
0:00:42 more adaptive, and more immersive.
0:00:46 Now, in today’s episode, A16C’s growth general partner,
0:00:48 David George, discusses this feature
0:00:52 with arguably one of the most influential builders of this era.
0:00:56 That is Meta CTO, Andrew Boz Bosworth,
0:00:58 who spent nearly two decades at the company,
0:01:01 shaping consumer interaction from the Facebook newsfeed
0:01:05 all the way through to their work on smart glasses and AR headsets.
0:01:09 Here, Boz explores the art of translating emerging technologies
0:01:12 into real products that people use and love,
0:01:14 plus how breakthroughs in AI and hardware
0:01:17 could turn the existing app model on its head.
0:01:22 In this world, what new interfaces and marketplaces need to be developed?
0:01:24 What competitive dynamics hold strong?
0:01:26 And which fall by the wayside?
0:01:28 For example, will brands still be a moat?
0:01:31 And if we get it right, Boz says,
0:01:34 the next wave of consumer tech won’t run on taps and swipes,
0:01:36 it’ll run on intent.
0:01:39 So, is the post-mobile phone era upon us?
0:01:40 Listen in to find out.
0:01:43 Oh, and if you do like this episode,
0:01:45 it comes straight from our AI Revolution series.
0:01:48 And if you missed previous episodes of this series
0:01:50 with guests like AMD CEO Lisa Su,
0:01:52 Anthropic co-founder Dario Amadei,
0:01:55 and the founders behind companies like Databricks,
0:01:56 Waymo, Figma, and more,
0:02:00 head on over to a16z.com slash AI Revolution.
0:02:06 As a reminder, the content here is for informational purposes only,
0:02:09 should not be taken as legal, business, tax, or investment advice,
0:02:11 or be used to evaluate any investment or security,
0:02:14 and is not directed at any investors or potential investors
0:02:15 in any A16Z fund.
0:02:18 Please note that A16Z and its affiliates
0:02:21 may also maintain investments in the companies discussed in this podcast.
0:02:24 For more details, including a link to our investments,
0:02:27 please see a16z.com slash disclosures.
0:02:34 Boz, thanks for being here.
0:02:35 Thanks for having me.
0:02:36 Appreciate it.
0:02:38 Okay, I want to jump right in.
0:02:42 How are we all going to be consuming content
0:02:44 five years from now and ten years from now?
0:02:45 Ten years, I feel pretty confident
0:02:49 that we will have a lot more ways to bring content into our view shed
0:02:50 than just taking out our fun.
0:02:54 I think augmented reality glasses, obviously, are a real possibility.
0:02:55 I’m also hoping that we can do better
0:02:58 for really engaging in immersive things.
0:03:00 Right now, you have to travel to, like, the sphere,
0:03:02 which is great, but there’s one of them.
0:03:03 It’s in Vegas since a trip.
0:03:05 Are there better ways that we can have access to
0:03:07 if we really want to be engaged in something,
0:03:09 not just immersively, but also socially?
0:03:10 So it’s like, oh, I want to watch the game.
0:03:11 I want to watch it with my dad.
0:03:12 I want to feel like we’re courtside.
0:03:15 Sure, we can go and pay a lot for tickets.
0:03:16 Is there a better way?
0:03:16 I think there is.
0:03:18 So ten years, I feel really good
0:03:20 about all these alternative content delivery vehicles.
0:03:21 Five years is trickier.
0:03:24 For example, I think the glasses, the smart glasses,
0:03:26 the AI glasses, the display glasses
0:03:28 that we’ll have in five years will be good.
0:03:31 Some of them will be super high-end
0:03:32 and pretty exceptional.
0:03:34 Some of them will be, like, actually little
0:03:37 and, like, not even tremendously high-resolution displays,
0:03:38 but they will be, like, always available
0:03:39 and on your face.
0:03:41 I wouldn’t be doing work there,
0:03:44 but, like, if I’m just trying to grab simple content
0:03:45 in moments between, it’s pretty good for that.
0:03:47 So I think what we are seeing is,
0:03:50 as you’d expect, we’re at the very beginning now
0:03:53 of a spectrum of super high-end
0:03:54 but probably very expensive experiences
0:03:57 that will not be evenly distributed across the population.
0:03:57 Yeah.
0:03:59 A much more broadly available set of experiences
0:04:01 that are, they’re not really rich enough
0:04:03 to replace, like, the devices that we have today.
0:04:05 And then hopefully a continually growing number
0:04:07 of people who are having experiences
0:04:10 that really could not be had any other way today.
0:04:12 You know, thinking about what you could do
0:04:13 with mixed reality and virtual reality.
0:04:13 Yeah.
0:04:15 We’re going to build up to a lot of that stuff.
0:04:17 So throughout your career,
0:04:20 I would say one of the observations I would have
0:04:22 is you’ve been uniquely good
0:04:26 at piecing together various big technology shifts
0:04:28 into new product experiences.
0:04:31 So in the case of Facebook, early days for you,
0:04:34 obviously you famously were part of the team
0:04:35 that created the news feed.
0:04:35 Yeah.
0:04:37 And that’s a combination of social media,
0:04:38 a mobile experience,
0:04:41 and applying your, like, old-school AI.
0:04:42 Yeah, that’s right.
0:04:42 To it.
0:04:43 The old-school AI.
0:04:43 Yeah, exactly.
0:04:44 But that’s pretty cool.
0:04:46 And, like, a lot of times these trends,
0:04:47 they come in bunches.
0:04:47 Yeah.
0:04:49 And that’s what creates the breakthrough products.
0:04:53 So maybe take that and apply it to where we are today
0:04:55 with the major trends that are in front of you.
0:04:56 Let me say two things about this.
0:04:58 The first one is I think if there was a thing that,
0:04:59 not me specifically,
0:05:02 but I think me and my cohorts at Meta were really good at,
0:05:04 was, like, we really immersed in, like, what the problem was.
0:05:05 Like, what were people trying to do?
0:05:06 What did they want to do?
0:05:08 And when you do that,
0:05:12 you are going to reach for whatever tool is available
0:05:13 to accomplish that goal.
0:05:15 That allows you to be really honest about
0:05:18 what tools are available and see trends.
0:05:21 I think the more oriented you are towards the technology side,
0:05:24 you get caught in a wave of technology,
0:05:25 and you don’t want to admit when that wave is over
0:05:27 and you don’t want to embrace the next wave.
0:05:29 And you’re building technology for technology’s sake.
0:05:29 Yeah, yeah.
0:05:31 So, like, solving a product problem.
0:05:31 But if you’re embracing, like,
0:05:34 what are the issues that people are really going through in their life
0:05:35 and they don’t have to be profound,
0:05:37 I bring that up just because I think we’re in this interesting moment
0:05:40 where I think all of us have been through a phase
0:05:43 where a lot of people wanted a new wave to be coming
0:05:45 because it would have been advantageous to them.
0:05:45 Yeah.
0:05:48 But those things weren’t solving problems that regular people had.
0:05:50 I think the reason we’re so enthusiastic about
0:05:53 the AI revolution that’s happening right now
0:05:56 is it really feels tangible.
0:05:58 These are real problems that are being solved.
0:05:59 And it’s not solving every problem.
0:06:00 It creates new problems.
0:06:01 It’s fine.
0:06:04 So it feels like a substantial real NUCA capability that we have.
0:06:09 And what’s unusual about it is how broad-based it can be applied.
0:06:12 And while it has these interesting downsides today on factuality
0:06:15 and certainly compute in cost and inference,
0:06:18 those types of trade-offs feel really solvable
0:06:20 and the domains that it applies to are really broad.
0:06:22 And that’s very unusual.
0:06:23 Certainly in my career,
0:06:25 you almost always, when these technological breakthroughs happen,
0:06:27 they’re almost always very domain-specific.
0:06:28 It’s like, cool, this is going to get faster
0:06:31 or that’s going to get cheaper or that’s now possible.
0:06:33 This kind of feels like, oh, everything’s going to get better.
0:06:36 Every single interface that I interact with,
0:06:40 every single problem space that I’m trying to solve
0:06:43 are going to be made easier by virtue of this new technology.
0:06:44 That’s pretty rare.
0:06:47 Mark and I always believed that this AI revolution was coming.
0:06:48 We just thought it was going to take longer.
0:06:48 Yeah.
0:06:50 We thought we were probably still 10 years away at this point.
0:06:51 Yeah.
0:06:52 But what we thought would happen sooner
0:06:54 was this revolution in computing interfaces.
0:06:59 And we really started to feel 10 years ago
0:07:02 like the mobile phone form factor, as amazing as it was,
0:07:05 this is 2015, was like already saturated.
0:07:06 That was what it was going to be.
0:07:09 And once you get past the mobile phone,
0:07:10 which is, again, the greatest computing device
0:07:12 that any of us have ever used to this point,
0:07:14 of course, it’s like, okay, well, it has to be more natural
0:07:18 in terms of how you’re getting information into your body,
0:07:20 which is obviously ideally usually through our eyes and ears,
0:07:24 and how we’re getting our intentions expressed back to the machine.
0:07:25 You no longer have a touchscreen.
0:07:26 You no longer have a keyboard.
0:07:29 So once you like realize those are the problems,
0:07:31 it’s like, cool, we need to be on the face
0:07:33 because you need to have access to eyes and ears
0:07:35 to bring information from the machine to the person.
0:07:37 And you need to have these neural interfaces
0:07:40 to try to allow the person to manipulate the machine
0:07:41 and express their intentions to it
0:07:43 when they don’t have a keyboard or mouse or a touchscreen.
0:07:47 And so that has been an incredibly clear-eyed vision
0:07:49 we’ve been on for the last 10 years.
0:07:53 But we really did grow up in an entire generation of engineers
0:07:56 for whom the system was fixed.
0:07:57 The application model was fixed.
0:07:59 The interaction design.
0:08:02 Sure, we went from a mouse to a touchscreen,
0:08:04 but it’s still a direct manipulation interface,
0:08:06 which is literally the same thing that was pioneered in the 1960s.
0:08:09 So we really haven’t changed these modalities.
0:08:11 And there’s a cost to changing those modalities
0:08:14 because we as a society have learned
0:08:18 how to manipulate these digital artifacts through these tools.
0:08:20 So the challenge for us was,
0:08:22 okay, you have to build this hardware,
0:08:24 which has to do all these amazing things
0:08:27 and also be attractive and also be light
0:08:28 and also be affordable.
0:08:31 And none of these existed before.
0:08:32 And what I tell my team all the time is like,
0:08:34 that’s only half the problem.
0:08:37 The other half of the problem is, great, how do I use it?
0:08:39 Like, how do I make it feel natural to me?
0:08:41 I’m so good with my phone now.
0:08:44 It’s an extension of my body, of my intention at this point.
0:08:47 How do we make it even easier?
0:08:49 And so we were having these challenges.
0:08:51 And then, what a wonderful blessing.
0:08:54 AI came in two years ago, much sooner than we expected.
0:08:56 And it’s a tremendous opportunity
0:08:58 to make this even easier for us.
0:08:59 Because the AIs that we have today
0:09:01 have a much greater ability to understand
0:09:02 what my intentions are.
0:09:04 I can give vague reference,
0:09:06 and it’s able to work through the corpus of information
0:09:09 it has available to make specific outcomes happen from it.
0:09:11 There’s still a lot of work to be done
0:09:13 to actually adapt it.
0:09:15 And it’s still not yet a control interface.
0:09:17 Like, I can’t reliably work my machine with it.
0:09:19 There’s a lot of things that we have to do.
0:09:21 We know what those things are.
0:09:24 And so, now you’re in a much more exciting place, actually.
0:09:25 Whereas before, we thought, okay,
0:09:28 we’ve got this big hill to climb on the hardware.
0:09:30 We’ve got this big hill to climb on the interaction design.
0:09:31 But we think we can do it.
0:09:33 And now we’ve got a wonderful tailwind,
0:09:35 where on the interaction design side, at least,
0:09:39 there’s the potential of having this much more intelligent agent
0:09:43 that now has not only the ability for you
0:09:46 to converse with it naturally and get results out of it,
0:09:50 but also to know by context what you’re seeing,
0:09:51 what you’re hearing, what’s going on around you.
0:09:51 Yeah.
0:09:54 And make intelligent inference based on that information.
0:09:56 Let’s talk about, like, reality labs
0:09:58 and this suite of products, what it is today.
0:10:01 So, you have Quest headsets, you have the smart glasses,
0:10:03 and then on the far end of the spectrum is Orion
0:10:05 and some of the stuff that I demoed.
0:10:07 So, just talk about the evolution of those efforts
0:10:10 and what you think the markets are for them
0:10:12 and how they converge versus not over time.
0:10:14 So, when we started the Ray-Ban Meta Project,
0:10:16 they were going to be smart glasses.
0:10:18 And, in fact, they were entirely built,
0:10:20 and we were six months away from production
0:10:22 when Llama 3 hit.
0:10:24 And the team was like, no, we got to do this.
0:10:25 And so, now they’re AI glasses, right?
0:10:27 Like, they didn’t start as AI glasses,
0:10:28 but the form factor was already right.
0:10:30 We could already do the compute.
0:10:31 We already had the ability.
0:10:32 So, yeah, now you have these glasses
0:10:33 that you can ask questions to.
0:10:36 And, in December, to the early access program,
0:10:37 we launched what we call Live AI.
0:10:39 So, you can start a Live AI session
0:10:41 with your Ray-Ban Meta glasses,
0:10:43 and for 30 minutes until the battery runs out,
0:10:44 it’s seeing what you’re seeing.
0:10:45 Yeah.
0:10:46 And it’s funny because, on paper,
0:10:49 the Ray-Ban Meta looks like an incremental improvement
0:10:50 to Ray-Ban Stories.
0:10:52 And this is kind of the story I’m trying to tell,
0:10:54 which is, the hardware isn’t that different
0:10:55 between the two,
0:10:58 but the interactions that we enable
0:11:00 with the person using it
0:11:02 are so much richer now.
0:11:03 When you use Orion,
0:11:05 when you use the full AI glasses,
0:11:08 you can imagine a post-phone world.
0:11:09 You’re like, oh, wow.
0:11:11 Like, if this was attractive enough
0:11:12 and light enough
0:11:13 and had battery life enough
0:11:14 to wear all day,
0:11:15 this would have all the stuff I need.
0:11:17 Like, it would all be right here.
0:11:18 And when you start to combine that
0:11:20 with the images that we have
0:11:21 of what AI is capable of.
0:11:21 So, you did the demo
0:11:23 where we showed you the breakfast.
0:11:24 Yeah, it did.
0:11:25 And it’s, yeah,
0:11:25 and for what it’s worth,
0:11:26 I mean, I’ll explain it
0:11:27 because it’s very cool.
0:11:28 Got to walk over
0:11:30 and there’s a bunch of breakfast ingredients laid out.
0:11:32 And I look at it
0:11:33 and I say,
0:11:33 hey, Meta,
0:11:35 what are some recipes?
0:11:35 That’s right.
0:11:36 And these ingredients.
0:11:38 So, that is, for me at least,
0:11:39 when we think about Orion,
0:11:41 initially,
0:11:43 it didn’t have that AI component
0:11:44 when we first thought about it.
0:11:45 It had this component
0:11:47 that was very direct manipulation.
0:11:48 So, it was very much modeled
0:11:49 on the app model
0:11:49 that we’re all familiar with.
0:11:50 Yeah, of course.
0:11:51 And I think there’s a version of that.
0:11:52 Yeah, of course,
0:11:53 you’re going to want to do calls
0:11:53 and you’re going to want to be able
0:11:54 to do your email
0:11:56 and be able to do your texting
0:11:57 and you want to be able to play games.
0:11:58 We have to play our Stargazer game
0:12:00 and you want to do your Instagram reels.
0:12:01 What we’re now excited about
0:12:01 is, okay,
0:12:02 take all those pieces
0:12:04 and layer on the ability
0:12:07 to have an interactive assistant
0:12:08 that really understands
0:12:10 not just what’s happening
0:12:11 on your device
0:12:13 and what email’s coming in,
0:12:14 but also what’s happening
0:12:16 in the physical world around you
0:12:17 and is able to connect
0:12:19 what you need in the moment
0:12:20 with what’s happening.
0:12:21 And so, these are concepts
0:12:21 where you’re like,
0:12:23 wow, what if the entire app model
0:12:23 is upside down?
0:12:24 What if it isn’t like,
0:12:25 hey, I want to go fetch
0:12:26 Instagram right now.
0:12:26 It’s like, hey,
0:12:28 the device realizes
0:12:28 that you have a moment
0:12:29 between meetings,
0:12:30 you’re a little bit bored.
0:12:30 Hey, do you want to catch up
0:12:31 on the latest highlights
0:12:33 from your favorite basketball team?
0:12:34 Those things become possible.
0:12:35 Having said that,
0:12:36 the hardware problems are hard
0:12:36 and they’re real
0:12:37 and the cost problems
0:12:38 are hard and they’re real.
0:12:39 And come for the king,
0:12:40 you best not miss.
0:12:41 The phone is an incredible
0:12:43 centerpiece of our lives today.
0:12:45 It’s how I operate my home.
0:12:46 I use it in my car.
0:12:46 I use it for work.
0:12:47 It’s everywhere, right?
0:12:50 And the world has adapted
0:12:51 itself to the phone.
0:12:53 So, it’s weird that my ice maker
0:12:53 has a phone app,
0:12:54 but it does.
0:12:54 Like, I don’t know.
0:12:55 I’m not sure.
0:12:56 It seems excessive,
0:12:57 but like,
0:12:58 so somebody today
0:12:58 who’s like,
0:12:59 I got to make an ice maker,
0:13:00 number one job,
0:13:01 got to have an app.
0:13:03 It’s like the smart refrigerator.
0:13:04 You’re like,
0:13:04 I don’t need this.
0:13:05 Take it out of me.
0:13:06 I do think it’s going
0:13:07 to be a long,
0:13:07 that’s why I said
0:13:09 the 10-year view for me
0:13:10 is, I think, much clearer.
0:13:11 I think these things
0:13:13 are going to be available,
0:13:14 widely accepted,
0:13:16 increasingly adopted.
0:13:17 The five-year view is harder
0:13:18 because, man,
0:13:19 like, even if it’s amazing.
0:13:20 Knocking out the dominance
0:13:21 of the phone in five years,
0:13:22 it just seems so hard.
0:13:22 It seems unthinkable.
0:13:24 It’s unthinkable for us, right?
0:13:24 That’s why I said,
0:13:25 like, Orion was the first
0:13:27 time I thought me.
0:13:27 Orion, like,
0:13:28 putting that in my head,
0:13:28 I was like,
0:13:31 okay, it could happen.
0:13:32 Like, there does exist
0:13:33 a life for us as a species
0:13:34 past the phone.
0:13:34 Yeah.
0:13:35 Yeah, it still has
0:13:36 the whole dynamic of,
0:13:37 well, how do I envision
0:13:38 my life without the operating
0:13:39 system that I’m so accustomed to?
0:13:39 Totally.
0:13:40 So I see the physical stuff
0:13:41 that you do,
0:13:42 but just the familiarity
0:13:43 and all the stuff
0:13:44 that’s working in there.
0:13:45 So what do you think
0:13:47 of the interim period?
0:13:49 So maybe you get to the point
0:13:50 where the hardware is capable,
0:13:52 it is market accessible,
0:13:54 but do you tether
0:13:54 to the phone?
0:13:56 Do you take a strong view
0:13:57 that you will never do that
0:13:58 and let the product stand?
0:13:59 Like, how do you think
0:14:00 about that piece?
0:14:02 The phones have this huge
0:14:03 advantage and disadvantage.
0:14:04 Huge advantage,
0:14:05 which is like,
0:14:06 the phone is already central
0:14:07 to our lives.
0:14:08 It’s already got this huge
0:14:09 developer ecosystem.
0:14:10 It’s this anchor device,
0:14:11 and it’s a wonderful anchor
0:14:12 device for that.
0:14:13 The disadvantages,
0:14:14 I actually think what we found
0:14:17 is the apps want to be different
0:14:19 when they’re not controlled
0:14:20 via touchscreen.
0:14:22 And that’s not super novel.
0:14:23 A lot of people failed
0:14:24 early in mobile,
0:14:25 including us,
0:14:26 by just taking our web stuff
0:14:27 and putting it on
0:14:27 the mobile phone
0:14:28 and being like,
0:14:29 oh, the mobile phone,
0:14:30 we’ll just put the web there.
0:14:32 But because it wasn’t native
0:14:33 to what the phone was,
0:14:34 and I mean everything
0:14:36 from interaction design
0:14:38 to the actual design
0:14:39 to the layout
0:14:40 to how it felt,
0:14:41 because we weren’t doing
0:14:43 phone native things,
0:14:44 we were failing
0:14:45 with one of the most popular
0:14:46 products in the history
0:14:46 of the web.
0:14:48 It’s just like the major
0:14:49 design field,
0:14:50 like the skeuomorphic idea
0:14:51 versus the native idea.
0:14:52 Yeah, and I think
0:14:53 having the developers
0:14:54 is a true value,
0:14:54 and I think having all
0:14:55 this application functionality
0:14:56 is a true value.
0:14:58 but then once you actually
0:15:00 reproject it into space
0:15:01 and you’re manipulating it
0:15:04 with your fingers like this
0:15:05 as opposed to a touchscreen,
0:15:06 you have much less precision.
0:15:08 It doesn’t respond
0:15:08 to voice commands
0:15:10 because there’s no tools
0:15:11 for that.
0:15:11 There’s no design
0:15:12 integration for that.
0:15:14 So having a phone platform
0:15:16 today feels like,
0:15:17 wow, I’ve got this huge base
0:15:17 to work from
0:15:18 on the hardware side,
0:15:19 but I’ve also actually got
0:15:20 this kind of huge anchor
0:15:22 to drag on the software side.
0:15:24 And so we’re not opposed
0:15:24 to these partnerships,
0:15:25 and I think it’ll be interesting
0:15:26 to see once the hardware
0:15:27 is a little bit more developed
0:15:28 how partners feel about it.
0:15:30 And I hope they continue
0:15:32 to support people
0:15:33 who buy these phones
0:15:34 for $1,200, $1,300,
0:15:35 being able to bring
0:15:36 whatever hardware
0:15:37 they want to bring
0:15:38 and take the full functionality
0:15:39 of that with them.
0:15:41 The biggest question I have
0:15:42 is whether the entire app model,
0:15:43 because we were imagining
0:15:45 a very phone-like app model
0:15:46 for these devices,
0:15:47 admittedly a very different
0:15:48 interaction design,
0:15:50 input, and control schemes
0:15:50 are very different
0:15:51 and that demands
0:15:52 like a little extra
0:15:52 developer attention.
0:15:54 I am wondering if like
0:15:55 the progression of AI
0:15:56 over the next several years
0:15:57 doesn’t turn the app model
0:15:58 in its head.
0:15:59 Like right now,
0:16:00 it’s kind of an unusual thing
0:16:01 where I’m like,
0:16:03 I want to play music.
0:16:04 So in my head,
0:16:05 I translate that to
0:16:06 I have to go open Spotify
0:16:07 or open Tidal,
0:16:08 and the first thing I think of
0:16:09 is who is my provider
0:16:10 going to be?
0:16:11 Yeah, of course.
0:16:12 As opposed to like,
0:16:12 that’s not what I want.
0:16:13 It’s extremely limiting.
0:16:14 What I want is to play music.
0:16:14 Yes.
0:16:15 And I just want to be like,
0:16:16 go to the AI,
0:16:16 I’m like, cool,
0:16:18 play this music for me.
0:16:18 Yeah.
0:16:20 And it should know,
0:16:21 oh, like you’re already
0:16:23 using this service.
0:16:24 We’ll use that one.
0:16:25 Or these two services
0:16:26 are both available to you.
0:16:27 This one has a better quality song.
0:16:29 Or this one has lower latency,
0:16:29 whatever the thing is.
0:16:30 Or it’s like, hey,
0:16:31 the song you want
0:16:31 isn’t available
0:16:32 on any of these services.
0:16:33 Do you want to sign up
0:16:34 for this other service
0:16:34 that does have the song
0:16:35 that you want?
0:16:36 I don’t want to have
0:16:36 to be responsible
0:16:37 for orchestrating like
0:16:38 what app I’m opening
0:16:39 to do a thing.
0:16:40 We’ve had to do that
0:16:41 because that’s how
0:16:41 things were done
0:16:43 in the entire history
0:16:43 of digital computing.
0:16:45 You had an application-based model
0:16:46 that was the system.
0:16:47 So I do wonder
0:16:49 how much AI inverts things.
0:16:50 That’s a pretty hot take.
0:16:51 Yeah, that’s a hot take.
0:16:51 Inverts things.
0:16:53 And that’s not about wearables.
0:16:54 That’s not about anything.
0:16:54 That’s just like,
0:16:55 even at the phone level,
0:16:57 if you were building
0:16:57 a phone today,
0:16:59 would you build an app store
0:16:59 the way you historically
0:17:00 built an app store?
0:17:01 Or would you say like,
0:17:03 hey, you as a consumer,
0:17:04 express your intention.
0:17:05 Express what you’re
0:17:06 trying to accomplish.
0:17:07 And let’s like,
0:17:07 see what we have.
0:17:08 Let the system see
0:17:08 what it can produce.
0:17:09 Yeah, for you.
0:17:10 But I do think
0:17:11 if you were starting
0:17:12 from scratch today,
0:17:14 you probably wouldn’t build
0:17:15 this like app-centric world
0:17:16 where I, as a consumer,
0:17:17 I’m trying to solve a problem
0:17:18 and first have to decide
0:17:20 which of the providers
0:17:20 I’m going to use
0:17:21 to solve that problem.
0:17:21 Yeah, of course.
0:17:23 That’s fascinating.
0:17:24 And again,
0:17:25 I think it’s a function
0:17:26 of where the capabilities
0:17:27 are today
0:17:27 and I think where we have
0:17:28 line of sight
0:17:29 into orchestration capabilities.
0:17:30 Because I’d say,
0:17:31 knowledge-wise,
0:17:33 that is probably capable today.
0:17:35 I think orchestration-wise,
0:17:35 it’s probably
0:17:37 we’re a little bit away.
0:17:38 And then, of course,
0:17:39 you’ve got to build
0:17:40 the developer ecosystem
0:17:41 to develop on the platform.
0:17:42 Which is incredibly hard.
0:17:43 That’s the thing
0:17:43 I want to see
0:17:44 That’s the hardest piece, right?
0:17:45 That’s the hardest piece.
0:17:45 Yeah.
0:17:46 The stronger we get
0:17:48 at agentic reasoning
0:17:49 and capabilities,
0:17:50 the more I can rely
0:17:52 on my AI
0:17:53 to do things in my absence.
0:17:54 And at first,
0:17:55 it will be knowledge work,
0:17:55 of course.
0:17:56 That’s fine.
0:17:57 But once you have a flow
0:17:59 of consumers
0:18:00 coming through here,
0:18:00 what you’re going to find
0:18:01 is that they’re going to have
0:18:02 a bunch of dead ends.
0:18:02 Yeah.
0:18:03 Where they’re going to ask the AI,
0:18:05 hey, can you do this thing for me?
0:18:05 And it’s going to say,
0:18:06 no, I can’t.
0:18:09 That’s the goldmine
0:18:10 that you take to developers.
0:18:10 And you’re like,
0:18:13 hey, I’ve got 100,000 people a day
0:18:14 trying to solve this problem.
0:18:15 They’re trying to use your app.
0:18:15 Yeah.
0:18:16 They don’t know they are,
0:18:17 but they’re trying to use their app.
0:18:18 Look, here’s the query stream.
0:18:19 Here’s what’s coming through.
0:18:21 And we’re going to tell them no today.
0:18:23 If you build these hooks,
0:18:24 you’ve got 100,000 people
0:18:25 clamoring for something today.
0:18:27 Coming in for your service.
0:18:27 Yeah.
0:18:28 And it’s totally fine
0:18:30 for RAI to go back and say,
0:18:31 hey, you’ve got to pay for this.
0:18:32 There’s a guy who does this for you,
0:18:33 but you’ve got to pay for it.
0:18:34 Yeah.
0:18:34 And by the way,
0:18:35 I’m not just talking about apps.
0:18:38 There’s some kind of a marketplace here
0:18:41 that I think emerges over time.
0:18:43 So that’s how I see it playing out.
0:18:44 I don’t see it playing out
0:18:46 as like someone goes into a darkroom
0:18:47 and comes up with this app platform.
0:18:48 No.
0:18:49 What’s going to happen is
0:18:50 there’s going to become a query stream
0:18:53 of people using AI to do things.
0:18:55 And the AI will fail
0:18:57 repeatedly in certain areas
0:18:59 because that’s a type of functionality
0:19:00 that is currently behind
0:19:01 some kind of an app wall.
0:19:02 And there’s no…
0:19:04 Or it hasn’t been built native
0:19:06 to whatever consumption mechanism.
0:19:07 There’s no bridge that can be built.
0:19:07 Yeah, yeah, yeah.
0:19:08 And everyone wants to build the bridges.
0:19:08 They’re like, no, no.
0:19:10 It’s going to manipulate the pixels
0:19:12 and it’s going to manipulate…
0:19:13 It’s like, fine, it can do those things.
0:19:13 I’m not saying the AI
0:19:15 can’t cross those boundaries.
0:19:16 But I think over time,
0:19:18 that becomes the primary interface
0:19:21 for humans interacting with software
0:19:23 as opposed to the like
0:19:24 pick from the garden of applications.
0:19:25 Yeah, that makes a ton of sense.
0:19:29 That’s a very alluring end state
0:19:31 just as a consumer, right?
0:19:31 Yeah, it’s messy.
0:19:33 And I think it creates
0:19:34 these very exciting marketplaces
0:19:37 for functionality inside the AI.
0:19:39 It abstracts away
0:19:41 a lot of companies’ brand names,
0:19:42 which I think is going to be very hard
0:19:45 for an entire generation of brands.
0:19:45 Yeah.
0:19:47 Like the fact that I don’t care
0:19:48 if it’s being played
0:19:49 on one of these two music services,
0:19:51 that’s hard for those music services
0:19:54 who like really want me to care.
0:19:55 Yeah, yeah, yeah.
0:19:55 And like they want me
0:19:58 to have a stronger opinion about it.
0:19:58 And like they want me
0:19:59 to have an attachment.
0:19:59 Yeah.
0:20:00 I don’t want to have an attachment.
0:20:01 There are some things
0:20:02 where you may value the attachment
0:20:03 and you don’t whatever.
0:20:04 Yeah, in the world
0:20:04 where I’m like,
0:20:05 hey, there’s an app garden
0:20:06 and these two are competing
0:20:07 for my eyeballs,
0:20:09 the brand that they’ve built
0:20:11 is the hugely valuable asset.
0:20:13 In the world where
0:20:15 I just care if the song gets played
0:20:15 and sounds good,
0:20:17 a different set of priorities
0:20:18 are important.
0:20:20 I think that’s net positive
0:20:21 because what matters now
0:20:23 is performance on the job
0:20:23 being asked.
0:20:24 actual product experience
0:20:25 and value
0:20:26 and price per performance
0:20:27 like matters a lot.
0:20:28 Yeah.
0:20:29 I think a lot of companies
0:20:29 won’t love that.
0:20:31 Well, abstracting away,
0:20:33 that’s like effectively articulating,
0:20:35 abstracting away margin pools,
0:20:36 which puts a lot more pressure
0:20:39 on us trusting the AI
0:20:41 or the distributor of the AI.
0:20:42 And so far as I’m floating
0:20:43 between different companies
0:20:44 that are each providing AIs,
0:20:46 the degree which I trust them
0:20:47 to not be bought
0:20:48 and paid for in the back end,
0:20:49 they’re not giving me
0:20:50 the best experience
0:20:51 or the best price per money.
0:20:52 They’re giving the one
0:20:53 that gives them the most money.
0:20:54 Yeah, of course.
0:20:55 So, yeah, it’s the experience
0:20:56 of Google Search today, right?
0:20:57 It’s a very different world.
0:20:59 It’s a very different world.
0:21:00 It’s a very different world.
0:21:01 But you can actually see
0:21:03 inklings of it today, right?
0:21:04 So certain companies
0:21:05 are willing to work
0:21:06 with the new AI providers
0:21:08 in agentic task completion.
0:21:09 Yeah, yeah.
0:21:09 And then they’re like,
0:21:10 well, actually, wait a minute.
0:21:12 I don’t just want the bots
0:21:13 executing this stuff.
0:21:14 I want the humans coming to me.
0:21:15 I think I need that.
0:21:16 Yeah, right.
0:21:16 It’s existential
0:21:18 that I have this brand relationship
0:21:19 directly with the demand side.
0:21:20 Yeah.
0:21:22 So that’s potentially messy,
0:21:24 but a bright future,
0:21:24 especially if we don’t have
0:21:26 to pay that like brand tax.
0:21:27 Yeah, it’ll be very messy.
0:21:30 I don’t know it’s avoidable
0:21:31 because I think once consumers
0:21:33 start to get into these tight loops
0:21:35 where more and more
0:21:35 of their interactions
0:21:38 are being moderated by an AI,
0:21:39 you won’t have a choice.
0:21:40 That’s like where
0:21:41 your customers will be.
0:21:42 But it’s going to be
0:21:42 a pretty different world.
0:21:43 Yeah, it’ll be a different world
0:21:44 and there’ll probably be
0:21:45 some groups that try
0:21:46 to move fast to it.
0:21:47 as a way to compete
0:21:48 with things that are branded.
0:21:48 Yeah.
0:21:49 And just say,
0:21:49 I’m going to compete
0:21:50 on performance and price.
0:21:50 Yeah, that’s right.
0:21:51 Where do you think
0:21:52 that could potentially
0:21:53 happen first?
0:21:55 It probably will mirror
0:21:56 query volume.
0:21:57 I think of this a lot.
0:21:58 We do have a model of this,
0:22:00 which was in the web era
0:22:01 when Google became
0:22:03 the dominant search engine.
0:22:04 So before that,
0:22:05 the web era was like
0:22:07 very index based.
0:22:07 It was like Yahoo
0:22:09 and it was like links
0:22:11 and getting major sources
0:22:12 of traffic to link to you
0:22:12 was the game.
0:22:14 And then once Google
0:22:16 came to dominance,
0:22:17 which happened very quickly
0:22:18 over maybe a couple of years,
0:22:18 I feel like.
0:22:19 All that mattered
0:22:20 was like SEO.
0:22:21 All that mattered
0:22:21 was like where you were
0:22:22 in the query stream.
0:22:22 Yeah.
0:22:23 And the query stream
0:22:26 dictated what businesses
0:22:27 came over and succeeded.
0:22:28 Yeah.
0:22:29 Because like the queries
0:22:30 that were the most frequent,
0:22:31 those were the ones
0:22:32 that came first.
0:22:32 Yeah.
0:22:37 Travel came right away.
0:22:38 It was a huge disruption
0:22:39 and travel agents
0:22:40 went from a thing that existed
0:22:41 to a thing that didn’t exist
0:22:42 in a relatively short-
0:22:42 Immediately.
0:22:44 And they all competed
0:22:45 on the basis of like
0:22:46 execution of the best deal
0:22:47 in a seamless fashion
0:22:49 with the highest conversion.
0:22:50 I think SEO has gotten
0:22:51 to a point now
0:22:53 where it’s kind of a bummer.
0:22:54 It’s like made things worse.
0:22:55 It’s like everyone’s gotten so good.
0:22:56 It’s just like game.
0:22:57 Everyone’s gotten so good at it.
0:22:59 Especially with AI actually now.
0:22:59 That’s right.
0:23:00 So I actually think it’s like
0:23:01 we had this incredible
0:23:02 flattening curve
0:23:02 and now it’s like starting
0:23:03 to kind of rise up
0:23:03 in terms of-
0:23:05 Especially with paid placement too.
0:23:05 Yeah.
0:23:06 That’s so dominant.
0:23:07 So duh.
0:23:08 Yeah, that’s right.
0:23:09 And this is like probably
0:23:10 the cautionary tale
0:23:11 for how this plays out
0:23:12 in AIs as well.
0:23:13 I think there will be
0:23:16 a pretty good golden era here
0:23:17 where the query stream
0:23:18 will dictate
0:23:20 what businesses come first
0:23:21 because those are the queries
0:23:22 that are-
0:23:23 That’s the volume of people
0:23:24 unsatisfied with the existing
0:23:26 solutions that they have.
0:23:26 Yeah.
0:23:27 Otherwise they wouldn’t
0:23:28 be asking about it.
0:23:29 And product providers
0:23:30 and developers will follow that.
0:23:31 And build specifically-
0:23:32 solve those problems.
0:23:33 That’s right.
0:23:34 Once it tips
0:23:36 in each vertical
0:23:37 we get a lot of progress
0:23:37 very quickly
0:23:39 towards better solutions
0:23:39 for consumers.
0:23:41 And then once it hits
0:23:41 a steady state
0:23:42 it starts to be
0:23:43 gamesmanship.
0:23:43 Yeah.
0:23:44 And that’s the thing to fight.
0:23:46 And that’s decaying or-
0:23:47 That’ll be the true test of AI.
0:23:48 The true test.
0:23:48 Can it get through that?
0:23:50 Can it avoid falling into that trap?
0:23:51 Can it avoid that trap?
0:23:51 Yeah, yeah.
0:23:52 That’s right.
0:23:52 Exactly.
0:23:53 Well a lot of that
0:23:54 is business model driven
0:23:55 and we’ll see how that evolves
0:23:55 over time too.
0:23:56 That’s right.
0:23:57 You guys have also been
0:23:59 leading from the front
0:24:00 on this idea of open source.
0:24:01 Yeah.
0:24:02 And so talk about
0:24:03 some of your efforts
0:24:05 on that side of the business
0:24:05 and then
0:24:07 what is the ideal
0:24:08 market structure
0:24:10 of the AI model side
0:24:10 for you guys?
0:24:11 There’s two parts
0:24:12 that came together.
0:24:13 The first one is
0:24:14 Llama came out of FAIR
0:24:16 our fundamental AI research group
0:24:17 and that’s been
0:24:18 an open source
0:24:19 research group
0:24:20 since the beginning.
0:24:20 Yeah.
0:24:21 Since John LeCun came in
0:24:22 and they established that
0:24:24 it’s allowed us to attract
0:24:24 incredible researchers
0:24:26 who really believe
0:24:26 that we’re going to make
0:24:28 more progress as a society
0:24:29 working together
0:24:29 across boundaries
0:24:30 of individual labs
0:24:31 than not.
0:24:33 And to be fair
0:24:33 it’s not just us
0:24:34 obviously
0:24:36 the Transformer paper
0:24:36 was published at Google
0:24:37 and like you know
0:24:39 big self-supervised learning
0:24:39 was our contribution
0:24:40 like everyone’s contributing
0:24:41 to the knowledge base
0:24:43 but when we open source Llama
0:24:44 that’s how all models
0:24:45 were open source
0:24:45 at that point.
0:24:45 Yeah.
0:24:47 Of course.
0:24:48 Like everyone was open
0:24:49 the only thing that was unusual
0:24:50 was everything else
0:24:50 just went closed source
0:24:51 over time.
0:24:51 Yeah.
0:24:52 Effectively.
0:24:52 That’s right.
0:24:53 But before that
0:24:55 every time someone built a model
0:24:55 they open sourced it
0:24:56 so that other people
0:24:56 could use the model
0:24:57 and see how great
0:24:58 that model was.
0:24:58 Like that was like
0:24:59 mostly how it was done.
0:25:00 Sure.
0:25:01 If it was worth anything.
0:25:02 Certainly some specialized models
0:25:03 for translations and whatnot
0:25:04 were kept closed
0:25:05 but like if it was a general model
0:25:05 that was what was done.
0:25:06 Llama 2 was probably
0:25:08 the big decision point for us.
0:25:08 Llama 2
0:25:09 and this is where I think
0:25:10 the second thing
0:25:11 that came in
0:25:12 was a belief that I’ve had
0:25:12 that I was advancing
0:25:14 really strenuously internally
0:25:15 that Mark Ridley believes in too
0:25:16 and he’s written
0:25:16 his post about this
0:25:17 which is first of all
0:25:18 we’re going to make way more progress
0:25:19 if these models are open.
0:25:20 Yeah.
0:25:21 Because a lot of these contributions
0:25:22 aren’t going to come
0:25:24 from these big labs
0:25:24 like they’re going to come
0:25:25 from these little labs
0:25:26 and we’ve seen this already
0:25:26 with DeepSeq in China
0:25:28 which was put in a tough spot
0:25:29 and then innovated
0:25:31 incredibly in the memory architectures
0:25:31 and a couple other places
0:25:33 to really get amazing results.
0:25:34 And so we really believe
0:25:35 we’re going to get
0:25:36 the most progress collectively.
0:25:37 The second thing is
0:25:38 inside this piece is
0:25:39 you know this is a classic
0:25:40 I believe these are going to be commodities
0:25:42 and you want to commoditize
0:25:42 your complements.
0:25:43 Yes.
0:25:44 And we’re in a unique position
0:25:45 strategically where
0:25:46 our products are made better
0:25:47 through AI
0:25:48 which is why we’ve been
0:25:49 investing in it for so long.
0:25:50 Whether it’s recommendation systems
0:25:51 in what you’re seeing
0:25:52 in feed or reels
0:25:54 whether it’s simple things
0:25:55 like what friend
0:25:56 do I put at the top
0:25:57 when you type
0:25:57 you want to make a new message
0:25:58 who do I think
0:25:58 you’re going to message right now?
0:26:00 Little things like that
0:26:01 to really big expansive things
0:26:03 like hey here’s an entire answer
0:26:04 here’s an entire search interface
0:26:05 that we couldn’t do before
0:26:06 in WhatsApp
0:26:07 that like now
0:26:09 is a super popular surface.
0:26:10 So there’s all these things
0:26:11 that are possible for us
0:26:12 that are made better
0:26:13 by this AI
0:26:14 but nobody else
0:26:15 having this AI
0:26:16 can then build our product.
0:26:18 The asymmetry works in our favor.
0:26:18 Yeah of course.
0:26:19 And so for us
0:26:20 like commoditizing your complements
0:26:21 is just good business sense
0:26:22 and making sure that there is
0:26:24 a lot of competitively priced
0:26:26 if not almost free
0:26:27 models out there
0:26:29 helps the entire industry
0:26:31 helps a bunch of small startups
0:26:32 and academic labs
0:26:33 it also helps us.
0:26:34 Yeah you as the application provider
0:26:35 are huge beneficiaries.
0:26:36 So we’re all super aligned.
0:26:37 Yeah you’re aligned.
0:26:38 Business model alignment
0:26:38 and industry alignment.
0:26:40 It’s a strong alignment there.
0:26:40 Yes.
0:26:42 So it comes from both
0:26:43 this fundamental belief
0:26:44 in how this kind of research
0:26:45 should be done
0:26:46 and then aligns
0:26:47 with the other business model
0:26:48 and so there’s no conflict.
0:26:49 Yeah societal progress
0:26:50 plus business model alignment.
0:26:51 It’s all together.
0:26:52 It’s all great.
0:26:53 It’s all going the same direction.
0:26:53 That’s awesome.
0:26:54 It’s great.
0:26:55 I want to shift gears
0:26:56 to talking about
0:26:58 the impediments to progress
0:26:59 and like what you think
0:27:00 you know are kind of
0:27:01 linear versus not.
0:27:03 So the risks to the vision
0:27:04 to the overall vision
0:27:05 that you articulated
0:27:06 obviously hardware
0:27:07 Yep.
0:27:08 AI capabilities.
0:27:08 Yep.
0:27:10 Vision capabilities
0:27:11 and screens
0:27:12 and all that
0:27:12 resolutions.
0:27:14 We talked about the ecosystem
0:27:15 and developers
0:27:17 and native products.
0:27:18 So maybe just talk about
0:27:18 what you see
0:27:19 are kind of
0:27:20 the linear path things
0:27:22 and the things
0:27:22 that may be harder
0:27:23 or riskier.
0:27:26 we have real invention risk.
0:27:27 There exists risk
0:27:28 that the things
0:27:28 that we want to build
0:27:30 we don’t have the capacity
0:27:31 to build as a society
0:27:32 as a species yet.
0:27:33 Yeah.
0:27:34 And that’s not a guarantee.
0:27:36 I think we have windows to it.
0:27:37 You’ve seen Orion
0:27:38 so like it can be done.
0:27:38 Yeah there’s probably.
0:27:39 Yeah it feels like
0:27:40 it’s a cost reduction exercise
0:27:42 it’s a materials improvement exercise
0:27:43 but it can be done.
0:27:44 There is still some invention risk.
0:27:46 Far bigger than the invention risk
0:27:47 I think is the adoption risk.
0:27:48 Is it considered socially acceptable?
0:27:50 Are people willing
0:27:51 to learn a new modality?
0:27:52 Like we all learned to type
0:27:53 when we were kids at this point.
0:27:54 We were born with phones
0:27:55 in our hands at this point.
0:27:55 Yeah.
0:27:56 Are people willing to learn
0:27:57 a new modality?
0:27:57 Is it worth it to them?
0:27:58 Ecosystem risk
0:27:59 even bigger than that.
0:28:00 Like great you build this thing
0:28:02 but if it just does like
0:28:03 your email and reels
0:28:04 that’s probably not enough.
0:28:05 Do people bring
0:28:06 the suite of software
0:28:07 that we require
0:28:08 to interact with
0:28:09 modern human society
0:28:11 to bear on the device?
0:28:13 Those are all huge risks.
0:28:14 I will say
0:28:15 we feel pretty good
0:28:16 about where we’re getting
0:28:17 on the hardware
0:28:18 on acceptability.
0:28:19 We think we can do
0:28:20 those things.
0:28:20 That was not a guarantee
0:28:21 before I think
0:28:23 with the Ray-Van Metaglasses
0:28:23 we’re feeling like
0:28:24 okay we can get through
0:28:25 You feel like the acceptability
0:28:26 Humans will accept
0:28:28 that I’m using technology.
0:28:29 Within that
0:28:31 super interesting
0:28:32 regulatory challenges
0:28:33 here I have
0:28:34 an always on machine
0:28:35 that gives me
0:28:36 super human sensing.
0:28:37 My vision is better.
0:28:38 My hearing is better.
0:28:39 My memory is better.
0:28:40 That means
0:28:42 when I see you
0:28:43 a couple years from now
0:28:44 and I haven’t seen you
0:28:44 on the internet
0:28:45 I’m like
0:28:45 oh god I don’t remember
0:28:47 we did a podcast together
0:28:47 what’s the guy’s name?
0:28:49 Can I ask that question?
0:28:49 Am I allowed
0:28:50 to ask that question?
0:28:50 Yes.
0:28:51 What is your right
0:28:53 it’s your face
0:28:54 you showed me your face
0:28:55 and if I was somebody
0:28:55 with a better memory
0:28:57 I could remember the face
0:28:59 so like that happened
0:29:00 but I don’t have
0:29:00 a great memory
0:29:01 so am I allowed
0:29:02 to use a tool
0:29:03 to assist me or not?
0:29:04 So there’s really
0:29:05 subtle regulatory
0:29:06 privacy
0:29:07 social
0:29:07 acceptability
0:29:08 questions
0:29:08 that are like
0:29:09 embedded here
0:29:10 that are super deep
0:29:10 individually
0:29:12 and can derail
0:29:13 the whole thing
0:29:14 like you can easily
0:29:14 derail
0:29:15 easily derail
0:29:16 the whole thing
0:29:17 and slow progress
0:29:17 that’s the thing
0:29:18 I think we sometimes
0:29:20 think in our industry
0:29:21 it’s like feel the dreams
0:29:21 if you build it
0:29:22 they will come
0:29:22 and it’s like
0:29:23 no a lot of things
0:29:24 have to happen right
0:29:25 well you can also
0:29:26 overstep too
0:29:27 that’s the risk
0:29:28 you’re sure you get
0:29:28 your hands locked
0:29:29 great technology
0:29:31 can get derailed
0:29:32 for long periods
0:29:32 of time
0:29:33 nuclear power
0:29:34 got derailed
0:29:35 yeah for absolutely
0:29:36 stupid reasons
0:29:37 for 70 years
0:29:38 for bad reasons
0:29:39 we know we’re bad now
0:29:40 and it was like
0:29:41 they just played it wrong
0:29:42 yeah of course
0:29:42 and they were like
0:29:43 ah ignore this
0:29:43 it’s like no
0:29:45 people actually feel this way
0:29:46 so I think yeah
0:29:47 I feel pretty good
0:29:47 about the invention risk
0:29:48 acceptability risk
0:29:49 is looking better
0:29:50 than it has been
0:29:51 but like I think
0:29:51 there’s still a lot
0:29:53 of big hedges
0:29:53 to cross there
0:29:55 I actually think
0:29:56 the ecosystem risk
0:29:56 was one
0:29:57 I would have said
0:29:57 previously
0:29:58 was the biggest one
0:30:00 but AI is now
0:30:01 my potential
0:30:02 silver bullet there
0:30:03 if AI becomes
0:30:04 the major interface
0:30:05 then it comes for free
0:30:06 yeah
0:30:07 and I will also say
0:30:08 that we’ve had
0:30:09 such a positive response
0:30:10 from even just
0:30:11 set aside Orion
0:30:12 even the Ray-Ban Metas
0:30:14 companies that want
0:30:14 to work for us
0:30:15 and build on that platform
0:30:16 it’s not a platform yet
0:30:17 yeah it’s not
0:30:17 there’s so little
0:30:18 there’s so little compute
0:30:19 there’s so little compute
0:30:20 we just connect an app
0:30:21 we literally don’t
0:30:22 have any space yet
0:30:22 yeah
0:30:23 but we did do a partnership
0:30:24 with Be My Eyes
0:30:25 which like helps blind
0:30:26 and hard of vision
0:30:26 people navigate
0:30:28 and it’s really spectacular
0:30:29 and so there’s a little window
0:30:29 there where we can start
0:30:30 building
0:30:30 so yeah
0:30:31 I would say the response
0:30:32 has been more positive
0:30:33 than I had expected
0:30:35 so everything right now
0:30:36 tailwinds abound
0:30:36 right now
0:30:37 and to be honest
0:30:38 after eight years
0:30:40 of nine years
0:30:42 of headwinds
0:30:43 having a year of tailwinds
0:30:43 is nice
0:30:43 yeah
0:30:44 I’ll take it
0:30:44 I’ll take it
0:30:45 I’m not gonna look
0:30:45 in the face
0:30:46 no victory laps
0:30:47 yeah but that’s good
0:30:47 okay
0:30:48 but it’s all hard
0:30:50 at every point
0:30:50 it could all fail
0:30:51 yeah I like that you
0:30:51 just started with
0:30:52 it’s invention risk
0:30:53 it’s I don’t know
0:30:54 there’s many ways
0:30:55 this just won’t work
0:30:55 yeah that’s right
0:30:56 even if it does work
0:30:57 it might not take
0:30:59 well I’ll say two things
0:31:00 about this
0:31:01 and this is where
0:31:02 Mark just deserves
0:31:02 so much credit
0:31:04 is we’re true believers
0:31:06 like we have actual conviction
0:31:07 yeah
0:31:08 Mark believes
0:31:10 this is the next thing
0:31:12 it needs to happen
0:31:13 and it doesn’t happen
0:31:13 for free
0:31:15 like we can be the ones
0:31:15 to do it
0:31:16 our chief scientist
0:31:17 Michael Arash
0:31:18 who’s one of my favorite people
0:31:18 I’ve ever gotten a chance
0:31:19 to work with
0:31:20 he talks a lot about
0:31:22 the myth of technological eventualism
0:31:23 it doesn’t eventually happen
0:31:24 there’s a lot of people in tech
0:31:25 who are like
0:31:26 yeah AR will eventually happen
0:31:27 that’s not how it fucking works
0:31:28 that would not
0:31:28 that would actually
0:31:30 AR is a specific one
0:31:30 that would just absolutely not
0:31:32 you have to stop
0:31:33 and put the money
0:31:34 and the time
0:31:34 and do it
0:31:36 somebody has to stop
0:31:36 and do it
0:31:37 and that is the difference
0:31:38 the number one thing I’d say
0:31:39 is like the difference
0:31:40 between us and anybody else
0:31:41 is we believe in this stuff
0:31:42 in our cores
0:31:44 this is the most important work
0:31:45 I’ll ever get a chance to do
0:31:46 this is Xerox PARC level
0:31:48 new stuff
0:31:49 where we’re rethinking
0:31:50 how humans are going to interact
0:31:50 with computers
0:31:52 it’s like JCR Licklider
0:31:53 and the human in the loop computing
0:31:54 we’re seeing that with AI
0:31:55 it’s a rare moment
0:31:56 it’s a rare moment
0:31:57 it doesn’t even happen
0:31:58 once a generation I think
0:31:59 it may happen every other generation
0:31:59 every third generation
0:32:00 like you don’t get a chance
0:32:01 to do this all the time
0:32:03 so we’re not missing it
0:32:03 we’re just like
0:32:04 we’re going to do it
0:32:05 and we may fail
0:32:06 like it’s possible
0:32:07 but we will not fail
0:32:09 for lack of effort or belief
0:32:09 great
0:32:10 thanks a ton boss
0:32:11 cheers
0:32:11 yeah cheers
0:32:20 and we’ll see you in the next video
0:32:21 we’ll see you in the next video
0:32:22 we’ll see you in the next video
0:32:23 we’ll see you in the next video
0:32:23 we’ll see you in the next video
0:32:23 we’ll see you in the next video
0:32:24 we’ll see you in the next video
0:32:24 we’ll see you in the next video
0:32:24 we’ll see you in the next video
0:32:25 we’ll see you in the next video
0:32:25 we’ll see you in the next video
0:32:26 we’ll see you in the next video
0:32:26 we’ll see you in the next video
0:32:27 we’ll see you in the next video
0:32:27 we’ll see you in the next video
0:32:28 we’ll see you in the next video
0:32:28 we’ll see you in the next video
0:32:29 we’ll see you in the next video
0:32:30 we’ll see you in the next video

0:00:03 – Frankly, in the beginning, they didn’t work at all.
0:00:05 That’s when the DOD first started to realize
0:00:07 that it was a totally new threat.
0:00:08 – You can learn them the hard way
0:00:10 by just start a company and go do it,
0:00:11 or maybe you can have some training wheels,
0:00:14 but ultimately there’s no way to learn how to do this job
0:00:15 other than to just do it.
0:00:19 – As taxpayers, as veterans, and as founders,
0:00:21 we should want that level of competition
0:00:22 and that cutthroat competition
0:00:25 to onboard things quickly and offboard things quickly.
0:00:28 – You can’t have a company where people are split
0:00:29 on what they care about
0:00:31 and what those foundational values are.
0:00:33 You have to be all on board.
0:00:35 – Do you want to do something meaningful?
0:00:37 Do you want to do meaningful work?
0:00:38 – It’s very clear what the stakes are
0:00:42 and it’s very clear why it’s important.
0:00:45 – In 2025, the battlefield is not just overseas.
0:00:48 It’s in the tech stacks, the logistics networks,
0:00:50 and airspace security systems
0:00:52 that power our national defense.
0:00:54 And who better to design solutions
0:00:57 than those who have faced these problems firsthand?
0:00:58 Those who have served.
0:01:01 In this episode, recorded live at our third annual
0:01:04 American Dynamism Summit in the heart of Washington, D.C.,
0:01:07 A16Z’s Matt Shortle, operating partner and chief of staff,
0:01:11 plus a veteran himself, sits down with three veterans
0:01:12 who have turned to the private sector
0:01:14 to build technology companies,
0:01:16 tackling secure communications,
0:01:19 military logistics, and drone defense.
0:01:21 Joining this conversation is John Doyle,
0:01:23 founder and CEO of Cape,
0:01:25 David Tuttle, co-founder and CEO of Rune,
0:01:29 and Grant Jordan, founder and CEO of SkySafe.
0:01:31 Today, you’ll get to hear on-the-ground stories
0:01:32 from these veterans,
0:01:35 like how their time in uniform has shaped the scope of problems
0:01:38 that they see and ultimately what they’ve chosen to build.
0:01:41 They also discuss whether a jump directly from military
0:01:43 to entrepreneurship is possible,
0:01:46 or whether a stint in the private sector is needed,
0:01:48 like the time that David spent at Anderil,
0:01:51 or that John spent at Palantir.
0:01:54 And finally, how does this unique group of individuals
0:01:57 who have spent time on the battlefield and in the boardroom
0:02:01 think our collective culture is changing around the national project?
0:02:05 Listen in to find out.
0:02:09 As a reminder, the content here is for informational purposes only,
0:02:12 should not be taken as legal, business, tax, or investment advice,
0:02:14 or be used to evaluate any investment or security,
0:02:19 and is not directed at any investors or potential investors in any A16Z fund.
0:02:22 Please note that A16Z and its affiliates may also maintain investments
0:02:25 in the companies discussed in this podcast.
0:02:27 For more details, including a link to our investments,
0:02:31 please see A16Z.com/disclosures.
0:02:38 I’d like to start with just hearing your initial story,
0:02:40 what you did in the service,
0:02:42 and then also what inspired you to join initially.
0:02:46 What inspired me to join initially was what I think inspires plenty of people,
0:02:48 which is paying for school.
0:02:51 So the Air Force was good enough to pay for me to go to MIT.
0:02:54 And for that, when I commissioned as an officer,
0:02:56 I went and worked in AFRL for four years,
0:02:57 working in the Air Force Research Lab,
0:03:01 and focusing on how do we actually innovate in the Air Force?
0:03:03 How do we bring new systems out to the warfighter?
0:03:07 And how do we rapidly get those systems out there?
0:03:09 I mean, similar family history of service to the nation.
0:03:12 I just service to our people, went to Cornell,
0:03:14 did ROTC there, commissioned into the Army.
0:03:16 I left active service for a period of time,
0:03:17 went to the private sector,
0:03:20 and then actually went back into active service.
0:03:22 Ended up down at the Joint Special Operations
0:03:24 Camp for a number of years and ended up in some technology orgs
0:03:27 at the end of my time there, and then came back out and said,
0:03:28 “Hey, I think we can do some great work
0:03:32 from the private sector technology side going into the defense world.”
0:03:34 I should say I’m still serving,
0:03:36 still serving Army National Guard officer up in New York.
0:03:38 So I do the one weekend a month driving back there,
0:03:41 and that’s been a number of years now at this point.
0:03:43 That’s great. John?
0:03:45 So I almost joined, like a lot of folks of my vintage,
0:03:48 I almost joined on 9/11 or on 9/12.
0:03:50 I’m obviously heavily affected by that.
0:03:52 By that day, I was a senior in college,
0:03:57 and I ultimately made the decision to finish my computer science degree.
0:03:59 But then when we invaded Iraq in 2003,
0:04:02 I felt that same call to service decided it wasn’t just a whim,
0:04:04 it’s something I shouldn’t and couldn’t ignore.
0:04:06 And so the day after the invasion, I went to the recruiter,
0:04:07 and signed up.
0:04:10 I think I’m unique on the panel in that I enlisted.
0:04:14 I was on a program called 18 X-ray, which I think still exists.
0:04:19 You can join the Army as an enlisted soldier and commit to five years up front.
0:04:24 And in exchange, the Army will give you a shot to try out to be a Green Beret,
0:04:27 to be Army Special Forces, after doing basic training in airborne school.
0:04:32 And so I signed that contract the day after the invasion and I was successful.
0:04:36 And so I spent two years doing really amazing training in the Special Forces training pipeline,
0:04:39 and three years on an SF team at the Special Forces Group.
0:04:43 Was entrepreneurship something you envisioned while you were serving?
0:04:44 Did that come after?
0:04:47 Tell me how all that works, because you’re all three founders now.
0:04:52 I don’t think it was something I thought about at the time in service, but I think part of my frustration
0:05:01 in the Air Force was seeing how some of the acquisitions programs were broken or were outdated for what the pace of current technology was.
0:05:09 And I think I saw a lot of opportunities to try to change that from the outside and to try to innovate and bring those technologies in from the outside.
0:05:12 To be honest, it was not something that was on my radar.
0:05:17 When I left the active service for the first time, I went to about the biggest companies you could imagine.
0:05:23 I ended up as an investment banker on Wall Street and did that, obviously massive international bank presence.
0:05:24 I think that was great for me.
0:05:33 I don’t think I had at the time, at my stage in career at that point, I don’t think I had what I thought I needed to go out and start my own company or do the thing that we’re all doing now.
0:05:39 So I think for me, it was like, how do I build that toolkit of knowledge and how do I do those sorts of things in a large organization?
0:05:46 And then now that the time was right later on, opportunity presented to us, hopefully, give back to the mission and help the mission that we can from a startup standpoint.
0:05:49 I think for me, I’ve always been a builder.
0:05:52 The thing I loved about computer science in college was building things.
0:05:54 I’m very proud of my service.
0:05:58 In some ways, that was a little bit of a detour on my inevitable path to starting a company.
0:06:00 In fact, when I left the service, I went to law school.
0:06:05 And the reason I never really practiced law was because I figured out pretty quickly that that was not a path to,
0:06:09 or not, it’s really not a direct path to entrepreneurship or the kind of work I wanted to be doing.
0:06:13 So I wound up at Palantir instead in 2013 and spent nine years at that company.
0:06:17 It was not among the first hundred employees, but it felt very early stage.
0:06:23 And I got a lot of exposure that I think helped me on the eventual journey to founding Cape three years ago.
0:06:30 As you look at when you’re in the service and your problems, you work in aerospace management, logistics, and then privacy first, mobile care.
0:06:35 Were those problems you saw while serving, or did that help influence what you founded later?
0:06:36 Absolutely.
0:06:45 So we at RUN are focused on how do we enable military logistics, cross-services, but really on the Army and the Marine Corps side right now, how do we do field logistics?
0:06:46 How do we enable that?
0:06:52 And I think from my career side, we saw that not to say that military logistics can get the job done in the GWAT era, in our era of serving.
0:07:05 But, you know, I saw very quickly that the same things we had in GWAT from a military logistics standpoint were not necessarily to enable the force to be successful in a peer adversary, whether that’s in competition phase or conflict phase.
0:07:10 When Peter was my co-founder at RUN, we were still at Anderil before we left and founded RUN.
0:07:14 We were thinking about this problem setting, saying like, yeah, this is something that actually needs help and needs attention.
0:07:17 No great technology companies necessarily were really focused on it.
0:07:25 And I said, hey, this is a need that I felt still serving when I was serving in the military, and then this is a need that we think we can help advance now from the private sector.
0:07:38 When I was in AFRL, it was very, very early in the era of DOD thinking about drones on the battlefield and thinking about especially small, inexpensive drones, which at that time, we’re talking like 07, 08 kind of era.
0:07:45 Expensive drones on the battlefield were like, oh, wow, a $100,000 drone, that’s so inexpensive, or $50,000.
0:07:51 Whereas now, what anybody can buy off the shelf for $1,000 or $500 is greater than any capabilities we had at that time.
0:08:02 But at that point, the DOD was first starting to think about what does it really mean when adversaries can bring low-cost capabilities to the battlefield, adversaries who may not have a traditional air force.
0:08:19 And what I saw early on in those initial tests and those initial developments was kind of a disconnect of the old-school, big, heavy military systems trying to be thrown at these small, light, fast threats that were kind of a totally new challenge.
0:08:25 And I saw how difficult it was for this traditional structure and the traditional defense contractors to actually do that.
0:08:36 We saw a lot of big systems, missiles and lasers and all of these things, which were big, expensive, bespoke systems built for a prior era of warfare, trying to be thrown at this small problem.
0:08:38 And frankly, in the beginning, they just didn’t work at all.
0:08:43 And I think that’s when the DOD first started to realize that it was a totally new threat.
0:08:43 It was a totally new thing.
0:08:52 And it’s something we’ve seen borne out since then in pretty much every conflict since then, you know, certainly in Ukraine, where we’ve got tens of thousands of drones flying around all the time.
0:08:59 And so I think that’s what really inspired me and a lot of the folks on our team to do something about that and to try to think about it in a new way.
0:09:03 As you brought up Ukraine, I think, obviously, that’s changed the whole paradigm in your space.
0:09:04 And then also you.
0:09:06 I think, frankly, I would say.
0:09:07 Oh, three, actually.
0:09:22 I don’t want to speak for John, but I think all of us, right, as we look at what has happened in the Russia-Ukraine conflict, right, the new technologies that have been effectively used or ineffectively used by both sides and how that changes things, right, from a logistics standpoint, certainly from a U.S. standpoint, from a communications standpoint.
0:09:23 Yeah.
0:09:31 I think it’s changed, like, that laboratory, for better or worse, has really certainly impacted the way I think about things and we think about things at Rune, and I’m sure it has for both of you.
0:09:38 I left Palantir in February 2022 to start CAPE, left Palantir on a Friday, checked into that we work on a Monday morning, and the invasion was happening.
0:09:52 The very first question that we asked ourselves at CAPE was, if you recall, commercial cellular was a huge factor, continues to be a huge factor in Ukraine, but was enormous in the initial days as a force multiplier, but also as a risk.
0:09:59 Both sides were using commercial cellular heavily and enjoying benefits of that, but also literally targeting missile strikes against each other based on that signature.
0:10:12 And so the first question we asked ourselves as a company was, what tech do we wish the Ukrainians were using right now, and how do we get that tech into Taiwan ahead of an eventual Chinese invasion that hopefully never happens, but we certainly need to prepare for?
0:10:15 And that’s really been a motivating question for the company from the beginning.
0:10:18 I think if you look at logistics, obviously they stopped outside of Kiev.
0:10:19 Yep.
0:10:21 So the Russians had a few issues there.
0:10:34 As we look at going from veteran to founder, can you jump straight from veteran and found a company, or do you need to stop somewhere in between an Anderol, a Palantir, and know what right looks like, know what culture you need to set?
0:10:36 I’d love to hear your thoughts on that.
0:10:39 I think you can become a founder at any point in your career.
0:10:44 If it’s something you’re going to do and it’s something that’s sort of in your blood, I think that it’s inevitable.
0:10:54 I will say personally, my stint, which turned out to be almost a decade, at Palantir in between service and starting a company, I learned a ton.
0:11:09 The culture at Palantir, I think, is borderline legendary at this point, and the reputation is warranted, but also the mechanics of how to operate in a company, what does an innovative company look like, fostering innovation and allowing a little bit of chaos and a little bit of things to bubble up within the stew, but without the wheels coming off.
0:11:12 Those are important lessons for a founder to learn.
0:11:18 You can learn them the hard way by just start a company and go do it, or maybe you can have some training wheels for a few years at a Palantir in Anderol.
0:11:29 I would say my experience at Palantir certainly was helpful and I think has helped set us up for success in some ways, but ultimately there’s no way to learn how to do this job other than to just do it.
0:11:30 Right.
0:11:34 Yeah, I would say I think it’s a very individual and a very personal question, like what are your life experiences, right?
0:11:39 I think we all come out of the military with a great set of leadership skills, a great set of communication skills.
0:11:42 We understand we have some domain expertise, right, in what we’re doing.
0:11:50 But I think it’s really boils down to, again, to use the toolkit analogy of like how strong do you as a veteran feel coming out of this and committed to it?
0:11:52 And do you think you have the skills to be able to do it?
0:11:58 To John’s point, I think for me personally, I wanted to get out in the time that I had in finance and then the time that I had at Anderol.
0:12:00 Very similar to like the Palantir ethos there.
0:12:03 How do you move fast and break things, but that’s okay to get towards a mission and do those things.
0:12:07 To me, that is super valuable experience as I now lead a team and lead the company.
0:12:11 But that’s not to say that we have different paths to get here.
0:12:12 I mean, all three of us have different paths.
0:12:17 That doesn’t mean you might not have had different experiences prior to service, right, before serving or coming out.
0:12:22 And I think that’s a very individual, I hate to be a cop out on it, but I think it’s a very individual choice to make.
0:12:24 I would agree with that.
0:12:28 I would say for me, when I left the Air Force, I went to grad school after that.
0:12:36 And I think having some distance from it is helpful, whatever that distance is, whether that’s a big corporation or doing something entirely different or grad school.
0:12:42 I think having a little bit of distance, seeing a little bit of a different set of culture or a different set of problem sets.
0:12:54 Also, you know, it helps you maybe in some ways distance yourself from some of the frustrations of the government bureaucracy and the military and all of that to then kind of refresh yourself to go face those challenges and jump back into it.
0:13:06 But I think also just trying to figure out how you should do things differently out in the private sector, out in the commercial sector, how you can take the good parts of the military, of government service and apply them.
0:13:13 Take the good parts, ignore the bad, drop the bureaucracy bits that are not helpful to anybody and figure out how you can benefit from it.
0:13:18 But I think there’s a lot of good things to take from it, especially a lot of the kind of mission-driven focus.
0:13:27 We’ve had lots of folks on our team over the years who have been prior military or prior intelligence, and there’s no replacement for that mission-driven focus.
0:13:30 There’s no replacement for that actual commitment to making a difference.
0:13:33 Matt, where do you come down on that question?
0:13:35 So that’s a good question, and we debate that internally.
0:13:41 Personally, I transitioned out, and I was part of Lehman Brothers’ 08 MBA class, which went really well.
0:13:42 I was there 22 years.
0:13:44 Probably not technically your fault, though.
0:13:45 That’s not your fault.
0:13:45 No, it wasn’t my fault.
0:13:47 Me and Dick Folder were really tight.
0:13:50 But I think it helps to go to somewhere and see what right looks like.
0:13:57 For me personally, especially Anderol, Palantir, SpaceX, and then Tesla, just some companies that are producing these founders.
0:13:58 I think it really helps.
0:14:01 But there’s obviously folks that can make that jump.
0:14:03 So I think the answer is either.
0:14:06 But for me personally, I would want to stop along the way.
0:14:08 You made a great point as well that I think is important.
0:14:12 I think you always have to be careful that you don’t want to just create the thing that you thought you knew.
0:14:14 needed when you were in the military, right?
0:14:16 And I think that is sometimes a trap, right?
0:14:17 We all need to use our experience to do that.
0:14:22 But when I talk to folks about that, if you try to build the thing that you thought you needed,
0:14:28 every month that goes by, every year that goes by, what you thought you needed then is now out of date and it’s stale, right?
0:14:34 So especially as you lead a company, and I’m sure Sean can talk about it too, is you need to be constantly thinking forward,
0:14:42 not just creating the thing that you thought you needed in the Argandab River Valley in 2011 or 12 or in Herat or wherever it happened to be.
0:14:42 Yeah, I agree.
0:14:44 And we went 4G to 5G.
0:14:44 Let’s go to 10G.
0:14:46 Why stop at 4?
0:14:50 Or we built the Hornet, which is a Category 4 fighter, and we built the Super Hornet with Category 4+.
0:14:52 Skip that like the Marine Corps did.
0:14:57 I think that’s the essential skill for a founder, is that mindset.
0:14:59 In my opinion, you have that or you don’t.
0:15:00 That’s part of who you are or it’s not.
0:15:07 There are other skills that support that overall mindset, that insistence on way forward thinking, ambitious goal setting.
0:15:10 And you can learn those skills, certainly in the military.
0:15:12 You can learn those skills at great companies like the ones you mentioned.
0:15:17 So they’re supporting skills you can learn, but the core, that core mindset, I think, is just innate.
0:15:22 I don’t know your founding stories necessarily, specifically, but finding the right co-founder was important.
0:15:23 For me, right?
0:15:26 So finding the right software engineer, technical co-founder, right?
0:15:29 So Peter is my co-founder and CTO.
0:15:32 Like, I don’t know how I would have done that straight coming out of service, right?
0:15:35 So I needed to go to a place, thankfully a great place like Anderil,
0:15:39 where we’re surrounded by those types of people to be able to find that partner.
0:15:43 And now he and I, our Venn diagrams don’t overlap that much, which is great in a co-founding pair.
0:15:46 Let’s pull in a thread that Grant brought up, and let’s go culture.
0:15:47 You might have some good responses here.
0:15:50 Culture in the military, culture in startup land.
0:15:54 And then you could even add finance, city, which is very different.
0:15:54 So that’s a third.
0:15:59 But let’s focus on military and then startup land differences and similarities.
0:16:06 I think I probably have a unique perspective from the military side by virtue of being an 18x-ray
0:16:10 and only ever experiencing the special operations community in my military service.
0:16:16 That is a naturally more meritocratic, flatter organization, I think, than the rest of the DoD.
0:16:19 Having said that, there’s still a rank structure.
0:16:20 There’s still centralized planning.
0:16:26 There’s still a very clear hierarchy and a very clear set of doctrine and a book that literally tells you how to do your job.
0:16:30 Even in the context of special operations where you’re granted a lot more creativity and a lot more autonomy,
0:16:32 that structure still exists.
0:16:36 What is amazing about a startup is none of that exists.
0:16:41 You know, Monday morning, you check into a WeWork, and now you’ve just got a company and a couple bucks in the bank, hopefully.
0:16:43 And there’s literally nothing else.
0:16:44 You don’t have an email address necessarily.
0:16:47 And so you get to write the entire thing yourself.
0:16:50 That can be really disorienting, especially in my experience.
0:16:52 We have a ton of veterans at the company, and they’re all amazing.
0:16:54 And I love to hire veterans.
0:16:56 But that can be disorienting coming out of service.
0:17:02 And so you have to create, in my opinion, a culture around, I think it’s a little bit trite to say celebrating failure,
0:17:06 but at least openly accepting failure when it happens.
0:17:11 If people are working hard and people are taking risks and pushing the envelope and their efforts fail,
0:17:16 at CAPE anyway, the only thing you need to do is be transparent about that and call it out when we fail
0:17:19 and adjust and maybe name a couple of lessons learned for next time.
0:17:22 And it doesn’t do any damage to your reputation at the company or your ability to operate.
0:17:27 I think that’s the most important cultural aspect at a startup because it encourages risk-taking.
0:17:29 It encourages people to lean forward and push hard.
0:17:33 I find that there’s like common threads across all of them.
0:17:38 So when I think of like out of broad buckets, military service, my finance experience, and now startup experience,
0:17:43 there are like broad threads that make somebody or make something successful there,
0:17:46 which is initiative, drive, taking chances and doing those things, right?
0:17:49 Whether that is part of a deal team in investment banking,
0:17:53 whether that was part of a team within the JSOC enterprise or the military writ large,
0:17:55 or now it’s part of a startup, right?
0:17:56 So it’s, yeah, how do you celebrate success?
0:17:58 How do you encourage risk-taking?
0:18:01 I think also I view it as like, how do you hire talent?
0:18:05 But you understand that when that talent comes in, you got to train, you got to mentor, you got to develop, right?
0:18:08 And it’s not necessarily always at a startup, at least from my standpoint, at Rune.
0:18:11 It’s not been hiring just because somebody has a skill.
0:18:17 It’s hiring somebody that I think can grow into a position or do these things or with the right coaching and development can take chances
0:18:20 and maybe fail, as John says, but like you move on from that.
0:18:23 So I think there are like commonalities across all of these.
0:18:28 Obviously, at the startup, you don’t have the structure in place and the support structure that’s in place.
0:18:31 It’s certainly at a large company, but also somewhat in the military side.
0:18:35 So there’s some differences there, but I think from a human standpoint and a culture standpoint,
0:18:39 you can have some threads of success that are common across those.
0:18:44 Probably for most of our companies, it’s not just about government or military.
0:18:45 Everything is dual use.
0:18:48 Technology crosses a lot of different industries.
0:18:52 For us, it’s not just the kind of like military and federal market.
0:18:57 It’s also critical infrastructure, oil and gas, power companies, all of these large organizations
0:19:02 that face a lot of the same challenges that military or federal government face.
0:19:07 I think for us, it’s about balancing that culture and trying to figure out how to take the really mission-driven
0:19:11 and what that really translates to is customer-driven, user-driven.
0:19:13 I think there’s commonality there.
0:19:18 Not an over-obsession with the kind of military portion of it, but taking the part about it
0:19:23 about what are we trying to accomplish and how are we trying to help these people, these organizations.
0:19:25 Yeah, you mentioned dual use.
0:19:28 And first of all, I’d love you to define, are you dual use?
0:19:30 Would you consider all three of you dual use or not?
0:19:36 And then any cultural issues with that, bringing folks on board on the national defense side?
0:19:40 Number one, I would say we’re definitely dual use because kind of threats and concerns
0:19:46 about drones in the national airspace, in critical areas, absolutely cut across all sorts of different areas.
0:19:52 Whether it’s an Air Force base or a power plant or border security, everyone is concerned.
0:19:56 And not just because they’re concerned about just malicious drones, it’s about being able
0:20:00 to use drones for good things and use drones for inspecting infrastructure and doing logistics
0:20:01 and all of these things.
0:20:06 But I think in that, you do need to make sure that you’re building a culture that understands
0:20:09 why you’re there and what your values are and what you care about.
0:20:12 In our world, right, we’re going to end up dealing with serious stuff.
0:20:17 And you can’t just have people who wanted to build iPhone games and so they just happen
0:20:17 to get a job.
0:20:20 They’re like, no, you have to understand we’re here to do serious stuff.
0:20:26 We had folks actually go over and assist the Ukrainians and provide support and services
0:20:28 and technology to them.
0:20:33 You can’t have a company where people are split on what they care about and what those foundational
0:20:33 values are.
0:20:35 You have to be all on board.
0:20:39 Yeah, there’s some great quotes out there by Palmer Luckey talking about taking technology
0:20:43 talent from the commercial sector and how do you convince them to come into the defense
0:20:43 world.
0:20:45 And I will say that, frankly, it’s not that hard.
0:20:46 Do you want to do something meaningful?
0:20:47 Do you want to do meaningful work?
0:20:51 And I think that’s the genius behind a lot of the American Dynamism companies, to include,
0:20:54 I would say, all three of ours, is like, how do you bring that amazing technical talent
0:20:58 and those amazing software engineers or hardware engineers, the case may be, and bring them
0:21:03 to something like, hey, do you want to do something meaningful for the nation, for the people and
0:21:04 do those types of things?
0:21:07 And how do you fuse that together with those of us who have come out of the military and
0:21:08 do those sorts of things?
0:21:12 I don’t want to say it’s easy, but it’s not as hard as people would think to get them motivated
0:21:16 about doing what’s right for the country, doing what’s right for our people and to do those
0:21:17 sorts of things.
0:21:21 And I’ll also add, I think what really sorts people out on whether or not they want to
0:21:26 join a company that does these sorts of work and these sorts of missions is you show them
0:21:27 what you’re doing, right?
0:21:30 Nothing sorts that out faster than real-world examples.
0:21:36 When we show drones in places that are being run by prison gangs to smuggle fentanyl into
0:21:40 prisons or in military environments, whatever, it’s very clear what the stakes are, and it’s
0:21:41 very clear why it’s important.
0:21:46 And for those people who it resonates, it’s instantaneous, and it’s, wow, I want to be working on this.
0:21:48 And for those, it doesn’t, great.
0:21:49 Like, that’s fine.
0:21:52 There’s plenty of iPhone game companies to go join, and that’s great.
0:21:57 I think one of the things I really love about Cape and the company we’ve built is we really
0:22:03 are truly dual-use, and we’re dual-use in some sense defense and consumer, much in the same
0:22:04 way that Signal is dual-use, right?
0:22:08 It’s like Signal is used by privacy-preferring people all around the world doing all kinds
0:22:08 of stuff.
0:22:13 And also everyone here at the Summit, everyone on Capitol Hill has Signal installed on their
0:22:13 phone.
0:22:15 It’s very valuable technology.
0:22:19 I’m personally very motivated, and the company is very motivated by building this tech to
0:22:24 keep folks doing special operations work safe and allowing them access to commercial cellular
0:22:26 without incurring all the risk that typically comes with that.
0:22:30 It’s the same set of issues, and frankly, everyone has the same use case for their phone.
0:22:32 We have customers who are journalists.
0:22:36 We have customers who are survivors of domestic violence.
0:22:41 We do work with the Electronic Frontiers Foundation on detecting cell-size simulators at protests
0:22:42 and at demonstrations.
0:22:46 All those people have the same set of issues and the same desire to leverage the value of
0:22:50 the commercial cellular network without making that typically inherent set of compromises.
0:22:54 And you ask a really good question, which is, does that change the kind of people you can
0:22:55 attract, or how does that affect recruiting?
0:23:00 I think at Cape anyway, first of all, I’m like, I’m incredibly proud of the team that we’ve built,
0:23:02 and so I’m biased here.
0:23:06 But you just have, you have to work hard to find people who embrace both of those missions
0:23:07 simultaneously.
0:23:11 But when you find them, they just tend to be awesome, awesome folks, and the kind of people
0:23:13 that I really like to work with, and the people I want to work with.
0:23:18 One thing I’ll just say, tactically, that we’ve done internally that I really believe in is
0:23:24 we have not split the business into commercial and defense divisions, for lack of a better
0:23:24 word, right?
0:23:29 There’s an engineering team, and the engineering team, we all agree on a set of priorities that
0:23:32 serve the entire customer base, and I think that’s worked out really well.
0:23:34 We’re building one product, and it really serves both purposes.
0:23:35 That’s great.
0:23:38 You’re all busy building your own companies.
0:23:43 When you served, did you see any gaps there that you wish people would start looking to
0:23:44 build companies?
0:23:45 Any requests for startups?
0:23:47 It’s a big question.
0:23:49 This is such an unfair question.
0:23:53 I realize now that I ask every government person I talk to, if you have the magic technical
0:23:55 wand, what would you wave it and create?
0:23:57 Who, height or weight, body armor, and kit?
0:24:01 Give me all that stuff that I need to carry around and make it weigh 30% of what it weighs.
0:24:02 80 pounds of gear in your back.
0:24:03 Yeah, yeah.
0:24:03 Right, right, right.
0:24:05 Make it 25 pounds and make it less cumbersome.
0:24:08 I mean, I was actually just having a conversation this morning.
0:24:13 As we think about logistics and how do you enable autonomous logistics systems, maritime,
0:24:14 assets, ground assets.
0:24:19 One of the things, though, that we’re interested in is heavy lift aerial resupply platforms,
0:24:19 right?
0:24:21 Which is a hard problem, right?
0:24:23 Autonomous air vehicles, obviously a thing.
0:24:27 UASs, how do you actually lift substantial cargo with that?
0:24:33 As we think about driving the automation of military logistics or through autonomous vehicles, right?
0:24:36 Getting an air system that actually can carry the weight for that is something that’s like
0:24:39 particularly interesting to me and is a very hard thing to do.
0:24:48 Part of the thing with both military and government now is the challenges they face are less kind of hardware centric than they used to be.
0:24:51 It’s not about owning a particular set of kit.
0:24:57 A lot of it is about data and data sharing and being able to have the correct information to make decisions.
0:25:03 A lot of what’s needed now is how does data get shared between organizations.
0:25:12 So for us, right, in the drone tracking world, we actually shifted business models a lot over the last few years from what was initially traditional hardware sales,
0:25:15 building the fancy piece of military hardware and selling it.
0:25:19 And pretty much as soon as it goes out the door, you kind of never see it again, mostly if things go well.
0:25:28 But what we saw was that the need really was for the information to make those decisions to take action and that it wasn’t about having a fancy piece of hardware.
0:25:36 But I think changing the way that the government looks at that and the way that they buy things and the way that they’re able to actually share information, that’s the hard part.
0:25:46 And part of dual use and part of the cross problems between government agencies and private companies is being able to share information between those two.
0:25:54 It doesn’t do the power plant any good if the federal law enforcement agency next door can’t share with them the threats that are happening.
0:26:04 So I think anything that encourages and enables data sharing between organizations with government is super, super valuable and definitely needed in a lot of these cases.
0:26:05 You touched on hardware, software.
0:26:09 As we look at the future of war, we don’t want to be fighting the last war.
0:26:10 Obviously, we want to fight the next war.
0:26:15 We look at Ukraine, perfect test case for all three of your companies there.
0:26:22 Do you think we have the right weapons, whether it’s hardware, software, or do we need to be looking out further and making a bigger leap?
0:26:26 Do we need different types of programs, vice trillion-dollar programs?
0:26:34 I think that maybe the common thread to pull out as we think about the next war is an emphasis on asymmetric capabilities.
0:26:39 A lot of attention is being paid, I know, rightfully, to the need for low-cost achievable systems, right?
0:26:44 That’s, I think, there’s a growing consensus that’s the future of warfare in a lot of ways.
0:26:52 I think, although it’s certainly no secret, people don’t always contextualize cyber warfare in the same way as an asymmetric capability.
0:26:56 Relatively low amounts of investment can cost huge amounts of damage in an adversary.
0:27:01 That’s sort of where CATE plays, is in the cyber vulnerabilities in the global cell network.
0:27:04 But in general, that’s the theme to pay attention to.
0:27:10 Rather than investing in trillion-dollar bets on absolute dominance in a relatively small number of exquisite systems,
0:27:15 how do we win the scrappy, messy, asymmetric war that’s sort of inevitable next time around?
0:27:18 Do we feel we have the things now?
0:27:19 I mean, no.
0:27:23 I mean, that’s why we started, or at least in the middle of the time, that’s why all of us started our companies, right?
0:27:28 Is we feel like there is something there that we can give back to the military that doesn’t exist right now,
0:27:30 or we can do it better, or we can help enable operations.
0:27:33 One of the things we need to look at is, how do we do these things faster, right?
0:27:35 At least as a software company, right?
0:27:42 The idea of a 20-month prototyping effort to potentially get to something that then maybe a year later is going to transition.
0:27:44 I’m using that as a very specific example right now.
0:27:46 It’s just, it’s absolutely insane.
0:27:49 Would be insane to the commercial sector, right?
0:27:51 Is insane when you think about software development cycles.
0:27:57 So I think the ability to rapidly prototype things, develop them, get them out to the force, test, iterate with them,
0:28:00 and then frankly for the government to say, yes, that worked, great, cool, move on.
0:28:02 No, it didn’t, and off-ramp.
0:28:07 I think that’s what we are totally on board with, is the ability for the government to force us to perform,
0:28:09 to compete, to do things for the warfighter.
0:28:11 And if we are doing great, then awesome.
0:28:15 And if not, then off-board us and bring somebody else in, right?
0:28:20 And I think how you do those things, I think, as taxpayers, as veterans, and as founders,
0:28:26 we should want that level of competition and that cutthroat competition to onboard things quickly and off-board things quickly.
0:28:30 And I think that’s finally, the administration’s obviously moving towards that way, that direction.
0:28:33 And I think that’s the goodness for the force and the goodness for the nation to move that way.
0:28:38 I think a lot of attention has rightfully been invested in, how do we onboard things quickly?
0:28:42 You know, the resurgence to OTAs is the most visible example of this, and it’s great.
0:28:44 DIU’s doing amazing work.
0:28:49 I think there’s sneaky work to be done on killing stuff that’s not working still.
0:28:51 And there are big, high-profile, very expensive examples.
0:28:55 But also within the innovation ecosystem, just call it when you see it.
0:28:58 And if something’s not working and something’s not going to pan out, just kill it.
0:29:03 You do everybody a favor, including the folks at that company, if you say, this isn’t working, and you cut it off.
0:29:08 Yeah, the worst thing, like you have a zombie program that just goes on with millions or tens of millions of dollars,
0:29:11 or sometimes even more of that, that’s not being used by warfighters.
0:29:14 I think to John’s point, like, you’re not doing the service any favors.
0:29:15 You’re not doing the warfighter any favors.
0:29:16 Definitely not the taxpayer.
0:29:22 And frankly, you’re not doing that company any favors because you’re putting them into a false sense of that solution is useful,
0:29:27 and you’re stifling potentially that company to, like, reinvent itself or go forward and do something new on that.
0:29:28 So I think you’re right.
0:29:30 I think off-boarding is just as important as onboarding.
0:29:35 Yeah, I think one of the gaps right now is, like you said, DIU’s done a lot of great stuff.
0:29:42 A lot of different efforts have been done to try to get early-stage startups involved in those first government contracts,
0:29:44 that kind of $1 million phase, $2 million phase.
0:29:50 But we still have that gap, that kind of valley of death, of how do we take any of those successful programs
0:29:53 and scale them up to that mid-size level?
0:29:57 Not everything is going to jump from a $1 million to a $200 million contract.
0:30:01 How do we scale them up, get them out to the forces, get them out to the services,
0:30:03 and see the potential once you actually scale it up?
0:30:11 Because I think that’s one of the challenges is, it’s easy to deploy a couple systems in a little POC and see,
0:30:12 okay, well, that’s pretty useful.
0:30:16 But a lot of the types of things we’re talking about, you don’t really see the real value of them
0:30:19 until you have them at some sort of scale.
0:30:22 If I’m talking about monitoring the airspace and looking at all the drones,
0:30:27 it’s great at the kind of tactical level to see, oh, we’ve got a drone crossing into this airbase here.
0:30:34 But as soon as you scale it up to the larger scale of saying, oh, well, now we can see patterns of activity.
0:30:39 We can see that, oh, that’s funny, this drone showed up at three different bases over time.
0:30:44 And we can do that only because we’ve started to build that out at a not large scale.
0:30:49 We’re not fully committing to everything, but we can see the potential of it at a kind of mid-size scale, larger scale.
0:30:51 And I think that’s a missing piece that we have.
0:30:55 How do you scale it up to actually try it in the real world at a scale that matters?
0:30:57 Yeah, you brought up programs.
0:31:02 CIA’s got In-Q-Tel, DIU’s for the DoD, not only programs, but how to select the best tech.
0:31:05 And obviously, we think you three are all the best.
0:31:06 You’re our portfolio companies.
0:31:14 But you need to build that market map, find out the category leader, because it’s going to go 90% leader, 10%, everybody else gets steak knives.
0:31:17 So we need to help them select the best tech also.
0:31:22 And I think one of the challenges on the government side with that, especially on the DoD side,
0:31:29 is because people and officers rotate out so frequently, you run into situations where somebody started out a really good program.
0:31:31 They started to see the benefits of it.
0:31:34 And then a brand new person walks in the door the next day and says, OK, what are we doing?
0:31:36 And you’re starting from scratch.
0:31:38 And they don’t really understand why were the decisions made?
0:31:39 What was the intent?
0:31:47 And so I think there needs to be a little bit of a better focus on continuity of ushering these programs across those eras and understanding why they’re doing them.
0:31:51 Military units are renowned for their trust and confidence, their cohesion.
0:31:53 We all saw that, especially you and the Special Forces.
0:31:57 How do you instill that in your companies?
0:32:00 How do you build that teamwork, that trust, confidence, cohesion?
0:32:05 We’re an in-person company, which I don’t think is the final answer, but really helps.
0:32:07 Proximity matters, in my opinion.
0:32:13 When people are in the office working together, you get a lot of that more easily than you do in a remote configuration.
0:32:20 The key ingredient, the thing you’re really trying to build that the military has in space, you really want at your startup,
0:32:27 you want to know when you need someone on your team that they’re there and they’re picking up on the first ring.
0:32:30 I think that’s the most important thing in life, certainly in company building.
0:32:37 And so my barometer for this within the company is when someone, it just happened this weekend,
0:32:42 we had a PZERO with someone who was traveling abroad in support of a customer and was having an issue,
0:32:46 and they spun up an internal thread, and immediately seven people were on the thread,
0:32:49 and it was the right seven people, and they were all helping to troubleshoot,
0:32:51 and within 20 minutes they had the guy turned around and good to go.
0:32:57 And that sort of mindset, no one complained about it, and we didn’t even mention it.
0:33:00 On Monday morning, it was just, that’s how we do business at Cape, right?
0:33:06 And that’s where you want to get, in my opinion, and the way you do that, to quote Ben Horowitz,
0:33:07 culture is what you do, it’s not what you say.
0:33:11 And so as the founder, your job is to be on every single one of those Slack threads,
0:33:12 especially in the early days, right?
0:33:15 And to be at the office every single day and to be working on those things.
0:33:17 And people will follow your lead.
0:33:21 And then if you do it enough, and you do it over and over and over again, it becomes part of the culture.
0:33:22 I’ll amplify that.
0:33:27 The collaboration that happens across engineering, product, growth, company leadership, all in one place.
0:33:31 I can’t imagine having not had that when we first opened our first initial hires,
0:33:34 because your first hires are really also building the culture with you.
0:33:38 Yes, it’s top-down from founders and from the executive level,
0:33:40 but like those initial hires are part of that culture for us.
0:33:42 And then I think it’s the normal things.
0:33:44 After work, during Sun-Fridays, we do a happy hour.
0:33:46 Every single Friday, we go to the same bar.
0:33:48 But it’s little things like that.
0:33:49 I mean, what does it cost?
0:33:50 It costs us an hour of time.
0:33:52 We cut out a little bit early on a Friday,
0:33:54 but I also know that everyone’s working generally right now.
0:33:56 We’re working Saturdays over the weekend.
0:33:59 You know, in one day, let’s presume success, and we have thousands of employees.
0:34:01 I don’t think I can still fit in the same bar.
0:34:03 But right now, that’s important, right?
0:34:05 It’s important to know that, especially as a software company,
0:34:07 where my capital is my people, right?
0:34:10 It is my engineers, and it is my growth professionals.
0:34:12 And that’s really what we need to grow and build.
0:34:17 I think the biggest thing for us has been actually recognizing and calling out
0:34:19 real-world impacts of the things we do.
0:34:24 I think, especially initially, we were actually very bad at celebrating our successes
0:34:25 and calling attention to it.
0:34:29 I think part of that is because so many of our folks are very mission-driven
0:34:31 and very like, okay, we got to do this thing.
0:34:35 And so you accomplish something, and you never even take the time to notice.
0:34:40 And I think that is one of the good things of having that mix in the culture
0:34:46 of kind of the mission-driven military and intelligence people versus the normal people
0:34:48 is sometimes the normal people kind of step back and they’re like,
0:34:50 holy shit, like, what did we just do?
0:34:51 That’s amazing.
0:34:52 That’s incredible.
0:34:54 And there’s plenty of other people who are just like,
0:34:55 oh, well, yeah, we do this all the time.
0:34:56 It’s no big deal.
0:35:00 But actually stepping back and recognizing that and celebrating those wins,
0:35:04 I think that’s part of what really builds that culture and that cohesion.
0:35:08 And then that builds a culture in which when things are going bad,
0:35:10 when things are going wrong, which are always going to happen,
0:35:12 there’s always going to be new things popping up.
0:35:16 Suddenly, that’s a culture where everyone understands why they’re like,
0:35:18 okay, cool, let’s pull together a team.
0:35:18 Let’s hop on a flight.
0:35:21 Let’s go fix this thing, do this thing right now.
0:35:25 And everybody’s on board because they understand why they’re doing it
0:35:25 and what the stakes are.
0:35:27 All right, big question here.
0:35:29 Take your time before you answer.
0:35:32 Would you ever go back to the military,
0:35:35 either in a leadership, advisory, or an innovation role?
0:35:38 I think innovation roles are really interesting.
0:35:40 There are so many of them across government.
0:35:43 And without picking on any in particular,
0:35:44 a lot of them never get anything done.
0:35:46 So I know internally,
0:35:49 someday if I’m lucky enough to be approached about that sort of a job,
0:35:51 I’m going to have a list of demands.
0:35:52 Here are the ways I need to be empowered.
0:35:56 Here’s the authorities that I need so that I can actually make a real impact
0:35:57 and be empowered.
0:35:59 But given the right set of conditions, I would love it.
0:36:02 That would be a defining career milestone just like starting a company.
0:36:05 This is a little bit different for me because I am still serving
0:36:07 in the Army National Guard in that way.
0:36:09 And I love leading soldiers and I love being around soldiers.
0:36:13 But I think going back into like full-time some kind of role or governmental role,
0:36:16 if I am successful to the point that I’m offered that opportunity at some point,
0:36:16 same as John,
0:36:18 yeah, I think that would be an honor of a lifetime to go back.
0:36:19 I mean, I think we were all,
0:36:23 yes, we are doing these to grow companies and to do those things.
0:36:25 And I’m not ashamed to say we are here to make some level of money.
0:36:28 I mean, that is what we do as companies.
0:36:30 But there is a selfless service aspect to that.
0:36:34 And if that selfless service could eventually be back in a governmental role
0:36:37 and helping the mission, if I believe I can make an impact and be helpful there,
0:36:39 then yeah, that would absolutely be something I would do.
0:36:41 Yeah, I totally agree with that.
0:36:43 I think being able to know that you can have an impact
0:36:46 and being able to know that you actually have the authority
0:36:50 and the ability to go and do the things that need to be done as you see them.
0:36:52 Coming from the Air Force acquisition side,
0:36:55 I have strong opinions about what’s needed in that world
0:36:57 and acquisition reform and whatnot.
0:37:00 But one request, one thing that I know I would have
0:37:02 that I didn’t realize until I knew it was an option.
0:37:05 A number of years ago, I was talking with the Secretary of the Air Force
0:37:10 and she said, wow, we need a lot more folks like you coming back into service.
0:37:12 And I said, well, I don’t really know.
0:37:14 And she said, well, I can get you a beard waiver
0:37:16 so you wouldn’t have to shave your beard.
0:37:17 I’m like, okay, now you’re talking.
0:37:20 That’s my, gotta have a beard waiver.
0:37:21 That’s nice beard.
0:37:22 There you go.
0:37:22 Done.
0:37:26 Now, if you made it this far,
0:37:28 a reminder that this was recorded live
0:37:30 at our third annual American Dynamism Summit
0:37:32 in the heart of Washington, D.C.
0:37:35 And if you’d like to see more exclusive content from the Summit,
0:37:38 head on over to a16z.com
0:37:41 slash American dash Dynamism dash Summit.
0:37:43 Or you can click the link in our description.

0:00:04 The industrial capacity of China is fearsome.
0:00:08 Being able to deploy highly autonomous AI-driven drones at scale
0:00:10 is still a domain that we can win in.
0:00:14 I think the technologies that matter most to the future of war
0:00:17 are right there in front of us.
0:00:20 I think a modern conflict becomes basically like a software writing fight.
0:00:22 It will be the pace of deployment
0:00:27 that is the make or break for militaries around the world.
0:00:30 The game theory here is just as simple and obvious as it can be.
0:00:35 Unless we find a way for industry and government to work together,
0:00:39 we will find ourselves in a very tough situation.
0:00:43 Will Durant once said, quote,
0:00:46 war is one of the constants of history, unquote.
0:00:49 And while the presence of war has not changed,
0:00:51 the way it’s conducted has.
0:00:55 It is technology, whether steel, gunpowder, radio, GPS,
0:00:59 or nuclear weapons, which have defined conflicts over the eras.
0:01:02 And while the images of tanks or machine guns
0:01:04 dominate the visuals we have of war,
0:01:07 these are not the decisive technologies of the future.
0:01:09 And here’s the thing.
0:01:11 The future of warfare isn’t coming.
0:01:12 It’s already here.
0:01:16 It’s fought in the skies over Ukraine, Israel, and beyond
0:01:18 by AI-empowered drones.
0:01:22 These drones have become a crucial weapon of war with asymmetric capability,
0:01:26 where a handful of drones costing hundreds or thousands of dollars
0:01:30 can disable equipment like tanks or aircraft that cost orders of magnitude more.
0:01:33 We’re literally off by like three orders of magnitude.
0:01:37 Within a few short years, drones have gone from a reconnaissance tool
0:01:40 to one that ensures aerial and battlefield dominance.
0:01:43 The army with more drones has a decisive advantage.
0:01:48 So, with China dominating about 80% of the global drone market,
0:01:52 what does that say about our national security in the growing power competition?
0:01:54 Why have we fallen so far behind?
0:01:56 And what will it take to build up our domestic drone industry?
0:02:00 Plus, where should autonomy play a role in military decision-making,
0:02:03 when lives are literally on the line?
0:02:07 In today’s episode, recorded live at our third annual American Dynamism Summit
0:02:09 in the heart of Washington, D.C.,
0:02:13 A16Z’s Senior National Security Advisor, Matt Cronin,
0:02:16 sits down with two people who have been thinking about these questions
0:02:21 and building solutions dating all the way back to when this industry was full of hobbyists.
0:02:24 That is Ryan Tseng, co-founder and CEO of Shield AI,
0:02:28 and Adam Breit, co-founder and CEO of Skydio.
0:02:33 Shield has been building intelligent systems like AI-powered fighter pilots
0:02:35 and drones since 2015,
0:02:40 while Skydio has been manufacturing drones for use in the battlefield since 2014,
0:02:44 and today is the largest U.S. drone manufacturer by volume.
0:02:47 So who will command the skies in the years to come?
0:02:49 Listen in to find out.
0:02:54 As a reminder, the content here is for informational purposes only,
0:02:57 should not be taken as legal, business, tax, or investment advice,
0:03:00 or be used to evaluate any investment or security,
0:03:02 and is not directed at any investors or potential investors
0:03:04 in any A16Z fund.
0:03:07 Please note that A16Z and its affiliates
0:03:10 may also maintain investments in the companies discussed in this podcast.
0:03:13 For more details, including a link to our investments,
0:03:15 please see A16Z.com slash disclosures.
0:03:26 We’re here to chat about drones, autonomy, and great power conflict.
0:03:31 Now, both of you have started incredibly successful and innovative drone-focused companies.
0:03:35 At the same time, both of you started when the industry was truly nascent.
0:03:40 It was seen as a field for hobbyists rather than something that could actually shape the future of warfare.
0:03:46 So what led you to be interested in this field and to spend so much of your time,
0:03:50 blood, sweat, and tears, on this now incredibly important industry?
0:03:53 I had started and sold a company to Qualcomm.
0:03:59 And I’ve always been somebody that has had an intense passion to compete, fight, win at whatever I was doing.
0:04:02 And in my last year, Qualcomm was there for about four years.
0:04:05 I didn’t have that fire in the belly that I had had throughout my life.
0:04:09 And so I started thinking about what it was that would really motivate me,
0:04:11 not for the next five years, but for the next 50.
0:04:15 And I decided that if I could find the intersection of three things,
0:04:18 a noble mission, the chance to work with extraordinary people,
0:04:21 and a chance to define the possible, I’d have that fire in the belly.
0:04:26 And my brother went on to become a Navy SEAL, so a totally different track in life.
0:04:29 And he was getting ready to go to business school.
0:04:32 I encouraged him to think about what he wanted to do in life.
0:04:35 And he came to me with the idea of bringing the best of what was going on
0:04:39 in the autonomous driving sector to the mission of protecting service members and civilians.
0:04:45 He felt like if somebody would do that, it would have brought home a lot of his friends safely to their families.
0:04:50 And he felt like it would be a pillar for the future of American military dominance.
0:04:52 I thought it was an extraordinary mission.
0:04:54 I told you earlier, I thought it was a stupid business.
0:05:00 And so I wished him luck and suggested that he come up with a better business to make mission impact.
0:05:03 But SEALs are very persistent people.
0:05:04 My brother is no exception.
0:05:09 And for a long story short, I started to spend more time with him learning about the challenges.
0:05:13 And I was just shocked by the scope of the problem and how little was being done about it.
0:05:19 And I’ve been proud and humbled every day for the last 10 years to have an opportunity to contribute to a mission that I think is so important.
0:05:19 Incredible.
0:05:21 Well, I’m glad SEALs are persistent in general.
0:05:25 And I’m especially glad that your brother, the SEAL, dragged you along in this important mission.
0:05:26 Yes.
0:05:26 What about you?
0:05:29 Yes, we come at this from very different places.
0:05:42 First, I think that S.H.I.E.L.D. deserves just enormous credit for 10-plus years ago recognizing the need and the opportunity for what we now think of as defense tech at a time when that really did not exist.
0:05:44 And people thought they were crazy for even trying.
0:05:47 Candidly, I don’t think we get that same kind of credit.
0:05:49 So I grew up flying radio-controlled airplanes.
0:05:51 I’ve basically been doing grown stuff my whole life.
0:06:03 And that led me to be a grad student at MIT in the late 2000s, early 2010s, when you could basically take radio-controlled airplanes and put computers and sensors on them and write software to get them to do smart stuff.
0:06:10 So I really became obsessed with trying to build AI systems that could fly better than people could and wasn’t thinking so much about applications at the time.
0:06:18 But in 2013, 2014, my lab mate and I started to look out and see there were interesting things starting to happen with small light quadcopters.
0:06:23 And we felt like the applications and implications of the technology could be enormous.
0:06:27 But needing to have an expert pilot there flying it was just sort of a fundamental restriction.
0:06:35 So the big bet that we made when we started Skydio was that AI and autonomy built into a small light quadcopter was going to be very powerful for a wide range of industries.
0:06:40 And the government and enterprise applications were always part of the vision.
0:06:47 But we explicitly decided to start with a consumer product because we thought that something like light, integrated, easy to use would be a really good platform for this other stuff.
0:06:52 And for me personally, when I was in grad school, I was wrestling with, you know, I love this technology.
0:06:54 I wonder if I want to keep working on it.
0:06:56 Am I going to have to go work for a defense contractor?
0:07:02 I deeply believe in like the mission of the U.S. military, but the idea of working at a traditional defense contractor was very unappealing to me.
0:07:05 And so we started with a consumer product.
0:07:11 And it’s reflective of kind of the trajectory of the space that very quickly, you know, like 2018, 2019 timeframe.
0:07:19 And I think to the credit of the U.S. military, they realized that these small light civilian quadcopters had enormous value on the battlefield.
0:07:30 And the technology from consumer to enterprise to military is so tight that in the span of a year, basically, we won our program of record, the Army Short Range Reconnaissance Program.
0:07:34 And I think it was officially announced in 2021, but it really started in 2018, 2019.
0:07:38 And that was the beginning for us of expanding to serve a much broader set of markets.
0:07:40 Adam was also part of my early story.
0:07:43 I don’t know if you know this, Adam, but I read all of your papers.
0:07:44 We’ve learned a lot since then.
0:07:46 Who’s the foundation?
0:07:46 Yeah, yeah.
0:07:50 No, I mean, look, a lot of the ideas that we use at Skydio came from the research community.
0:07:53 Some of the research that we had done, some that other folks had done.
0:08:05 So one of the things I think has most shocked many and fascinated military strategists is how drones have reshaped the nature of warfare in the past decade, particularly in the past three years.
0:08:08 There’s just been huge proliferation, right?
0:08:11 Far more mass being brought to the battlefield via drones.
0:08:19 And it’s enabling much more distributed, decentralized, and lethal force structures, right?
0:08:29 Anybody running around in a truck can pop out with a drone that might go 1,000 nautical miles or a collection of drones that might go 1,000 nautical miles and hit who knows what far off into the distance.
0:08:36 There used to be much more concentration of forces, dependence on large, exquisite assets to deliver capabilities.
0:08:40 So it’s been just a complete transformation.
0:08:44 And the world has seen a lot of that unfold in the conflict in Ukraine.
0:08:50 I think one of the major questions is how can the United States and her allies adopt those lessons?
0:08:53 Because the things that we’re doing have a lot of merit.
0:08:56 It has been a force structure that’s dominated the last several decades.
0:09:01 But our adversaries have spent a long time thinking about how to counter what we’re doing.
0:09:12 And I think that it’s going to be important for us to think about how we can embrace 100 times more systems to empower our service members to be lethal and effective, to come home safely to their families.
0:09:15 And I think drones, in many ways, are the future of war.
0:09:16 Absolutely.
0:09:30 And Adam, one of the things that particularly shocked those observing, the Ukraine conflict in particular, is you would see, as writers are referring to, say a commercial quadcopter like Escadio in some instances, they’ll go and take out an exquisite system like a tank.
0:09:35 Two or three of those packed with sufficient munitions could cripple a large armored vehicle.
0:09:41 So how have you seen that sort of shift impact how military looks at dual-use commercial drones?
0:09:46 To be honest, I think that’s a question that the U.S. military has not fully digested yet.
0:09:52 And one of the realities of drones is that it just creates this massive asymmetry, right?
0:09:57 A system that costs a few thousand dollars can take out a system that costs a few million dollars.
0:10:01 And I don’t think we’ve fully grappled with that yet, to be honest.
0:10:04 I think that we’re seeing evidence of this in Ukraine.
0:10:14 And the Ukrainians, largely out of necessity, and really just incredibly impressive ingenuity, they have a broad array of ground robots or drones, air drones, sea drones.
0:10:16 They’re building them at an incredible rate.
0:10:17 They’re iterating very quickly.
0:10:19 And it’s very scrappy.
0:10:23 The U.S. military may not be subject to the same kind of constraints that the Ukrainians are.
0:10:29 But I think that we need to understand that the possibility for that asymmetry is real.
0:10:37 Like, the possibility of building very low-cost systems that are very capable and capable of delivering strikes or capable of maintaining surveillance is very real.
0:10:40 And our adversaries are likely to take advantage of it.
0:10:43 That’s a journey that we still need to go on to some extent.
0:10:46 There’s still quite a bit of inertia and momentum in our military.
0:10:49 I don’t know what your experience has been towards, like, larger, more traditional, exquisite systems.
0:10:51 I think that is starting to pivot.
0:10:54 But that’s a real question for us for the future.
0:10:58 There’s a phrase used in the military context often.
0:11:00 It’s that quantity has a quality all its own.
0:11:08 And there is an incredible disparity between what our chief adversary, the People’s Republic of China, can produce in a month for drones.
0:11:12 Whether it’s going to be formal military-style drones or dual-use commercial drones.
0:11:22 And that scale, I think a lot of people are wondering what that means in order to deter a future conflict or, God forbid, whether it would be a conflict in the Taiwan Strait or elsewhere.
0:11:33 Adam, would you mind just talking about the sort of differences in scale, just roughly speaking, between the production capabilities of both countries and why that is, particularly on a regulatory basis?
0:11:35 So I mentioned I grew up flying radio-controlled airplanes.
0:11:39 So basically all radio-controlled airplanes were made in China in the 90s and 2000s.
0:11:41 And nobody was thinking too hard about that.
0:11:42 They still are, right?
0:11:43 They probably still are, yeah.
0:11:47 And, you know, that didn’t seem like a national security issue at the time.
0:11:55 But if you think about what a drone is, it’s basically like the combination of radio-controlled airplane-type stuff, motors, and consumer electronics.
0:11:57 A lot of the same stuff that goes into a phone.
0:12:01 And, you know, as a country, we basically outsource manufacturing to China.
0:12:06 And I think that that was a series of policy decisions, expediency on the part of business.
0:12:19 That was a mistake from a military standpoint because one of the major themes is that the gap between kind of civilian technology and consumer technology and military technology is closing in many of these domains.
0:12:31 The general manufacturing capacity in China for low-cost, capable compute systems, which are really now becoming robotic systems, not just drones, but other kinds of robots, is substantial.
0:12:32 It’s a whole ecosystem.
0:12:35 And it’s not about cost either at this point.
0:12:46 It’s really about, like, technical expertise, built capacity in terms of all the different things that it takes to, like, mold and machine and build PCBs, the circuit boards, and place components on them.
0:12:49 So I don’t think this is something that we can solve overnight.
0:12:59 The ultimate thing in TBD on if this is attainable, I always say, like, wherever they’re building iPhones, they’re going to have a really good ecosystem for building drones and other kinds of electronics.
0:13:06 And so the real prize is, can we bring that level of scaled manufacturing back to the U.S.?
0:13:09 I don’t know if we can, to be honest, but I think it’s worth a shot.
0:13:12 And from the outset, as a company, we’ve been manufacturing our drones in the U.S.
0:13:18 And I would say that when we started doing that in 2016, it felt like we were swimming upstream into, like, a fast-flowing river.
0:13:26 I would say that, like, we’re maybe starting to get some, like, signs of tailwinds, especially with the new administration, which, you know, is cause for optimism.
0:13:31 But I think this is one that we just can’t give up on because robots are going to become more and more important.
0:13:39 They’re going to be using the same kind of ingredients from consumer electronics and other kinds of, like, relatively low-cost systems.
0:13:41 Cars are going in this direction as well.
0:13:45 Cars are starting to look more like laptops and phones in terms of the components that are in them.
0:13:49 This is a combination of industry and policy and all of us working together.
0:13:51 And I think there’s a cause for optimism over the last couple of years.
0:13:55 There’s still a lot of work to do, but I don’t think it’s an insurmountable hill for us.
0:13:58 I’ve got a, I guess, good news, bad news take.
0:14:00 I think the bad news, you sort of led with it.
0:14:04 The industrial capacity of China is fearsome.
0:14:10 And try as we might, I think that’s going to be a very difficult thing to close out in one year and even a decade.
0:14:16 And I’m an optimist and always like to believe that there’s a way, but it is a substantial gap.
0:14:23 And I think we’re sort of approaching, to use a rocket term, a Max-Q moment from a national security perspective where we’re undergoing a forced transformation.
0:14:27 Huge technology changes are afoot.
0:14:29 Things like AI are coming into play.
0:14:36 We’ve got an adversary that’s become extremely wealthy, huge industrial capacity, also investing massively in their military capabilities.
0:14:39 And so the question then becomes, what’s the right play?
0:14:49 And I think that history has shown, World War I, World War II, Vietnam, tremendous amounts of mass were brought to the battlefield.
0:15:00 And in some cases, despite tremendous amounts of mass being brought to sections of the battlefield, it just turned into grinded out, war of attrition without a lot of movement on either side.
0:15:02 And we start to see some of that in the Russia-Ukraine conflict as well.
0:15:07 Tremendous amounts of mass being brought to the battlefield, but front lines that are extremely difficult to move.
0:15:12 And I think a reason for that is there’s a difference between mass and effect.
0:15:20 And simply, the world is big, and targets are relatively small compared to the scale of the world.
0:15:25 And it turns out that just throwing a bunch of mass downrange, it can still be pretty hard to hit the things that matter.
0:15:33 And where the United States has dominated over the last couple decades is the software prowess, the AI, the autonomy.
0:15:45 And so I think that if we can combine the fantastic work that’s going to re-industrialize the United States to build more mass, to build more capability, we have to do that.
0:16:01 But if we can combine it with the software and autonomy capabilities, if we can close, like, the OODA loop so that we can push software updates at a moment’s notice and make every ounce of charge, every minute of flight time, maximally effective, I think that’s how we can compete.
0:16:06 There’s a second wave that hasn’t really broken yet, which is really AI and autonomy.
0:16:10 The vast majority of what’s happening in Ukraine is still basically one-to-one.
0:16:17 You’ve got these FPV pilots who are expert operators who are flying the thing, or they’re flying drones with very limited, pretty simple mission.
0:16:20 Go to these coordinates and deliver a strike.
0:16:31 And I think that we are going to see over the next decade another fundamental change as rather than having these drones animated by an operator on the ground or by, like, relatively simple algorithms on board,
0:16:35 they become animated by really advanced autonomy, and they can communicate with each other.
0:16:42 The implications of that, I think, are probably going to be even more significant than the first step to just having these things be unmanned.
0:16:47 And I do think that’s an area where, as a country, it plays more to our strengths.
0:16:51 And you can’t forget about the hardware and the manufacturing capacity.
0:16:52 You need to be able to do that stuff.
0:16:57 But winning on the AI front, I think, is even more important.
0:17:01 So to ensure we dominate, to ensure we win, Vice President Vance,
0:17:05 just spoke a few moments ago at the American Dynamism Summit, and he said,
0:17:11 our goal is to essentially remake the economy, to fix that mistake that you guys were referencing earlier,
0:17:15 about we just offshored all of our manufacturing, we can’t make things anymore.
0:17:21 And there are leading members in Congress committed to that, and also committed to defense procurement reform.
0:17:27 So if you had an opportunity to sit down with anyone, the leaders in the executive branch, leaders in the legislative branch,
0:17:31 or anyone who would be listening right now, those leaders or staffers for them,
0:17:37 like if you just fix one or two really pivot points, and if those were resolved, we could do so much more,
0:17:41 either for drones in particular, manufacturing generally, unleashing AI.
0:17:43 What sort of recommendations would you give?
0:17:47 Number one, I would ask that they continue to do what they’ve been doing,
0:17:52 which is reinforce their belief and investments in the incredible people that sign up to serve.
0:17:58 The one thing that I think we’ve got absolutely right is we have brilliant people that are brave.
0:18:03 They believe in the values of this country, and they just go and do extraordinary things.
0:18:05 And it’s been a privilege of mine to be able to meet them.
0:18:10 My brother drew me into that universe, and I just think it is such a gift to the nation
0:18:12 that we have these people that are willing to do what they do.
0:18:17 Now, we strive to contribute to make sure they can be as effective as possible
0:18:19 to fight when the Tura come home safely to their families,
0:18:22 and to provide them the best possible tools.
0:18:24 And look, what do I know?
0:18:27 The administration has hard jobs, so let me put that disclaimer up front.
0:18:31 I think the technologies that matter most to the future of war,
0:18:34 that matter most to the future of this country,
0:18:38 and all of our allies around the world are right there in front of us.
0:18:46 And I think the security challenge of our time is whether or not we can mobilize the bureaucracy
0:18:50 to go reach out and pick up what’s on the table.
0:18:54 Everybody already knows is one of the most important capabilities to the future,
0:18:59 and that is, like, taking and making real the large-scale deployment
0:19:02 and operationalization of autonomy technologies.
0:19:07 If you look in the kind of space that we play, which are small, light quadcopters
0:19:09 that weigh a few pounds that are soldier-carried,
0:19:13 the Ukrainians today are using these things at the rate of millions per year,
0:19:15 literally millions per year.
0:19:18 It’s the primary method through which they’re delivering strikes and surveilling the battlefield.
0:19:24 The U.S. military, I think, to their credit, has programs generally pointed in this direction,
0:19:28 but those programs are operating at the scale of thousands, like single-digit thousands.
0:19:32 So we’re literally off by three orders of magnitude, I would argue,
0:19:38 relative to what evidence suggests the modern battlefield demands.
0:19:40 There are similar trends in Israel as well.
0:19:44 I mean, Israel has rapidly adopted these class of systems at substantial scale.
0:19:47 And this is an area, to the point about re-industrialization,
0:19:51 where military purchasing power makes a massive difference.
0:19:58 The consumer-civilian quadcopter markets are measured in the scale of single-digit billions.
0:20:01 That’s, like, comparable, I would argue, to what the military should be spending in the space.
0:20:04 They’re not spending anywhere close to that today for this class of system.
0:20:11 And so I think that’s a pretty obvious lever that has a bunch of benefits from a national security perspective.
0:20:15 I mean, one, you’re equipping our soldiers with modern, relevant technology.
0:20:19 But two, the purchasing power that the military can bring to bear there is significant enough
0:20:24 to actually move the needle from an industrial-based standpoint for this class of technology
0:20:25 that serves other markets.
0:20:28 Not just us, but, like, the companies in our space serve public safety
0:20:32 and critical infrastructure inspection, and there’s huge technology overlap
0:20:35 between the products that you use to do that
0:20:37 and the products that folks use on the battlefield.
0:20:40 I think if you just went down all the quantities of everything
0:20:42 and just added a zero behind all of them,
0:20:44 and then, of course, there’s a cost challenge,
0:20:46 and we have to figure out how to do that.
0:20:48 Yeah, I mean, I think there’s some things that you can delete,
0:20:49 and deleting those things is also painful.
0:20:51 Well, maybe not a zero.
0:20:52 Sometimes a zero in front of something.
0:20:58 Yeah, yeah, no, I think that’s the trade-off that needs to be made.
0:21:02 Buying legacy exquisite systems
0:21:04 that cost hundreds of millions of dollars,
0:21:05 some cases billions of dollars,
0:21:08 there’s a trade-off between one or two of something
0:21:11 or 10,000 or 100,000 or a million drones.
0:21:13 And if you just look into the future,
0:21:15 which of those things is going to be more powerful?
0:21:18 I think, like, large quantities of AI-driven drones is,
0:21:20 it’s not the only thing that you need,
0:21:22 but I think in many situations is the right answer.
0:21:23 Well, let’s dive into it a little bit more.
0:21:26 So, as you both noted a few moments ago,
0:21:29 the U.S. military, starting around the 1980s, 1990s,
0:21:33 really went all in on highly expensive, exquisite systems,
0:21:36 small numbers, and that worked fine for us.
0:21:38 But our now near-peer adversary,
0:21:42 China, developed an asymmetric military designed to counter that.
0:21:45 So, they have a carrier-strike missile.
0:21:47 So, for a cost of, say, 100 million,
0:21:50 they take out tens of billions, if not more, on our side,
0:21:52 plus the horrible lives we lost.
0:21:55 So, there’s been a move within the DoD
0:21:57 to counter the counter,
0:21:59 to have a project replicator.
0:22:02 Let’s presume even if we move those decimal points,
0:22:03 so that all of a sudden,
0:22:05 the older systems has less procurement,
0:22:10 and then the newer systems have more dollars allotted to procure.
0:22:13 China, some would argue,
0:22:14 still have certain advantages,
0:22:17 perhaps because they’ve had more time working on it,
0:22:18 perhaps because they’re more subsidized.
0:22:20 One of them, people would argue,
0:22:22 and I welcome you to say that this is wrong,
0:22:24 would be in the area of swarm technology.
0:22:25 So, you can see the light shows
0:22:28 over Shenzhen and other cities
0:22:30 for the Chinese New Year,
0:22:32 where it was just extraordinary shows
0:22:33 with hundreds of thousands
0:22:36 that broke the Guinness World Record book again this year.
0:22:39 Do we have that level of sophistication,
0:22:40 and why or why not?
0:22:42 And what can we do to make sure
0:22:43 that we not only achieve parity if we have not,
0:22:45 but achieve superiority in that space?
0:22:48 So, this is an area that we’re quite focused on.
0:22:50 I mean, swarm is sort of one of these terms
0:22:52 that can mean a lot of different things, right?
0:22:53 And it’s really like,
0:22:55 what tasks are you trying to accomplish?
0:22:57 So, when you see the drone light shows,
0:22:59 they’re 100% relying on GPS,
0:23:01 so they’re using GPS to figure out where they are
0:23:03 and to position themselves precisely,
0:23:07 and they’re 100% reliant on a comms link
0:23:08 between all the drones all the time.
0:23:10 And both of those technologies
0:23:13 are basically irrelevant on the modern battlefield.
0:23:15 It’s very easy to jam GPS,
0:23:17 and comms is always contested.
0:23:19 And so, I think that those are representative
0:23:21 of their ability to build a bunch of drones
0:23:22 and get them into the air,
0:23:24 which is certainly part of the equation,
0:23:26 but from a core technology standpoint,
0:23:29 they’re less relevant for the things
0:23:32 that I think would be impactful on the battlefield.
0:23:34 And we have the great joy of competing
0:23:36 against the leading Chinese drone company, DJI,
0:23:39 in civilian markets that are unregulated,
0:23:40 where customers can buy everything.
0:23:42 And our biggest advantage competing against them
0:23:44 is AI and autonomy capability.
0:23:45 The stuff that we’ve been able to build into our drones
0:23:49 is far more advanced than what DJI has been able to do
0:23:50 in terms of being able to sense the environment
0:23:52 in real time, respond to it,
0:23:54 automate complex tasks and missions.
0:23:56 And the advantage that we have,
0:23:59 I think, is reflective of our strengths as a country.
0:24:02 We were talking about building on academic research.
0:24:04 The technology of Skydio is reflective
0:24:05 of a lot of smart investments
0:24:08 that the government has made over the years.
0:24:10 So, I think that being able to deploy
0:24:12 highly autonomous AI-driven drones at scale
0:24:15 is still a domain that we can win in,
0:24:17 and, in my view, is still up for grabs
0:24:19 and is something that we’re quite focused on as a company.
0:24:22 Right, and your company has also invested heavily
0:24:23 in autonomy.
0:24:25 Yeah, so our investments on autonomy,
0:24:28 kind of going back to one of my earlier statements,
0:24:29 is sort of predicated on this belief
0:24:31 that to make math effective,
0:24:33 it has to be intelligent, right?
0:24:34 That’s the difference between
0:24:37 grinded out, trenched situations
0:24:39 and being able to assert dominance in a space
0:24:43 is if you can find, fix, and finish targets at scale.
0:24:46 So, we just think that’s fundamentally important.
0:24:47 And the first 10 years of our journey
0:24:48 was defined by,
0:24:51 let’s strive to make the world’s best AI pilot
0:24:53 and climb the aviation food chain,
0:24:56 which culminated in us doing some work on F-16s
0:24:58 that has circulated the internet.
0:25:00 When we thought about the next 10 years,
0:25:01 the question was,
0:25:03 is the future really about Shield AI
0:25:04 building the world’s best AI pilot,
0:25:06 or is it about making a contribution
0:25:07 to the industrial base
0:25:10 so that everybody that’s building these systems
0:25:11 across America,
0:25:14 building sort of incredibly sophisticated machines,
0:25:16 striving to do it at larger scale,
0:25:19 can now deploy the best possible AI pilot
0:25:20 for their vehicles
0:25:22 for the customer’s missions.
0:25:24 And we think our best and highest contribution
0:25:26 is to enable the industrial base
0:25:29 to fast forward the large-scale deployment
0:25:32 of the world’s best AI pilots.
0:25:34 In Ukraine, you see drones come up,
0:25:35 drones come down,
0:25:36 oftentimes they’re DJI drones
0:25:38 in a matter of minutes or seconds,
0:25:39 many times failing in the mission.
0:25:42 How have you thought through
0:25:44 how your drones can operate
0:25:45 in a battlefield
0:25:46 where there is a high degree
0:25:47 of electronic warfare?
0:25:48 So, you do not have access to GPS,
0:25:50 you don’t have access to reliable signals
0:25:51 to the controller and operators.
0:25:53 When we started the company,
0:25:55 we made a huge bet on computer vision
0:25:58 as the right technology for AI and autonomy.
0:25:59 And this was back in 2014
0:26:01 when it was like much less clear
0:26:02 than it is today.
0:26:05 So, just sort of natively,
0:26:06 our drones have a bunch of cameras,
0:26:07 they look out and see the world,
0:26:09 they use that to figure out where they are
0:26:09 and how they’re moving
0:26:12 and what’s interesting and important around them.
0:26:13 We come from a place
0:26:15 of not being reliant on GPS,
0:26:16 having more of an ability
0:26:17 to do onboard things.
0:26:19 Having said that, candidly,
0:26:20 like our first round of drones
0:26:22 in Ukraine basically failed.
0:26:25 And we had built the system
0:26:28 largely informed by the U.S. Army’s requirements
0:26:30 for what they thought was important
0:26:31 for a quadcopter in this space.
0:26:32 And electronic warfare
0:26:34 was just nowhere on that list.
0:26:36 And so, from a radio standpoint
0:26:38 and from a navigation standpoint,
0:26:40 our first generation system
0:26:42 was just not set up for success.
0:26:44 And it was a painful process for us.
0:26:46 So, I’ve been to Ukraine twice myself.
0:26:48 We had a bunch of folks on our team
0:26:49 spend a bunch of time there
0:26:50 and we learned a bunch of hard lessons
0:26:51 about what it takes
0:26:53 to really operate in this environment.
0:26:56 And that really started to drive our development.
0:26:58 And more so, honestly,
0:27:00 than the requirements coming from the U.S. military,
0:27:02 we made an explicit decision as a company
0:27:05 that we think this is the real world situation
0:27:07 that matters both from an immediate impact standpoint,
0:27:09 but also whether or not they realize it,
0:27:11 this is what everybody else is going to need as well.
0:27:13 And so, there was a process for us
0:27:14 that really took a couple of years
0:27:18 of adapting the sort of native vision AI primitives
0:27:20 to work in this environment,
0:27:21 which we’ve now gotten to
0:27:23 with pretty phenomenal results
0:27:24 where the drone is incredibly resilient
0:27:26 and capable from both a comm standpoint
0:27:29 and from a GPS-denied navigation standpoint
0:27:31 in extremely harsh environments.
0:27:35 But it’s a real technology hurdle to get across.
0:27:36 And we’re now seeing this.
0:27:38 The bet that we made is in many ways paying off.
0:27:41 We’re seeing this with other militaries around the world.
0:27:42 They’re starting to come around
0:27:43 that this is the kind of thing that matters
0:27:45 and our systems are performing.
0:27:46 Now, electronic warfare testing
0:27:48 is becoming part of the evaluation protocol
0:27:49 for a lot of purchases.
0:27:51 I think Adam hit the nail on the head.
0:27:53 The effectiveness of your systems in an EW environment
0:27:56 is the difference between whether they’re relevant or not.
0:27:58 And a lot of people will say
0:28:00 that our next generation product
0:28:01 is going to work in an EW environment.
0:28:03 But I think that’s hard to say
0:28:05 unless you’re actually doing it.
0:28:06 I mean, that’s a pretty tall claim.
0:28:08 Another element of effectiveness
0:28:10 in these complex battlefields,
0:28:12 I think it’s just the adaptability
0:28:15 of your capabilities of your software.
0:28:17 So, anecdote from Ukraine,
0:28:19 to solve for this environment,
0:28:20 which is constantly changing,
0:28:21 we were going to go out.
0:28:22 We have a product called the V-Bad.
0:28:23 It’s a plane.
0:28:25 It’s 12 feet tall, 12-foot wingspan.
0:28:27 I can fly for about 12 hours.
0:28:30 We got sort of the brief from the Ukrainians
0:28:31 and the situation we expected.
0:28:32 Here’s the situation.
0:28:34 You guys are going to take off.
0:28:35 You’re going to go this way,
0:28:37 maybe 70, 80, 90 nautical miles.
0:28:39 You’ll have GPS on the ground.
0:28:40 As soon as you get to 200 feet,
0:28:41 the jammer’s going to hit you.
0:28:43 And you have to have no GPS
0:28:44 from that point forward.
0:28:46 So, engineer team’s like,
0:28:46 great, got it, good.
0:28:48 We’ll get the fix on the ground.
0:28:49 We’ll take off.
0:28:50 We’ll go take care of business.
0:28:50 It’s going to be sick.
0:28:52 We go out there,
0:28:53 and at three feet,
0:28:55 the airplane loses GPS.
0:28:57 And so, the Ukrainians
0:28:58 are constantly jamming to us.
0:28:59 The Russians, Ukrainians,
0:29:00 because the Ukrainians stop jamming,
0:29:02 they find themselves at risk.
0:29:02 Even if you’re launching
0:29:03 on the friendly side,
0:29:04 there’s intense jamming
0:29:05 at three feet off the ground.
0:29:08 So, the airplane takes off
0:29:09 and it just starts flying
0:29:10 the other direction.
0:29:11 I don’t know if you’ve seen
0:29:12 the movie Interstellar,
0:29:13 where he’s going through the cornfield,
0:29:14 the dude has a laptop,
0:29:15 and he brings down the drone.
0:29:18 Our team and the Ukrainians
0:29:19 just took off in a truck
0:29:21 and chased it for about two hours
0:29:22 and found it orbiting
0:29:23 over a cornfield
0:29:25 about, I think,
0:29:26 80 kilometers away.
0:29:29 And in the span of 24 hours,
0:29:30 the engineering team
0:29:31 re-architected the stack
0:29:33 to not use GPS
0:29:34 at any point in the mission.
0:29:36 We validated it
0:29:37 at our Texas facilities,
0:29:38 and then we pushed it forward,
0:29:39 and 24 hours later,
0:29:40 the team took off
0:29:41 and then conducted
0:29:42 a demonstration
0:29:43 that was written about
0:29:44 in the Wall Street Journal,
0:29:46 where the outcome of it
0:29:46 was ultimately
0:29:47 a Russian SA-11
0:29:48 getting found,
0:29:49 getting fixed,
0:29:49 and finished
0:29:50 by a high Mars.
0:29:52 I tell that anecdote
0:29:53 because it will be
0:29:55 the pace of deployment
0:29:56 that is the make
0:29:57 or break
0:29:58 for militaries
0:29:59 around the world.
0:30:00 If we want to make
0:30:01 a software change
0:30:02 in some of our programs,
0:30:03 it can take up to a year,
0:30:04 right?
0:30:05 it is considered
0:30:07 a big deal
0:30:08 to change the software.
0:30:10 When we took
0:30:11 that aircraft forward
0:30:11 to Ukraine,
0:30:12 we wrote new software
0:30:13 in 24 hours,
0:30:14 we pushed it,
0:30:15 and then we needed
0:30:16 24 hours
0:30:17 to properly plan
0:30:18 the operation.
0:30:19 And unless
0:30:21 we find a way
0:30:21 for industry
0:30:22 and government
0:30:23 to work together
0:30:24 to deploy
0:30:25 software capabilities
0:30:26 at that pace,
0:30:27 we will find ourselves
0:30:29 in a very tough
0:30:30 situation.
0:30:31 In a world
0:30:32 where you’re using
0:30:33 autonomous systems,
0:30:34 everything is basically
0:30:35 software-defined,
0:30:35 right?
0:30:36 The entire behavior
0:30:37 capability of the system
0:30:37 is software-defined.
0:30:39 Electronic warfare
0:30:40 is also software-defined.
0:30:41 The behavior
0:30:41 of the jammers
0:30:42 and everything
0:30:42 is coming largely
0:30:43 through software.
0:30:44 And so in many ways,
0:30:45 I think a modern conflict,
0:30:46 and you see this in Ukraine,
0:30:47 becomes basically
0:30:48 like a software writing fight.
0:30:49 And the speed
0:30:50 at which you can write it
0:30:51 and deploy it
0:30:51 really matters.
0:30:52 And I have no choice
0:30:53 but to be hopeful
0:30:54 and optimistic here
0:30:54 because we’ve had
0:30:55 the same experience
0:30:56 where sometimes
0:30:57 it’s taken us
0:30:57 two years
0:30:59 to push new software
0:31:01 into deployed systems.
0:31:02 I would like to think
0:31:02 that if we were
0:31:03 actually in a conflict,
0:31:04 that would just evaporate
0:31:05 and we could do it
0:31:06 in a day,
0:31:07 maybe that’s overly
0:31:08 optimistic and hopeful.
0:31:09 But I think that’s one
0:31:09 where the status quo
0:31:10 is unacceptable.
0:31:12 Two exceptional stories.
0:31:14 So we’re discussing autonomy.
0:31:16 And one of the concerns
0:31:18 either of you may hear
0:31:19 from time to time
0:31:20 is that,
0:31:20 well,
0:31:21 if you have autonomous drones,
0:31:22 that means humans
0:31:23 are not in the loop.
0:31:24 And that means
0:31:25 we’re ceding our authority
0:31:26 to essentially software.
0:31:28 is that an accurate assessment?
0:31:30 What does autonomy mean
0:31:31 in terms of humans
0:31:32 actually having control
0:31:34 over the end state
0:31:35 of what the drones
0:31:35 are doing?
0:31:37 And how do we best
0:31:38 configure the military
0:31:39 and also civil society
0:31:40 to a future
0:31:42 where autonomous drones
0:31:43 are not only available
0:31:43 but cheap,
0:31:44 widely adopted?
0:31:46 The first thing to say,
0:31:47 and this is really,
0:31:48 I think,
0:31:48 the important backdrop
0:31:49 for all of this
0:31:51 is that this is
0:31:52 terrible stuff, right?
0:31:52 I mean,
0:31:53 we’re talking about
0:31:53 weapon systems
0:31:54 that kill people
0:31:56 and create immense suffering
0:31:57 and that is the reality
0:31:58 of war.
0:31:59 The real goal
0:32:00 for everything
0:32:02 is to act as a deterrent
0:32:04 and to make conflict
0:32:04 less likely
0:32:05 and to make it such
0:32:07 that if conflict does happen,
0:32:08 you can be maximally targeted
0:32:08 and precise
0:32:10 and minimize human suffering.
0:32:11 I think that’s just
0:32:12 sort of like
0:32:13 an important backdrop
0:32:13 for all of this
0:32:14 when we’re thinking
0:32:15 about what we do
0:32:16 and the kinds of systems
0:32:17 that we’re building.
0:32:18 Now,
0:32:19 I think there’s a bunch
0:32:20 of legitimate concerns
0:32:21 about what does it mean
0:32:22 to have these AI-driven robots
0:32:23 and how automated
0:32:24 are they going to be
0:32:25 and how much authority
0:32:26 are we going to delegate to them?
0:32:27 I think the thing
0:32:28 that we have to also
0:32:29 keep in mind
0:32:29 is unfortunately
0:32:31 the game theory here
0:32:32 is just as simple
0:32:33 and obvious as it can be,
0:32:33 right?
0:32:34 Nobody thinks
0:32:35 nuclear weapons
0:32:36 are good for humanity
0:32:38 on an individual level.
0:32:39 Deploying a nuclear weapon
0:32:40 is like a miserable,
0:32:41 terrible, terrible thing,
0:32:43 but the only world
0:32:44 worse than one
0:32:44 where like you
0:32:45 and your adversary
0:32:46 have nuclear weapons
0:32:47 is one where only
0:32:48 your adversary does,
0:32:48 right?
0:32:49 So I think
0:32:50 one of our strengths
0:32:51 as a country
0:32:52 is our values
0:32:52 and the way
0:32:53 that we try
0:32:54 to conduct conflict
0:32:56 with a high ethical standard
0:32:57 and I think
0:32:57 that it’s important
0:32:58 to maintain that.
0:32:58 One of the things
0:32:59 that I’ve been impressed by
0:33:01 is really the level
0:33:01 of sophistication
0:33:02 within the military
0:33:03 on these issues.
0:33:03 I mean,
0:33:04 there’s people
0:33:05 whose job it is
0:33:06 to think deeply
0:33:07 about the implications
0:33:08 of different kinds
0:33:08 of weapon systems
0:33:09 and how authority
0:33:10 is delegated.
0:33:11 there’s very robust
0:33:12 controls in place
0:33:13 for how the military
0:33:14 thinks about these systems.
0:33:15 Those things are evolving
0:33:17 as the technology evolves,
0:33:18 but this is something
0:33:19 I’ve thought quite a bit
0:33:19 personally about.
0:33:20 The more that I’ve thought
0:33:20 about it,
0:33:21 the more time I’ve spent
0:33:21 with the military,
0:33:23 the U.S. military in particular,
0:33:24 the more comfortable
0:33:24 I’ve gotten
0:33:26 that this is a robust organization
0:33:27 that cares about
0:33:27 doing the right thing,
0:33:28 is thinking deeply
0:33:29 about the implications
0:33:30 of technology.
0:33:32 And my general view,
0:33:33 which I think is shared
0:33:34 by the U.S. military
0:33:35 policy and doctrine,
0:33:36 is that ultimately
0:33:36 human judgment
0:33:37 is really important.
0:33:39 A human exercising judgment
0:33:41 in how force should be used
0:33:42 is super important.
0:33:43 But the other thing
0:33:44 that people have to understand
0:33:45 is that the status quo
0:33:46 is not great.
0:33:46 Oftentimes,
0:33:47 your choice,
0:33:48 if you want to take out
0:33:48 a target,
0:33:49 is dropping a 500
0:33:51 or 2,000-pound bomb,
0:33:52 which is going to cause
0:33:53 widespread destruction
0:33:55 and a lot of collateral damage.
0:33:56 And so,
0:33:57 an AI system
0:33:58 that might be
0:33:59 using autonomy
0:34:01 to a pretty intense degree
0:34:01 to figure out
0:34:02 where something is
0:34:03 and what to do about it
0:34:04 is probably better
0:34:05 than dropping
0:34:07 a 2,000-pound bomb
0:34:07 and blowing up
0:34:08 a whole city block.
0:34:09 And so,
0:34:10 I think you’ve always
0:34:10 got to be thinking about
0:34:11 what’s the status quo
0:34:13 and can we use AI
0:34:14 and autonomy
0:34:15 to better,
0:34:16 more precisely,
0:34:17 accomplish the thing
0:34:18 that we care about
0:34:19 while inducing
0:34:20 as little human suffering
0:34:21 as possible.
0:34:23 Maybe the good
0:34:25 and true news
0:34:25 right now
0:34:26 is that
0:34:27 I think human-machine teams
0:34:28 are far more effective
0:34:29 than machine-only teams
0:34:30 right now
0:34:31 and for the next
0:34:32 several years
0:34:33 that’ll continue
0:34:33 to be true.
0:34:34 And maybe it’s true
0:34:35 for longer than that.
0:34:36 And the frameworks
0:34:37 that we have in place,
0:34:39 people are in the loop
0:34:40 on the decisions,
0:34:41 make a tremendous amount
0:34:41 of sense.
0:34:43 I think Adam brings out
0:34:43 an excellent point.
0:34:44 The game theory
0:34:45 is one that
0:34:46 if somebody finds
0:34:47 that a machine-only team
0:34:49 is the most effective team
0:34:50 in certain missions
0:34:50 and circumstances,
0:34:52 the question is
0:34:52 why and when
0:34:53 would that be used
0:34:54 and what do you do
0:34:54 about it?
0:34:55 And how do you make sure
0:34:56 that you’re ready for it?
0:34:57 And so I don’t think
0:34:58 you can live in a world
0:34:59 where you have blinders.
0:35:00 I was speaking
0:35:01 to a senior military leader
0:35:02 that had a nice framing,
0:35:03 which is his expectation
0:35:04 is the more defensive
0:35:06 you end up being,
0:35:07 the more likely you are
0:35:07 to turn things over
0:35:08 to machine control
0:35:10 to get the very fast reactions
0:35:11 and the dominance
0:35:11 that you need
0:35:12 to come out
0:35:13 of that situation.
0:35:14 And so a great,
0:35:15 very practical example
0:35:15 of that today
0:35:17 is the phalanx gun system
0:35:18 that protects ships, right?
0:35:19 If you come into
0:35:20 that thing’s weapon
0:35:20 engagement zone
0:35:21 and it’s turned on,
0:35:22 it will kill it, right?
0:35:23 And so you can think
0:35:24 about that
0:35:25 on a larger scale.
0:35:26 If a force is pressing
0:35:27 on another force,
0:35:28 they find themselves
0:35:29 in a defensive situation
0:35:30 and they flip the switch
0:35:31 and they go full auto,
0:35:33 what does that mean, right?
0:35:34 And how does that get
0:35:35 put back in the box?
0:35:35 And tactically,
0:35:36 what does that mean
0:35:36 for your forces?
0:35:37 Were they trained
0:35:38 to face something
0:35:39 that had that level
0:35:40 of capability
0:35:41 and that level
0:35:41 of discretion,
0:35:42 that level of speed?
0:35:43 And then how does it
0:35:44 play forward from there?
0:35:45 And so I think
0:35:45 that there are a lot
0:35:46 of hard questions.
0:35:47 I think that like
0:35:48 the convenient answer,
0:35:48 the easy answer
0:35:49 is that humans
0:35:50 are always going
0:35:50 to be on the loop.
0:35:51 It’s going to be fine.
0:35:52 Don’t worry about it.
0:35:53 But I think that
0:35:53 the world is a little bit
0:35:54 more complicated than that.
0:35:56 I don’t have answers for it
0:35:57 other than
0:35:58 a strong conviction
0:36:00 that America needs to lead.
0:36:01 So no matter
0:36:02 what you believe
0:36:03 about any of this,
0:36:04 whether you have conviction
0:36:05 on one side or the other
0:36:06 or you’re just uncertain
0:36:07 what the future
0:36:08 is going to be
0:36:08 and I think there’s
0:36:09 a lot of uncertainty
0:36:10 about how the future
0:36:11 will play out,
0:36:12 American leadership
0:36:12 is the answer.
0:36:14 At the conceptual level,
0:36:15 a lot of these things
0:36:16 are less new
0:36:17 than they seem.
0:36:18 So like dropping a bomb
0:36:19 in World War II,
0:36:19 like once that thing
0:36:20 leaves the bomber,
0:36:21 it’s autonomous.
0:36:22 It’s pretty dumb autonomy,
0:36:23 but human judgment
0:36:24 is over, right?
0:36:25 That thing is falling
0:36:26 and it’s guided
0:36:26 by gravity
0:36:27 and wind
0:36:27 and physics
0:36:28 and other things
0:36:30 and at some point
0:36:30 in that trajectory,
0:36:32 if it starts heading
0:36:32 towards the wrong place
0:36:33 or you realize
0:36:34 that it was the wrong target,
0:36:34 there’s nothing
0:36:35 you can do about it.
0:36:36 We are used
0:36:37 to relinquishing control
0:36:39 over the end outcome
0:36:40 and usually that results
0:36:41 in much less precision
0:36:42 and much less ability
0:36:44 to actually accomplish
0:36:44 the thing
0:36:45 that you care about
0:36:47 and AI fundamentally
0:36:48 changes that equation,
0:36:49 but I don’t think
0:36:50 that the concept
0:36:50 of the human
0:36:51 relinquishing control
0:36:53 over the ultimate
0:36:54 thing that happens
0:36:55 is actually new.
0:36:57 both of you
0:36:57 questioned the premise
0:36:59 that this is a new thing
0:37:00 that’s never happened before.
0:37:01 It is not the case.
0:37:01 And second,
0:37:02 both of you
0:37:02 zoomed out
0:37:03 and said,
0:37:04 listen,
0:37:05 we are not
0:37:06 at the end of history.
0:37:07 Fukuyama was wrong.
0:37:08 There are rival systems
0:37:09 of government,
0:37:10 totalitarian states
0:37:11 and free states
0:37:13 and we have to make sure
0:37:13 and we have to make sure
0:37:14 whether your concerns
0:37:16 may be in a conscientious,
0:37:17 thorough and democratic manner,
0:37:19 we ensure that we are the ones
0:37:19 who have dominated
0:37:20 this technology
0:37:21 to ensure deterrence.
0:37:22 Now let’s imagine
0:37:23 two different futures
0:37:24 10 years from now.
0:37:25 One,
0:37:27 where totalitarian states,
0:37:28 China,
0:37:28 Russia,
0:37:29 et cetera,
0:37:30 has dominance
0:37:31 over this technology.
0:37:32 They have not only
0:37:33 in terms of procurement,
0:37:33 in terms of production,
0:37:34 but also in terms of technology.
0:37:35 And another,
0:37:37 where we maintained
0:37:39 and then advanced
0:37:40 our technological advancement
0:37:41 and lead over them
0:37:42 and also achieved
0:37:43 manufacturing,
0:37:43 at least parity
0:37:44 if not superiority.
0:37:46 What do those two futures
0:37:47 look like?
0:37:49 It’s hard to predict exactly,
0:37:50 but I think AI
0:37:51 is the most important technology
0:37:52 really in the history
0:37:52 of humanity.
0:37:54 And a world
0:37:55 where our adversaries
0:37:57 have it and we don’t
0:37:57 is not a good one.
0:37:58 I mean,
0:37:58 a world where like
0:37:59 the Soviet Union
0:38:00 had nuclear weapons
0:38:01 and we didn’t.
0:38:02 A world where
0:38:02 Nazi Germany
0:38:03 had nuclear weapons
0:38:04 and we didn’t.
0:38:04 I mean,
0:38:05 these are not pleasant things
0:38:06 to think about.
0:38:06 And so,
0:38:07 in my standpoint,
0:38:08 I think we should just view that
0:38:10 as an unacceptable outcome,
0:38:12 like one that we cannot allow.
0:38:14 And some of this plays out
0:38:15 like beyond
0:38:16 the military domain,
0:38:17 but I think there is
0:38:18 in AI,
0:38:19 and I think this is changing rapidly
0:38:20 for the better
0:38:21 with the new administration,
0:38:23 there was a lot of talk
0:38:24 of like safety
0:38:24 and regulation
0:38:25 and we can’t do this
0:38:26 and we can’t have
0:38:26 this many parameters
0:38:27 and if you use more
0:38:28 than this much compute,
0:38:29 it’s not allowed.
0:38:30 If you’re sitting
0:38:31 in Beijing or Moscow,
0:38:32 I mean,
0:38:32 that’s just got to be
0:38:33 music to your ears,
0:38:33 right?
0:38:34 Of please,
0:38:34 yes,
0:38:35 slow down.
0:38:35 It’s not to say
0:38:36 that we shouldn’t be
0:38:36 thoughtful,
0:38:37 but like,
0:38:38 we’ve got to be real
0:38:39 about the game theory
0:38:40 dynamics here
0:38:41 and the implications
0:38:42 for this technology.
0:38:44 So, to close out,
0:38:45 there are, I’m sure,
0:38:46 a number of founders
0:38:48 watching online
0:38:49 and I’m sure many of them
0:38:50 would be interested
0:38:50 in getting into
0:38:51 public safety,
0:38:52 defense tech,
0:38:53 want to make a difference.
0:38:54 They see the mission,
0:38:56 they see the need.
0:38:57 Perhaps they’re scared,
0:38:57 they’re afraid,
0:38:58 they’re afraid of failure,
0:38:59 they don’t know how to start.
0:39:00 So, if you were to give
0:39:02 just one piece of advice
0:39:03 to these founders
0:39:03 that would be founders,
0:39:04 what would it be?
0:39:05 Number one,
0:39:07 the mission is worth it.
0:39:08 I don’t think that there is
0:39:10 a thing in the world
0:39:11 that you can care about
0:39:11 that doesn’t build
0:39:12 from a foundation
0:39:13 of security and stability.
0:39:15 And I wake up every day
0:39:18 just incredibly excited
0:39:19 and honored to have
0:39:19 the opportunity
0:39:20 to contribute
0:39:21 to something that I think
0:39:22 is so important.
0:39:23 The other thing I would say
0:39:23 is this stuff
0:39:25 is just extremely challenging
0:39:25 and I think that
0:39:26 especially if you’re
0:39:27 building hardware
0:39:27 and you’re serving
0:39:28 these critical industries
0:39:29 where the stakes
0:39:30 are really high,
0:39:31 you should just expect
0:39:32 it to be really hard.
0:39:33 And it’s hard
0:39:34 in different ways
0:39:35 at different points in time.
0:39:37 But at some level,
0:39:38 I think to be successful
0:39:39 at it over the long term,
0:39:40 you’ve got to kind of love that.
0:39:41 And I think probably
0:39:41 both of us,
0:39:42 different things
0:39:43 in our backgrounds, we do.
0:39:44 I love the challenge
0:39:45 of trying to solve
0:39:46 these difficult problems
0:39:47 and I think probably
0:39:48 both our companies
0:39:48 tend to attract people
0:39:50 who want to be really pushed
0:39:51 and challenged
0:39:53 and work on hard problems
0:39:54 and struggle with things
0:39:55 over days, months, years.
0:39:56 And that is definitely
0:39:57 part of the journey here.
0:39:59 If you want to get to something
0:40:00 that is going to have
0:40:00 real impact,
0:40:01 you know, you’ve got to
0:40:02 embrace the struggle.
0:40:03 100%.
0:40:04 Wise words.
0:40:07 Now, if you made it this far,
0:40:09 a reminder that this
0:40:10 was recorded live
0:40:11 at our third annual
0:40:12 American Dynamism Summit
0:40:14 in the heart of Washington, D.C.
0:40:15 And if you’d like to see
0:40:16 more exclusive content
0:40:17 from the summit,
0:40:18 head on over to
0:40:19 a16z.com
0:40:20 slash American
0:40:21 dash dynamism
0:40:22 dash summit
0:40:24 or you can click the link
0:40:25 in our description.