Motley Fool Money - Unit X and the Future of Defense

Episode Date: August 3, 2024

Within the Pentagon, there’s an elite unit dedicated to bringing Silicon Valley innovation to slow-moving Washington. It’s called Unit X.  Chris Kirchoff and Raj Shah are the two men who built that unit. They join Ricky Mulvey for a conversation on the changing defense landscape and what it’s like to bring a venture capital mindset to bureaucrats. They also discuss: What it’s like to disrupt Washington’s “primes” Supersonic drones, EVOTLs, and  How investors and lawmakers can distinguish “statistical techniques” from genuine AI Companies mentioned: NOC, BA, GOOG, GOOGL, JOBY, PLTR, TSM Host: Ricky Mulvey Guest: Christopher Kirchoff, Raj Shah Producer: Mary Long Engineer: Tim Sparks Learn more about your ad choices. Visit megaphone.fm/adchoices

This episode is brought to you by Indeed. Stop waiting around for the perfect candidate. Instead, use Indeed sponsored jobs to find the right people with the right skills fast. It's a simple way to make sure your listing is the first candidate C. According to Indeed data, sponsor jobs have four times more applicants than non-sponsored jobs. So go build your dream team today with Indeed. Get a $75 sponsor job credit at Indeed.com slash podcast. Terms and conditions apply.

It's been a real wake-up call to everybody that the old weapons platforms can no longer work in the way they were originally designed to work, that there is a whole new hybrid way of fighting war, and that we need to master it quickly if we're going to have military as dominant as we had before the conflict kicked off. I'm Mary Long, and that's Chris Kirkoff. He's an expert on emerging technologies who, along with venture capitalist Raj Shah, launched and led the Pentagon's Defense Innovation Unit. Also called Unit X, this group piloted the use of flying cars and microsatellites for military missions. Shaw and Kirkhoff's new book is called Unit X, how the Pentagon and Silicon Valley are transforming the future of war. It tells the story of how that elite unit came to be and takes a look at the cutting-edge technology

that's shaping modern warfare. Shaw and Kirkoff joined my colleague, Ricky Mulvey, for a conversation about what happens when the Pentagon and Silicon Valley come together to develop defense technology. They also discuss bringing a VC mindset to Washington, D.C., why the military is interested in evodals, and how to distinguish statistical techniques from genuine artificial intelligence.

Let's set the table because for those who are not in this world, they may think, oh, there is an innovation unit within the government in defense, and it's called DARPA, where they're creating all of these really cool technologies for the military to use. How is Unit X different from that? Well, Ricky DARPA is an incredible part of the government that has been, of course, a part of some of the most story developments in technology, not only inventing the Internet and global positioning systems, but also importantly for warfare stealth.

All of the technology or most of the technology that DARPA works with is experimental in some way. So it requires essentially a team of really advanced scientists and engineers that are pursuing sort of cutting-edge science, things that have never been done. before, audacious things. And the mission that Raj and I were given with Defense Innovation Unit, which is separate from DARPA, the sort of a sister organization, was to take a different approach, to look at technology that was already being created by the consumer technology market and that you could just kind of buy off the shelf without having to do a lot of bench and engineering work to, but that nevertheless would be really instrumental and

important to military missions. And an even more basic question, why? I wasn't the Pentagon just buying this already? They have a massive amount of money to spend, and it seems like they could sprinkle a little bit of that at some of these startups creating things like drones that they can bring into wars. Ricky, I think you have to look back historically, right? And the Pentagon was designed in its processes, you know, in the 50s and 60s when the things that the Pentagon needed, aircraft carriers, fighter jets, could only be made. made by what we now call the defense, industrial base, defense primes, really large companies like

Boeing and Lockheed. And if you wanted the best technology in the world, that's where you went, where you went to IBM, and you got massive mainframes, and they literally had the best technology in the world, right? If you wanted to be a cryptographer in the 60s, then you wanted access to supercomputers, you went to the government. And what had happened was there was a revolution in technology development. in the private sector, from the microchip to the problems of software, cloud, and all of these

things moved so quickly that it really caught the government and the Pentagon with their feet back on their heels. They could not imagine that a plucky group of technologists and engineers working out of a garage in a couple of years could build something that was better and faster than their billion-dollar programs. And so the DODs, the DOD, sort of has missed much of this revolution. And interestingly, Secretary Ash Carter in 2001 had written a seminal paper while he was Harvard basically saying, look, commercial technology is going to be decisive to prevent and win wars. We need to be part of it. And it wasn't until he was secretary, you know, 15 years later that he put together the Defense Innovation Unit. But just to give you a

tangible example in the cover of the book, we have an F-35 and an iPhone. The F-35's design was set and complete in 2001, and it reached full operational capability in 2016. The iPhone didn't even exist in 2001, and now we're on like version 16, and the iPhone today has about 100 times of processing power as our frontline fighter jet. It's just a different way of thinking about technology and imagining how fast things can move. You described facing these antibodies when you're essentially trying to bring new technology and faster, disrupting these large defense contractors, which have, are getting, you know, hundreds of millions of dollars in contracts from the Pentagon.

What was it like disrupting the primes? What was that process like? Well, maybe I'll start and then I'll turn it over to Chris. The real disruption and fights were actually internal to the government. It was within the Pentagon and it was within Congress, right? The first branch of government, they write the checks. They have the power of the purse. And so day two of our tenure in this organization, Secretary Carter is coming in on Air Force 2 and made a big splash.

We get a phone call from Congress that says, hey, we got some bad news. Your budget has been zeroized. Zeroized. I don't know what that means. but it sounds terrible. And it's worse than you think. Not only is our budget zero is done with prejudice, so funds can't be transferred to that inside the Pentagon.

So Chris and I get on a flight to D.C. And there we meet the Senate and House staffers that are in charge of this, right? There's about 100 professional staff members that control this $800 billion budget. And they were really upset with the Pentagon because one of the things Congress does is go on delegations around the world to visit companies and places.

And oftentimes the Pentagon, the Air Force, provides the air transportation. And for whatever reason, the Secretary of Carter's staff had denied air transport. And so this group of staffers were really upset. And they said, look, we know this is important to the secretary. We don't like the fact we didn't provide this transport. And I was like, well, Secretary doesn't work for me. I work for him. But it gives you insight into the type of things that we had to deal with.

plus the skepticism that places like the Valley could actually create technology that had military capability. Well, you know, the story is actually worse because, you know, so our first trip to Washington was supposed to be this beautiful moment, right, where we go and sit with the heads of each of the military services and the head acquisition officials in each military service to understand their needs and to really calibrate our priorities to make sure that our new organization would be serving them. And halfway through our flight to Washington, we get this email message on the United Airlines flight that we were flying on that says your government hotel reservation has been canceled

because your credit card no longer works. And so we quickly realized that the old office that had set up DAU that had been relieved of its command when we got announced out of spite had canceled our government credit cards. So here we are taking an Uber over to Capitol Hill to understand why it was that some staffers were trying to end our organization and having to pay out of her pocket. You know, it's the kind of small things like this that you don't forget that sometimes petty infighting, you know, can ruin your whole day. Yeah, some of it came from, what was it, you didn't have enough presence in the state of Indiana, which one of the, people in charge of the budget was really upset about. And I see the petty infighting, but did any of that pressure you think come from those large government contractors that are like, you know,

here's this scrappy group that's trying to get in the middle of the hundreds of millions of dollars that we're getting for these contracts? You know, it's hard to, hard to draw a straight line. So I don't think we have any direct evidence. But clearly, I think some of these entrenched interests were worried about what we were doing. And they had a really good lobbying game and knew and had the right relationships with these members. But I think you have to even step back, right? And the pettiness, of course, is there. And that's in any bureaucracy. But there was just some fundamental disagreement or understanding of how this modern technology world could solve problems, right? So if you're a acquisition officer or program manager in the Pentagon,

you know, you're probably not going to get fired if you choose Boeing to build the next airplane or fighter jet. But if you went to some plucky little startup and then it failed, you know, you're going to get a lot of career risk. So there's just a ton of conservatism inside there, plus the whole view, again, I say lack of imagination that how could a small team in the valley do better than this massive engineering team at a traditional company. And I think people forget that Google Maps, there was a team of like five people that built that. And the modern tools allowed people to move so much faster, particularly in building software. And so I don't think people are necessarily doing it bad.

They just, you know, these worlds had divided. In fact, when Secretary Carter came to Silicon Valley, he was the first sitting secretary. in 20 years to do so. One of the reasons, I think, is because the large defense contractors were running these tech projects. One you mentioned is Northrop Grumman, who had almost $750 million over 10 years for a tech overhaul, but they essentially produced nothing. So we've gone from the Pentagon spending.

Now to these defense, these defense conglomerates, why are they, to use a kind word, inefficient it running tech projects or bringing innovation to? the armed forces? Well, Ricky, it's important to take a, you know, a look at the economics and the incentive structure that these companies are forced to operate under. And, you know, because you, you can't kind of go on Amazon and price comparison shop for advanced military hardware like aircraft carriers or, you know, submarines, oftentimes there's a very small number of firms that have the capability to actually build those. And so once a contract is awarded to one of them, to ensure that there aren't cost overruns that are unfair to the taxpayer, that misused taxpayer funds.

The Pentagon has developed a very elaborate system of auditing and cost accounting. In fact, these companies have to have a unique bespoke system of financial management to be compliant with how the Pentagon does audits. And not only that, but they have to follow to the letter a very sophisticated set of requirements that the Panagon will lay out in the contract. And to go back to the F-35, if those requirements are frozen in time in the late 1990s and then the production contract isn't awarded to 2001, if you're still constructing, you know, that system in 2015 before making it fully operational, you're bound by history.

You're not like a private company in Silicon Valley able to be as agile. So there are reasons why cultures of cost overrun and scheduled delay have become the norm. This is part of the structure of the older system of production that was the goal to move beyond by introducing this new office in Silicon Valley Defense Innovation Unit. So your first project was a program that scheduled mid-air refuelings. And until that point, there was a command center in the Middle East where people were literally using hockey puck. on whiteboards to try to schedule these things. Something that's kind of interesting about these two worlds is that you have a world with

very conservative incentives and then the world of Silicon Valley where, Raj, I know you're a VC, if you hit a home run in one in 10 shots, that's really good. But when you're doing something in this position, you really have to, you can't miss on your first shot. So how did you decide upon that being your first project? And what was that pressure like for trying to essentially have a project that in Silicon Valley may have a lower chance of success, but you also know that it really can't fail for this project to continue?

Look, the military has very important missions, many of which are no fail. It's literally life and death. And so I think it's a little bit of misnobre to think that, you know, the Valley or this iterative approaches, I'm going to put $2 billion behind a company, come back in 10 years and see if it succeeded. There's multiple steps along the way. And I think, in fact, the method that the Pentagon was using, which is we're going to have this really big program,

this was to overhaul that combined Air Operation Center, which probably has 50 different individual applications, we're going to give it to one contractor, we're going to give them $700 billion, and we're going to see what they come back with in eight years. Well, that's very risky because you're now stuck to one performer and you won't know for a very long time, but it's working and it's hard to change.

Our approach was let's take an iterative approach. We'll take a little bit of money, in this case. I think it was just a one or two million dollars. We'll start with one little program. This is the tanker planning tool, which is, again, where should air refuelers be so fighter jets can get gas in the Middle East? And then if it works, we'll scale it.

So I actually think it takes less macro risk to do this iterative approach, but it's a whole new way of thinking about software. that coded the story, which is quite positive, is that at the end, the Air Force in particular really began to realize the value of software. And they built the first software factory. It's called Castle Run. It's based in Boston. Think 200 engineers in a valley style office. And the first commander of it was Colonel Enrique Odie, the person on our team that led this tanker planning refresh. There's a lot of positives in your story.

It was good for me to read just to honestly hear about people who are trying to make a positive change in the government to make our nation safer right now. So I really appreciated that. In the latter half of your book, though, you mentioned that essentially it's still, it's harder to build or it's harder to sell a drone to the Pentagon than it is to build an autonomous drone that can carry 300 pounds of cars. What's sort of the state of play for these companies now that are trying to sell tools and technology to the Pentagon? So it's progressed significantly from 10 years ago, right? There is a boom in venture investing in companies focused on national security. At last count, it was nearly $40 billion last year.

And I've never seen more entrepreneurs and technologists want to solve it. The group that Chris and I ran, the Defense Innovation Unit, this last year got a billion-dollar budget, which is, you know, real money even for the Pentagon. But the question now, Ricky, do you point out is, will this scale? If we believe that drones are the future of warfare, and I'll let Chris talk about our recent visit Ukraine here shortly, if we believe that's the future, should we spend 1%, 2%, 10% of the Pentagon's 8,000? billion budget on this. So I really think it's still a matter of scale for us to achieve this true integration. Yeah, so, Ricky, I mean, on the one hand, there is a lot more attention to this innovation ecosystem. But on the other hand, developments in war are moving quite fast. And if you just look at the battlefield in Ukraine, you know, we all know that Ukraine is the first drone war,

and we've seen it on both sides. But just, you know, a few weeks ago, there was a striking development where the Ukrainians had to pull back from the front, all 31 of the M1-A1 Abrams battle tanks that we had provided the Ukrainian military. So this is the most advanced battle tank in the world in our arsenal and the arsenal of all of our allies. And a quarter of the 31 tanks have been disabled or destroyed by Russia and Kamakazi drones. So, you know, when you see something like that, you realize that drone warfare is on the verge of ending a century. of mechanized warfare that began in the First World War. So, if you think about that, that means that the incredible investments that we, that NATO,

that our other allies have made in tanks, might now be a sunk cost. It could be that in modern warfare, tanks are not any longer survivable. So, yes, the ecosystem is surging forward. Yes, last year, there were, I think, 537 venture deals with new defense tech startups. but the end of mechanized warfare demands, I think, a much larger response. Raj, you're an F-16 pilot. And now that job has become even more dangerous with drones that can do these sort of swarm attacks. If you were starting out today, do you think you would still learn to be a fighter pilot or would you learn to be a drone operator?

It's a really good question, right? I think the future of aerial combat is going to change with, drones, the ability of drones to collaborate and work in teams. And again, these things that Chris described in Ukraine. You know, I think there's still a role for man fighters and maybe I'm part of that old crude problem. But, you know, there's still a role, but it's rapidly changing. And it's going to change in ways that I think, you know, are going to surprise us all as AI begins to make more and more tactical decisions. And we can leave with, humans as strategic decisions, particularly matters of life and death.

And, Ricky, there's a neat thread here, actually, from one of the early projects that we led a Defense Innovation Unit that actually involved a supersonic drone that was autonomous and driven in part by artificial intelligence to be a windman to a manned fighter aircraft. And in just a few short years, the Air Force has really decided to harness this new generation of autonomous flying drones, such to the point that the current Secretary of the Air Force, Frank Kendall, has announced his intention to acquire something like 10,000, what he calls combat collaborative aircraft.

So these are autonomous drones that can serve as windmands, that can swarm, that have incredible range, that can do surveillance, can also carry weapons payloads. And the contract for that program is down to two final performers. And they're both non-traditional performers or not the traditional defense primes. One is General Atomics and the other is Anderol, one of the companies that was started in part by early contracts from Defense Innovation Unit Experimental. And that's, I think, hopeful because towards the beginning of your book, a lot of soldiers were using cheaper sort of Chinese-created drones. And essentially the idea was that American drones were pretty much unusable.

And now it seems like that situation has changed for something that's increasingly important for national security. I want to get to AI in a sec. You talk about Jobi and Avotals, these electric airplanes that are sort of these Jetson-style aircraft, air taxis is how we normally think about them. But what are the defense applications for Avotals? What are you hoping from this technology? I can take a crack at it. This vertical takeoff and land technology allows us. you to move people and material where there is no runways. So it can really help get supplies

to troops in the field. The second piece is not having a human. So if they're autonomous, they can go into higher risk areas. We've had in the wars in the Middle East many tragic occurrences where helicopters full of people have been shot down. And so if we can take those humans out, it really lowers the risk to our men and women uniform. So I think there's a lot of really interesting applications of this beyond even just the air taxis here in crowded cities. Now let's talk about AI.

So towards the end, you basically say that now these entrenched contractors are trying to repackage old technology, marketed as new, and quote, they take statistical techniques and with a straight face sell them is artificial intelligence to officials who are none the wiser, end quote. I think there may be some investors who are none the wiser. I might have trouble knowing the difference between a statistical technique in true AI. What's that look like? So, Ricky, artificial intelligence is a technology that's being developed primarily in the private

sector, which is something new for the government, right? The government is in a sense used to situations where a place like DARPA will sometimes take the lead. And so the only way our nation will be able to harness the most effective and the most cutting-edge artificial intelligence will be to work with the very small number of firms in the private sector that are leading its development. And what's also, I'd say unique about this is this is a situation where the United States has a clear lead over all of our potential adversaries. The small number of frontier labs that are creating the most advanced AI models are largely here in San Francisco. So if you're a military strategist, that makes you realize that

we have a unique advantage as a nation. And it'll be up to us to harness it to build it into our military technology and therefore create a level of battlefield advantage that no one else will have. What's the line between AI analysis and human decision making in the military right now? I can take a crack at that. The Pentagon has released a policy on, on the use of artificial intelligence. And one of the key tenets is that for life and death decisions, there must be a human in the loop. And that we're not going to outsource those types of decisions

to machines or computers, which makes a lot of sense. And I think most of us would agree with. But I think it's going to get more challenging in the future if our adversaries or potential adversaries don't have those same scruples and ethics. And if they do have automated decision-making, How do we respond to that? How do we make sure our decisions move fast enough to ensure that we still have the upper hand,

yet not move away from our ethical view of things? So I think these are tricky questions that will continue to unfold in the coming years. A lot of investors are enthusiastic about Palantir, which is a company you also write about. And that's one that's using an American company, using AI to help the military make more informed decisions where it's like tracking different ships and then the options you could use from that, let's say if a ship disappears, whether or not you try to track it with a satellite, if you put in an unmanned drone, that kind of thing. You all see it from the defense side. Is there a lot of enthusiasm about their software applications?

I think in general you're seeing enthusiasm for newer entrance. Palchre being one and there's a whole slew of other companies now focus on the national security customer base. And, you know, we started this discussion with the topic of that command center in Qatar. And I tell you, going into that center, at least a few years ago, it was like walking back into the 1980s, right, to see how technology used to be employed, right? Microsoft Excel and Merckchat for those that are. familiar with that, where the, which is like a text-based chat box, it was how we were doing operations. So I think there's a real desire for this generation that's grown up on the iPhone

to have modern software and the same quality of tools. So these new companies are getting a good reception and I'm optimistic that that will continue to scale. One thing that I'm afraid of is nuclear escalation. And I know it's something y'all think a lot about. And one of the one of the things you're right about is basically today it's Ukraine, and then next it might be China. And especially with their military exercises in the South China Sea, looking potentially to invade Taiwan. I've also heard the take that TSM, Taiwan Semiconductor, which produces the majority of the world's advanced chips, is a deterrent for that invasion.

Do you think that's a real deterrent or do you think that despite that we could still see that kind of escalation? Well, Rake, I think your question really points out the highest priority needs to be avoiding great power war, you know, war between sophisticated nations that have sophisticated militaries that would be incredibly violent and destructive in a way that we haven't seen in several generations. I mean, we're a long time away from anyone who's actually lived through great power conflict. So I think our highest goal needs to be to avoid it.

And one of the ways to avoid it is to come up with a set of technology that is so impressively powerful in the battlefield that you actually deter war from starting in the first place. And then so we can end this conversation, hopefully on a more optimistic note. Are there any defense tech trends or just, yeah, trends in keeping America secure that you're keeping an eye on that you're particularly optimistic about? I think there's a lot. First and foremost, these are human endeavors, building companies. companies, running the Pentagon, fighting wars. And so we really need our best and brightest working on these problem sets. And the level of interest that, you know, great engineers have in this space, the realization that America, while not perfect, you know, is the world's

oldest democracy. And democracy is worth preserving over an autocracy, as we've seen, you know, so blatant in Ukraine, is leading people to work on this. So I think that gives me the great this amount of optimism, is seeing talent that wants to build technology to keep our troops or our country safe rather than just doing photo sharing. Chris, anything to add it up before we go? Yeah, I think Ukraine is something actually that as tragic as it's been as a conflict, and, you know, Raj and I got a chance to visit and see the human toll up close. It's been a real wake-up call to everybody that the old weapons platforms can no longer

work in the way they were originally designed to work, that there is a whole new hybrid way. of fighting war and that we need to master it quickly if we're going to have a military as dominant as we had before the conflict kicked off. Christopher Kirkoff, Raj Shah. Thank you for your service to our country and thank you for joining us on Motley Fool Money. I appreciate your time and your insight. Thanks, Ricky. Thanks, Vicki. As always, people on the program may have interest in the stocks they talk about, and the Motley Fool may have formal recommendations for or against, so don't buy or sell stocks based solely on what you hear. I'm Mary Long. Thanks for listening. We'll see you tomorrow.