Odd Lots - This Is Palantir's Vision for Changing How the US Does Defense Spending

Episode Date: February 6, 2025

This week, we saw shares of software maker Palantir soar after reporting earnings and describing demand for AI as being "untamed." But what does Palantir do, and what do they actually sell to the federal government? More importantly, how does the government actually spend money on things related to security and national defense? Last year, Palantir's CTO Shyam Sankar published a document called "The Defense Reformation," containing 18 theses about arms procurement and how it should be changed. We talk to him about the history of US defense spending, and his vision for creating a more competitive, advanced environment for the US defense sector. Read More:Palantir Jumps Most in Year As ‘Untamed’ Demand Bolsters OutlookTrump’s Gaza Takeover Gets Israeli Approval, Global Scorn Only Bloomberg - Business News, Stock Markets, Finance, Breaking & World News subscribers can get the Odd Lots newsletter in their inbox each week, plus unlimited access to the site and app. Subscribe at  bloomberg.com/subscriptions/oddlotsSee omnystudio.com/listener for privacy information.

Bloomberg Audio Studios, Podcasts Radio News. Hello and welcome to another episode of the Odd Lots podcast. I'm Joe Wisenthal. And I'm Tracy Allaway. Tracy, you know my Twitter handle at the store. I'm vaguely familiar with it, yes. I was mentioned it. Do you know where I got it from?

Oh, that's a good question. I know you had it when you first started writing about the market. And I know you've said at various points in time that you kind of regret using it. Is that right? When I joined Twitter in like 2008, and like I didn't think it was going to be like a very big deal for any reason. It was just the name of my old blog at the time. And that was the name when we, you know, people just did like screen names. It was just the name of a blog.

The name means nothing to me. But before I was ever a journalist or even thought that I would get into journalism professionally, you know, I had this blog, the stalwart. You know, if you thought Twitter wasn't going to be that serious, I feel like you could have chosen a much worse name than the stalwart. But anyway, I had this blog, the stalwart.com. It was actually I had a co-author on it who went to do great things. Anyway, you know how like we sometimes tweet is like, hey, we're going to be in some city who wants to get a drink. I was visiting San Francisco in like 2006 and I posted a thing. And I think I had about 200 readers of my blog.

And I said, does anyone want to get a dim sum in San Francisco? And I got two people responded and show up. That's, I mean, that's not a terrible hit rate. No, it's not bad. Yeah, that's right. It's a not bad at all. But anyway, one of those people who actually showed up to get dim sum and a few friends of mine showed up is going to be coming on the podcast today. One of those two.

See, it's about quality, not quantity, Jay. That's right. The two people that showed up are useful to you. That's right. So I went on, I stayed in media, and I did the blogging and podcasting and journalism thing. And another became the CTO of one of the hottest companies in the entire world. And so, you know, some interesting people at that line. in a Chinatown in San Francisco. What did you tell him over Dim sum? I don't remember. Anyway,

this is a very sort of inward-facing intro, but to sort of zoom forward, you know, we've wanted to do more on defense in the year 2025 and how we spend money, et cetera. Anyway, and it turns out that this person, this guy who came to Dim sum in 2006 is a great person to talk about defense because he recently wrote a piece called The Defense Reformation, 18 Theses about how we currently spend money on defense in this country and why the way we do it is insufficient for this new world, especially with all of these geopolitical tensions. Of course, he works for a company that might benefit from a change in how we do spending on defense.

We're going to be speaking with Shamsankar. He is the CTO of Palantir, one of the hottest companies in the world, but which I'm still trying to wrap my head around exactly what they do, but we're maybe going to learn that and how they or Sham specifically envisions changing how we do defense and how Palantir could play a role in it. So, Sham, thank you so much for coming on Odd Lots. It's nice to talk to you again after almost 20 years. It's a pleasure to be on. Yeah, it's great to hear you again. Okay, here's a question. So, like, I watched these videos on YouTube about Palantir products, and it looks very like sort of like minority report

where it's like these like big walls of screens and they like zoom in on something and they're like oh and here is an asset in the South China Sea and why are these ships here and like someone's like looking up is that really what you guys sell? Like is it really this minority report? I know it's not minority report like pre-crime

but these sort of like visualizations like give me like the sort of like what does Palantir actually sell? Well the cynical way to think about Palantir is that it took something as sexy as James Bond to motivate engineers to work on a problem as boring as data integration. But I do think that is the seminal problem. It's not that we don't have this information,

it's that we don't have it in a way that we can easily access it to make a decision. I like saying, maybe 10 years ago, the idea that data was the new oil became very popularized. I strongly disagree with that. I think data is the new snake oil. It's not inherently valuable.

What really matters are the decisions we as humans make, and am I making a better decision? Am I getting into a new snake oil? continuous learning loop of how I will make that decision better tomorrow. That is something that Colonel John Boyd would recognize as the Uda Loop, the Observe, Orient, Decide Act, a piece of tradecraft he developed to beat mig fighters in dogfighting. But it's really like the human decision-making process and how we see ourselves and win as a consequence of being better at it than our competitors in the commercial sector or our adversaries in the military space.

So Joe mentioned that we want to do more on defense. And it is a difficult topic. to sort of wrap your head around because the defense industry is sprawling in many ways. So neither of us know that much about it. But one thing I do know is defense is an industry that is characterized by monopsony. Oh, do we still have the monopsony, Claxton? Yeah, we're going to put a sound defect here. Monopsony claxon. Okay, monopsony.

So there's one big buyer for defense equipment and services, which is the government. And if you are a defense contractor, you are basically only pitching to the government. If you fail to convince the government, it's not like you can turn around and sell missiles to the private market, at least hopefully not. How did Palantir actually break into this particular space? And why did you decide to do it? Because, you know, it seems like a pretty binary bet in some respects. I mean, first of all, I love the fact that people listening to this podcast know what monopsie is. My core thesis is that is the root of what ails us in terms of defense.

But you're exactly right. It is what makes this market so challenging. I mean, look, no great company is principally created to make money. You have a mission, a purpose. We wanted to work on hard problems and national security. When we came together to work on these problems, it was before there was like a political tinge to working on defense. It was actually people just thought we were wasting our life.

Like maybe you should sell software to banks. Wouldn't that be a better business? later, maybe 10 years after that, it became somehow highly unpopular and taboo to work on problems of national security. And now it's become in vogue, which I think is great. It's really a call back to a prior era. The birth of Silicon Valley was absolutely the base program and defense and missiles. Lockheed made those missiles in Sunnyvale in the heart of Silicon Valley, who was the largest employer

in the 50s in the Silicon Valley. So to your question of how did we break in, very painfully is the honest answer, that all of our adoption was driven from the field backwards. The monopsony is strongest in DC. It's where the opinion of a single buyer really matters. But the closer you get to soldiers in harm's way, people who are actually doing things, the more you start to approximate something looks like a market. People have heterogeneous ideas of what they need. They're able to express their opinion of this is better than that. They may not have the resource or the buying power to actually affect that economically. But it starts with the very beginning of a demand signal that actually there is something better.

else that we would like out there. And then you're doing battle with the bureaucracy every day. When we first started the business, I thought our competition would be the existing defense players. I could not have been more wrong. Our competition was the program of record, as they call it. It was the government program that thought it had a monopoly on providing a capability in a certain area. And if you challenge that monopoly in any way, they view it as their job to squish you, to crush you. And in this system that we've built up, the defense primes are just essentially yes men to the monopsonists. You know, their job is to do whatever the monopsonists asked for. And I think this is an important phenomenon because it's not how it always was. We can look at the

present moment, which is very much an outgrowth of having won the Cold War. It's something we've built up without pure competitors as a nation since the 90s. But when we were at our best in World War II in the early Cold War, it looked very different. Our services, competed against each other all the time. Today we think about it as the Air Force's Minuteman as the Intercontinental Ballistic Missile, but the Army and the Navy, they put in their own competing effort. It's really that the Minuteman won that great free market competition for having the best product. When we were building submarine launch ballistic missiles, Admiral Raybourne had four competing programs running concurrently. And so the consequence of the monopsony is not that

there isn't competition within the industrial base. It's actually that there isn't competition within government. How did that work? It's not intuitive to me. And like, you know, you say there's not competition within government. Even within many large corporations, even in the private sector, you don't often see that phenomenon in which, say, different offices within a company are competing against each other in terms of what they're buying and who bought the better technology and then wins out. I'm sure it happens to some extent, but by and large, even at a private corporation, procurement decisions are corporate-wide for certain types of products. There's like, you know, there might be a Teams company and there might be a Slack company. There might be a company that

uses AWS and there might be a company that uses Google Cloud, etc. Talk more about how, as you say, at least at one point, there could be within the defense overall competition over technology development. And these things are dialectical, where you can think about it as, like, if we were always duplicating efforts and competing, that somehow seems really inefficient on the other extreme, which is where we tend to find ourselves. If duplication is viewed as taboo and inherently bad and something that we always need unitary efforts around, that's also really bad. Now, these modalities skewed towards, so if you're in this sort of, hey, we should have a unitary effort, we should all just pick Slack or teams, something like that. That makes a lot of sense for something that's in the one-to-end phase.

It's functionally a commodity. There's value to standardization. You're not expecting a lot of innovation. On the other end, where you're dealing with zero to one problem, where there's fundamental uncertainty, a lot of innovation is needed. Innovation is fundamentally messy and chaotic. And trying to wrap it with all this process just means that you don't get the innovation at all.

Then the lack of competition and the lack of competing ideas, heterodox approaches, means that you lose. I think one of the most important responsibilities of a leader is to grant and revoke monopolies within your institution. Look, I'm calling the shot. I'm saying this is the blessed way. I have high conviction and the people and the approach and we're all moving out. We're putting all of our wood behind this arrow. On the other end, it's like there's fundamental uncertainty here. I'd like to fund three teams to go after this in different ways. And I will have to manage that as we go through to figure out what is the answer going to be. Okay. So on this note, as you point out, there's a lack of

competition within the government, maybe a lack of creativity. There's a tendency to be very bureaucratic. And of course, one of the things we hear a lot is that the government can be very slow in its processes, and it's not very good at, like, changing course when that is needed because there's always the sprawling infinite number of stakeholders and all these boxes that you need to check and things like that. And all of that combined, I imagine, can make for a somewhat frustrating customer if you're a defense company. I mean, if you're a defense company, the government is your customer. The government is certainly Palantir's customer. How could they be basically a better customer? By having more buyers. You know, how can we start to approximate something that looks closer to the free market?

So I think defense is one of these interesting places where it's the only place I know of where we disaggregate supply and demand. You think about that, the integration of supply and demand, the sales and operation planning process. It's the beating heart of most companies or institutions. In defense, you have the services, the Air Force, the Navy, the Marines, et cetera, they're responsible for manning, training, and equipping. This is the supply side. They produce the things that people are going to need. On the demand side, you have what we call the combatant commanders.

We have 13 geographic and functional combatant commands. They actually do the war fighting. If something happens in China, it is the Indo-Pacific command that will have to do that fighting, and they will use the supply that's been provided by the services to do that. But they don't really have a say in what is being produced, what the supply would be, what their opinions are. I mean, it's a very indirect sort of say. And I think this disaggregation ends up being one of the major bugs in our system.

So you have 13 buyers, essentially. Now, I'm not sure you should ask these folks who have three to five-year horizons to go build you an aircraft carrier. But I do think 95% of the things that we need to fight and win fall within a time horizon envelope they're going to care about and where their feedback and they're heterogeneous. needs. I'm not sure why each of them should need exactly the same thing. There's so much alpha to be had by capturing what is different about them and how you make that work for each of them rather than averaging all their requirements out for the sake of efficiency, but at a loss

of effectiveness. The U.S. hasn't fought a war against a really powerful adversary in a long time. There are certainly wars that we've fought that have been frustrating outcomes from the U.S. perspective, but not necessarily because we were technically outmatched. I understand, like, you make these points, and obviously you make these points in part because you think it will benefit your company. How do you establish that we're not doing it well? It seems like we still have an incredibly powerful military. What is the evidence that there's actually something wrong with the existing system? Well, I think we need to look at the world around us here. So I'd say that empirically, we've lost deterrence as the West. You know, if you go back to roughly 2014, Crimea was annexed by the

Russians, the Spratly Islands were militarized, Iran got shockingly close to getting the bomb, and certain policy enabled them to get and continue to get much closer. I think that a decade ago, we lost deterrence, and that's really accelerated. We've seen that we've had a pogrom in Israel. We now have North Korean soldiers dying in Ukraine, and we have unprecedented phase zero grave zone operations in the South China Sea, not to mention suddenly undersea, tables keep getting cut and snapping. So there's this sense in which people aren't afraid of us anymore. And that the whole point of defense is not to fight the war. It's to be so strong that you're deterring conflict. And I feel like... But just to push back, yeah, but just to push

back on this for a second, like, how do you establish that this lack of deterrence, as you perceive it, has something to do with where we are technologically along the state of the art versus domestic U.S. politics in the appetite after the Iraq war, after Afghanistan, et cetera, the sort of like appetite to be active and involved in all of these different places. Because we're talking technology, we're talking weapon systems, but another aspect of the sort of perceived lack of deterrence could just be about American appetite to exert our might. I do think it's multifactorial. I don't think I would discount that to zero there. But let's think about this just in terms of the cost of air defense. If you're using a $2 million missile to

shoot down a $2,000, that'd be generous. It's more like a $200 drone. You have a problem. And if it takes you two years to make each one of those missiles, you have a bigger problem. And so I think our adversaries can sense the ways in which we have, you know, and that then feeds into political will. I'm not sure these things are entirely disaggregatable. Since Joe brought up political will or political appetite, this is one of the other things I wanted to ask you. But, you know, if you talk about the government may be changing its ways, being more experimental in the way it does things, encouraging competition, there's a risk involved in that, right? Which is at some point, if something doesn't work, then someone in Congress is going to have to stand up and explain why the government, quote, wasted. I don't know, $10 billion on a program that didn't actually work out. How do you actually like go about, I guess, encouraging like the culture or changing the optics around what constitutes success and failure? I think you put your hand right on the issue here, which is how did we get here?

You know, if we go back to the early Cold War, our industrial base looked very different. It was an American industrial base, not a defense industrial base. Chrysler made cars and missiles. Ford made satellites until 1990, not that long ago. General Mills, the cereal company, used everything they learned about processing cereal in their mechanics division to make torpedoes, artillery, and inertial guidance systems.

Every camera, car, and cereal box Americans bought was actually subsidizing our national security. So the breadth of the investments, the commercial R&D investments we were making, we could actually use to have price performance and capability improvements in defense. And that really changed, it changed very, slowly in the 70s and 80s because actually the monopsinist was not a great customer.

People realized we should try to sell off these defense businesses. It's too bureaucratic. It's too slow. Then it changed all at once. In 1993, there was this famous dinner at the Pentagon called The Last Supper. We had won the Cold War. The country wanted a peace dividend. We slashed defense spending 67%.

So for every dollar we were spending before, we were only going to spend 33 cents. And this led to a massive consolidation in defense. We went from 51 prime contract. down to five today. And the consequence most folks see in this is that this consolidation led to a lack of competition inside the industrial base, less choices. I think that happened, but I don't think that's the high order bit. The high order bit is that this consolidation bred conformity.

This was the real beginning of the financialization of defense, of founderless companies that were principally run for the shareholder. They really thought about the return on capital, what's happening this quarter. and that plays into the hands of the monopsonist. You're not going to behave like Elon. You don't have the grand vision of getting to Mars and as a consequence of doing so, reducing launch costs by 90%.

And that's not what it was, because we think about it as Northrop Grumman today, but it was Jack Northrop and Leroy Grumman. Lockheed Martin was the Lockheed brothers, and separately, it was Glenn Martin. We had a very different kind of structure of founder-driven American capitalism that meant that people were investing

in big ideas with bold ambitions that is, I think, really important to understanding how we've kind of... So when the Berlin Wall fell, only 6% of major weapon system spending went to defense specialists. The rest went to what I call as dual-purpose companies, the Chrysler's of the world, the General Mills of the world. Today, that number is 86%. So we have taken our defense industrial base. We've put them on the Galapagos Island. They're serving this one customer, the monopsonists.

They're largely deprived of market signal from the competitive and dynamic commercial markets. and they've evolved in certain ways. And if you bring them to the mainland, you know, the wolves will eat the giant tortoises. And that's not what the Chinese industrial base looks like. Chinese primes only get 27% of their revenue from the PLA. The rest comes from us as silly patsies buying cheap crap on Amazon. It comes from commercial sales. So every now camera, cereal box, and car we're buying might be subsidizing lethality against U.S. soldiers.

I get this point. But to me, this also speaks to a different dimension, because if this is the story, then it can't really just be about decisions made at the Pentagon to winnow down the number of official suppliers because this is really an economy-wide story, right? Setting aside defense, you know, a company like General Mills is in 2025 not going to devote much of its internal attention to, you know, milling machinery, which is I see on Google is what got transformed into a war plan.

That would be some other company that they would contract out. Like this phenomenon, you call it financialization or specialization or prioritizing shareholders, being good at one thing that you sort of like optimize. This is an economy-wide phenomenon that there must be some root of it that can't be explained in a change of defense policy because we truly see it across the board. Lots of U.S. great U.S. companies, In Vida, for example, example, choosing not to focus on the low margin manufacturing aspect of semiconductors. This is an economy-wide phenomenon where companies shed these capabilities.

I think that's right. And you don't need everyone to be participating in it. But you could look at, you know, Powner was the first defense company to enter the S&P 546 years, not from mergers or spinouts. And I think that looks much more like the sclerotic European capital markets than dynamic U.S. capital markets. I think that tells you something about the rate at which we are creating new ideas, new companies in the innovation that's happening here. Now, a little bit more directly to Tracy's point on how do we avoid, you know, getting hauled in front of Congress? I'm not sure you avoid that, but I think the risk aversion plays. So we built the Pentagon in 16 months. Kelly Johnson, the founder of Skunkworks at Lockheed, he built the U2 spy plane in 13 months.

You know, he built 41 airframes in his entire career, which today you're lucky if you work on a third of an airframe in your career. So there's this sense in which we did have these really great people. We bet on them. Not everything they did worked out. I love this story of Admiral Raybourne, who is building the submarine-launched ballistic missile, the Polaris. You know, Congress is breathing down his neck.

Other admirals were breathing down his neck. Hey, when is it going to work? Give me the Gant chart of Gant-chard. Show me the project plan. Are you managing this? We're spending a lot of money on this. And he came up with the latest in decision analysis techniques. He used Monte Carlo simulation.

It's called P-E-R-T. you can research this, and it produced the Gant chart of Gant charts. And Congress was happy, and the admirals were happy, but it was all a smokescreen. What he actually did, he created this total fiction of Gant charts so that he could protect the engineers so that they could actually just do the work. He didn't know when it was going to work or how it was going to work, but this is the reality of zero to one, that actually you're doing this messy, innovative process. And if you try to wrap it in so much process that you eliminate all the risks, you can't

cut off one end of the tail of distribution of outcomes, like all the bad things, without cutting off the good end, too. So you're locking yourself into a system that's driving mediocrity. And one of my favorite examples to illustrate this is David Packard, co-founder of HP. He was the Deputy Secretary of Defense at one period of time in the 70s. He came up with a simple set of rules to improve acquisition called the DoD 5,000 series. Seven pages, pretty small. today that 5,000 series has turned into 2,000 pages. Wow. That is an 11% compounded growth rate of barnacles of bureaucracy.

People love to look at the 5,000 series and they complain, look what Packer did to us. But we did this to ourselves. And we do this because when anything goes wrong, we need to come up with a new rule how it will never go wrong again, which all we're really doing is coming up with rules to make sure nothing can go right again. I'm a sucker for alliteration, but barnacles of bureaucracies is a nice one. Just to drill down on this point, Palantir has both military and commercial clients. Can you walk us through what the differences are in as much granularity as possible between designing a product for a commercial client versus designing something for the government? Like, how do those contracts actually come into being and how do they differ from each other?

Well, one of our heterodox takes, which you can see in the line of argument I'm doing here is that it shouldn't be that different. And what has made it hard for us is we're not designing what we do to spec that either the commercial company or the government are asking us. We have our own opinion of what ought to exist to solve the problem. If you don't think it solves the problem, you shouldn't buy it, but allow us to pursue our opinion, leveraging all the stimulus that we have out there in the market and to deliver the innovation to you as a consequence of doing so. And, you know, we would think of this is so simple. You go to the Apple store and if you like the iPhone, you buy it. If you don't like the iPhone, you don't.

you don't buy it. And of course, that's not how the government likes to think of it. They want to come up with a requirements document that lists out everything they think they're going to need and then have essentially what is a fiction writing contest for you to respond to that and say whether you can build those things and at what price you can build those things. And then through that process, you're making a decision of who's going to actually then go build those things. And I think you're transferring all the risk to the taxpayer by doing so. The taxpayer has to fund the development. The taxpayer is betting that the government got all the requirements right. And the history would suggest that's not a great bet.

You know, there's a reason we believe in the free market. Instead, why wouldn't you transfer the risk to a private company? Let me build it on balance sheet. If you don't like it, I go out of business. But if you do like it, there's opportunity there. And I was able to do that with much less risk to the taxpayer. But just to play devil's advocate for a second, there are reasons that the government might have specific requirements that a commercial company doesn't.

You know, the government might have more stringent rules around something like data. privacy, something like that. Is it really possible to standardize products across commercial businesses that might be sensitive in one way or another, but military operations that are extremely sensitive? I mean, I would say we're in existence proof that's possible. That's like essentially my technical risk to manage of like, where do I enable the government to plug in things that they need to build a plug in that are in fact government unique? But how do I have the greatest possible common baseline. One of my favorite examples of this is Operation Warp Speed and the supply chain for the vaccine. We were able to build that supply chain in two weeks for General Perna because we had

built a similar digital twin for BP to optimize hydrocarbon production two years prior. There's no way we could have met the moment of need for the nation had we not had this broad, diverse commercial stimulus across 50 different industries. Talk to us about what specifically you sell because you're not selling missiles, et cetera. You're a software company. I also know, I think it, you probably worked with them for a while, but obviously at the end of last year, there was some announcement where you're working with Andrew, Palmer Lucky's company. What in your vision is the DOD buying more from you years from now than they are today? And talk to us about like some of like what the, what the pitch is.

In the commercial world, you call it a value chain from the hand of your supplier to the hand of your customer. But you can think of this as a series of decisions that create dependencies on the ones that are downstream. So, you know, you go from procuring things to producing things to pricing those things with customers. Simple example of the value. In the military, you call that a kill chain, from sensor to shooter. But it's literally the same thing.

It's a decision chain, maybe if you thought about it more generally. And every one of those decisions, you want to make faster than your competitors, and you want to make them better than your competitors. And the way to do that is with AI. So we provide an AI decision-making platform that spans the breadth of your enterprise that allows you to get better and better, faster and faster. And so we don't make the hardware, but we make the intelligence. You're talking about on the battlefield. You're talking about software that when you say make a decision better than your competitor. It means you're selling software that you say

on the battlefield allows some entity to fire better than their competitor. That's right. Or in the commercial world allows Airbus to build aircraft more effectively or Chrysler to build cars better or HD Hyundai to build ships better. There's one other thing I wanted to ask. you. I guess this is very top of mind for everyone at the moment, but the Trump administration, there's been a lot of rhetoric from Trump about shrinking the military industrial complex, which I presume, you know, there aren't a lot of specifics here, but I presume it means eliminating waste and also reducing the overall amount of money that's allocated to the military in the form of government subsidies or whatever else. How does that square with your vision?

of revolutionizing the Defense Department? I think it's pretty congruent. I mean, if you thought about the $880 billion that we're spending a year, what percentage of it do you think we're spending efficaciously to deter our adversaries? We could all come up with different numbers, but largely people would agree there are a lot of legacy things we're spending them on. Some of those are because Congress wants to protect certain programs. Some of them are because the services want to protect certain programs. Some of them are just because we haven't yet realized they're not going to meet their moment.

So I think we could be substantially more efficient. We should be thinking about this as an acid allocation problem. Like, of the capital we have, how is it allocated against the problems and the threats that we face? And is it kind of the same allocation it was 30 years ago? How much of that have we retooled and re-geared towards the future fight that we're likely to have? You know, we did an episode recently, someone we've had on the podcast a couple of times, Jennifer Palka, talk about software. And, you know, we were talking about Doge. And this idea that, like, it's one thing to identify a problem of government procurement or government operation.

It's another thing to actually do it. And doing it is probably some combination of changing rules that are written down somewhere, maybe in Congress, maybe not, maybe bureaucratic rules. And then there's sort of like the cultural change, too. And that's also extremely difficult. And sometimes maybe you just need like a sort of bully type who is like very aggressive to change. When you think about your vision for intra-D-D competition or bringing more of that, how much of that is changing rules versus changing culture? And how do you think about sort of like we're adopting Shams 18 Theses? What does that actually look like in practice?

Yeah. So I think a lot of it is cultural. I might call it leadership. If we think about Hyman Rickover, so Admiral Rickover built the nuclear navy. He was a five-foot-tall, pretty middling to bottom-of-class U.S. Navy Academy grad, didn't even drive a warship. But after World War II, by happenstance, he got to go see the remnants of the Manhattan Project and realized, holy cow, we could be building nuclear-powered submarines and warships with this. Admiral Rickover was in place leading that effort for 30 years.

He was protected by Congress. He was a notoriously difficult human to deal with. Admiral Zomwald, who was the chief of naval operations, basically the senior most Navy person, said, you know, the Navy has three enemies, the Soviet Union, the Air Force, and Hyman Rickover. So you can see how difficult this human must have been, but he was a genius. And I think there's no simulation of the world where we said, what if we had only Rickover for three years? Or what if we had him for 10 years? Like our nuclear navy is one of our last remaining asymmetric advantages, and he built it

in the 50s and 60s. It's truly spectacular. And I think if we go back and look at our history here, Kelly Johnson, building 40 airframes, including the SR 71, John Boyd, we call it the FSA. 16, maybe more accurately, we should call it John Boyd's plane. We call it the Apollo program. Maybe we should call it Gene Kranz's program. There's this beautiful humility in this kind of Midwestern sensibility we have in America that we think of the program as not being defined by a single person. But oftentimes it really is. And so to the point of culture and leadership,

ever since the Goldwater Nickel Reform of, I think, 86, where we prioritize this sense of jointness. You know, we want people to rotate every two or three years. We want them to have lots of different experiences to fill out their bingo cards. It's been very hard to have actually the deep time and space that's required to be in a role long enough, not only to make mistakes, but to learn from those mistakes so that you can ultimately succeed in delivering cutting edge technology. You know, I like the idea like, oh, there was this guy, Kelly Johnson, he made 41 planes, etc. And why can't we get back to that? Or like someone is just some like visionary and they like build all this stuff. On the other hand, like it's clear that like objectively airplanes

must have been much simpler in some sense because we know the Wright brothers, for example, invented, and that was just two of them. It is inconceivable today on any scale that two individuals could go into a barn and build an airplane. I mean, so like I get this like story that's like cultural change, et cetera, but just the sheer technological complexity of the cutting edge technology, doesn't that just sort of like make some of these stories impossible to replicate today? My counterpoint would be SpaceX. Elon can build a starship faster than the FAA can give him launch approval. What they've done in the last 15 years when I grew up in the shadow of the space coast,

I used to wake up to the double sonic booms of the shuttle reentering for landing at Kennedy. The shuttle was amazing. It inspired me in so many ways to follow STEM. But the launch cost was $50,000 a kilogram to orbit. With Starship Heavy reuse, we're looking at $10 to $20. I'm not missing any zeros there, $10 to $20. And we were stuck on a linear cost curve between $20,000 and $50,000 until SpaceX came around. And I think the timelines they've done this, and people would say that's not possible.

It's within a small lifespan of someone working on these rockets. And it's largely been the same people who've been leading these efforts since the very beginning. So I think when we take new approaches, we can still do these things. We can still do these things very fast. I think one of America's great strengths, if we were literally at war, we would throw all these rule books away and we would just be entrepreneurs again. We'd be very innovative and we'd be very focused on the primacy of winning. But the sort of sclerosis that sets in with peacetime of following all these rules without a critical eye towards which rules are making us better and which ones are making us worse, you accumulate them. John Boyd, I used him as an example.

He never made it past colonel because he was so difficult. was unpromotable. We tend to give credit to someone like John Boyd, as we should. But we don't give enough credit to is the three-star Air Force General that protected Boyd his entire career. There was the talent, but there was also the talent manager who understood that there was something special here, and it was worth going out of their way to protect this person. The same is true for Rickover. He didn't survive on his own. Congress deserves a lot of credit for protecting him. So many admirals, so many civil political appointees tried to get him fired. And no one succeeded until Lehman in the 80s, and arguably layman succeeded for the right reasons.

I'm glad you brought up SpaceX because, you know, you're talking about encouraging commercial companies to do more defense stuff and the sort of co-mingling and the benefits that would come with that. But a lot of people would look at SpaceX and say, well, it's kind of worrying that we have a commercial company, which is now dominant in U.S. space exploration. Or they might look at Starlink and say, like, oh, wait a second. And Elon Musk owns all these really important satellites. And maybe that's not ideal for the U.S.

Maybe the U.S. government wants more diversification in its suppliers of extremely important infrastructure and technology. How do you ensure, you know, as more commercial companies come into the space, theoretically, how do you sort of deal with those types of concerns? Well, I think one of the great strengths of America is our skepticism. them towards power. So here we are. We've reduced launch costs from $50,000 a kilogram to $10 a kilogram, but we're still worried about the accumulation of power, and I think we should be. But on the other hand, to look at that without acknowledging the massive effectiveness, like our assured access to space exists because of SpaceX, our dominance in this category, it has become a national security strength. So I think we have to recognize that we've

gotten this massive outsized, nonlinear benefit, and we have these concerns. And the way to address this concerns is to have more options, more competition. We need more people. people who could achieve launch costs that are actually competitive. You know, when the alternative is Boeing and the Starliner, that's not a real alternative. What should we be watching for with the new administration? Obviously, we have the new head of the DoD, Pete Heggseth. In the months ahead, who are either the individuals or bureaus or whatever, where we like, oh, something is meaningfully changing here in terms of how we buy defense?

It's a good question. I mean, I have a lot of confidence in the incoming team here, folks like Emil Michael, really coming from the tech industry, leading R&E, folks like Steve Feinberg, who've been in and around many of these challenges in defense in the defense industry for a long period of time on the private side at Cerberus. So that gives me a great degree of confidence that these folks actually understand what the problems are and that there's a need to fundamentally change how we approach these things. I don't expect it to be perceived as being wildly discontinuous because the answer is not to get rid of America's primes. Like America needs her primes. Like the things that Lockheed and Northrop and the other three of them do are exquisite.

They are incredibly high tech. They're also low volume and they're not the entire solution. So it goes back to this idea of asset allocation. And I think that would be the first thing that we would see as a sign of change, which is a changing of how we're allocating those assets. I think a lot of the ideas that Senator Wicker has put forward in the forged DAC, where the first step is to cut red tape is a big part of this, which is I think we have all the talent in the world. This department is extraordinarily talented. We have the people. We need to actually just put them in the role and give them the discretion to actually execute these programs. Understanding to Tracy's point earlier, some of these are going to fail probably in pretty spectacular ways, but some of them are going to succeed in very spectacular ways.

And we just need a few of our asymmetric capabilities to really work to preserve deterrence. So I don't know if Elon Musk listens to this podcast, but I do know that he sometimes reads tweets about this podcast, or at least he did once and responded to it. But what would be your big piece of advice for Doge as it sets out on the very, very large task of eliminating waste at the government? You know, there are sort of infinite ways you could do this because the government is so sprawling. Gosh, I mean, far be it for me to tell Doge how to do this. It seems like a group of experts here. But my broad advice, 19 years crawling through and fighting the bureaucracy and DoD, is that there are two things that really matter. The first is that we should push more decision-making authority on what to buy and how to buy these things to our combatant commanders to introduce strategic competition. You don't need to give them the entire budget. I'm not sure they would succeed if they were given the entire budget.

But even having 5% of it to express where they disagree with the market signals or to be able to decide which of the services are going to build something for them, I think starts to introduce the necessary vector for new entrants to get on the field and new ideas to be tested in reality. And the second part to that is a very concrete way of how we could do that is we could introduce this notion of competing programs more formally. We used to have it informally. Like I mentioned earlier how we had four competing programs for the submarine-launch ballistic missile. that does not happen today because that duplication is viewed as heresy.

That would be being irresponsible with taxpayer money. So before we even get started, we need to agree ex ante on a unitary effort. I'll give you a counterfactual. Should we have had one F-35? We conceived of the F-35 in the mid-90s, and it was supposed to be all things to all people. There was going to be a variant for the Marines and the Navy and the Air Force. And if we just have a single platform, it'll be a Swiss Army knife of, fighter jets. And now, you know, it's pretty spectacular plane, but it's also 30% available and it's

going to cost us $2 trillion. Maybe a counterfactual would have been, we're going to have an F-35, an F-36, and F-37, and none of them have the birthright to win. And in fact, we're going to go to the quote-unquote legacy platforms, the F-18, the F-16, the F-15, and we're going to say, here's a little bit of money for you guys to innovate and to actually be so good that I decide I canceled the new programs. But instead, we said, you're the old, no more money for you. Before the new thing even exists, we've already told you that you're going to be sunset, and we don't have a burdened hand. And I think this is one of the major things that slows down innovation. So institutionalizing the concept of the competing PM, I think is very important. I'll give you an anecdote from my

own experience. We were the software prime on a hardware project. We were building a satellite ground station on wheels. And in the midst of this competition, it was Palantir v. Raytheon and the teams that we had underneath us. In the midst of this competition, chat GPT came out. And I thought, this is great. You know, this is a very complicated piece of equipment. When my son is in the back of this vehicle, I don't want him flipping through a 10,000 page PDF manual. I want him to be able to chat with the bot to figure out how to fix the vehicle or troubleshoot any situation that he's in. And I went to the PM and said, I'd like to give you this chat bot, you know, for free. It's just going to be a better product. And he said, there's no requirement for that.

even though there's no, I'm not asking for a resource, you know, I guess the perceived possible risk for schedule, there's no incentive to do it. Why do it? I've already decided what it is we needed and this not on it, even though the world's changed around us, even though the competition takes two years and two years in infinite time and technology,

let's not even consider it. So by the time we're going to end up considering it, it'll have been an obvious thing and we'll probably get to it like 10 years from now. But I'll give you a counterfactual where what if there are 2 PMs, one for Palantir, one for the Raytheon team. And they woke up every day thinking about how they were going, to destroy and humiliate and win the competition.

I'm pretty sure that PM would have said, absolutely. Let's show up to the next soldier touchpoint with this great capability. I can't wait to win. And I think if our only competing point is at a period of great conflict in war, that's too little too late. We need to have many competitions amongst each other all the time so that we're continuously innovating. Shamsankar, thank you so much.

Really appreciate you coming on Odd Lots. And great to talk to you again after all of these years. much heavier topics than steam buns, which I think is what we talked about last time. Well, thank you guys for having me. Tracy, I like to believe the idea that we could have intra-military competition. Like, it's a really nice sounding idea. Yes, intuitively sounds nice. I guess my reticence is slightly, you know, okay, everyone can agree.

Monopolys are bad. Monopsony is bad. But I guess my reticence is you brought up. the specialization point, right? Like the idea that companies can get really good and efficient by doing one thing and maybe that one thing is developing ballistic missiles or whatever, like that exists in the world for a reason. And there are lots of examples of that. And, you know, there are anti-examples of that. Boeing does both military and commercial. And I would say at the current moment in time, Boeing has not been especially successful. So there's that point.

which I think is a good one. And then the other concern, I guess, is the one around, you know, corporate dominance. Yeah. And the idea that, well, you bring more commercial players into the market for something that is really essential, as Sham emphasized many times. Public infrastructure. Yeah. There's a little bit of attention there, bigger than a little bit, I should say. No, absolutely.

I do take his point at the end that it would be nice. I mean, I think there's a lot of points. but one is the idea like it probably makes sense that if you're going to have competition, that means you're going to have failures, that means you're going to have wasted public money. That's like a cultural thing that strikes me is like very difficult. But obviously in the private sector, big winners emerge and we don't usually talk that much about all of the ones that were losers, but they had to be there. It strikes me as, you know, like a tremendously difficult task.

It also seems like it would be nice to have like good sense of. the state of our defense industry and not suddenly discover various weaknesses at a time of actual kinetic warfare where we're being put to the test. So I thought, anyway, we got to do way more on this topic. And I appreciated that as a sort of way into it, even obviously coming from a very specific perspective. A launch pad into the topic. But I do agree. The idea of being able to change course seems like a really important one. And the story, you know, the example he gave about RFQs growing from like 20 pages to 2,000 pages. That's something we hear over and over again.

I could also probably just listen to stories about, you know, random guys in Ohio building planes in the 1950s. You have truly reached Middle Age. Oh, God, I'm so boomer. I'm so boomer with this stuff. This is totally true. All right. Shall we leave it there?

Let's leave it there. This has been another episode of the Odd Thoughts podcast. I'm Tracy Alloway. You can follow me at Tracy Alloway. And I'm Jill Wisenthall. You can follow me at the stalwart. Follow our guest, Sham Sankar. He's at Sankar.

And you can read his big thoughts about defense at the website, 18theses.com. Follow our producers, Carmen Rodriguez, at Carmen Armand Dashel Bennett at Dashpot in Kailbrooks. For more Odd Lots content, go to Bloomberg.com slash oddlots. We have transcripts, a blog, in a daily newsletter. And you can chat about all of these topics 24-7 in our Discord. Discord.g.g. We have a defense channel, so go check it out. And if you enjoy Oddlots, if you like it when we try to learn about the defense industry,

then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes, absolutely ad-free. All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening.