Expanding the Defense Industrial Base | 2025 Air, Space & Cyber Conference

Please welcome the moderator of our next session, General Hawk Carlile, former commander of Air Combat Command.

Well, good afternoon everyone. It's great to be here. I appreciate the opportunity to spend a few minutes with you. You notice I had to lower the mic. I have always had that problem. But I will give AFA a lot of credit for 39 years in the Air Force. They kept introducing me as Herbert Carlile. I hate that name. So they introduced me as Hawk and I really appreciate that. So it's great to be here today and I'm honored to share the stage with three brilliant folks. I know we're the only thing between you and a cocktail probably or close to a cocktail. So we'll try to make this as entertaining and exciting as we can. So with me on this stage, we'll start with Brian Kroger. He is a former Air Force targeter, some of the, in my opinion, the unsung heroes of air power in the United States and the US Air Force. He was one of the original founders of Kessel Run. Since then, he went out on his own and he's the CEO and founder of Rise 8. So, Brian, it's great to have you on the stage with us. Next to him is Raj Shaw. Many people know Raj. Raj has been around for a long time now. He's a Viper driver, so I try not to hold that against him. I like them anyway. But Raj has been around the department and been instrumental in a lot of the advancements that have been made in technology. He ran DIUx out in Silicon Valley. I think he was the last guy that was a leader of that with the X next to it. I think Secretary Mattis removed that as he was leaving. But Raj did great work. Now he is CEO of Shield Capital. He's also the founder and running an organization called or a company called Resilience. So doing great work. Raj, great to have you on the stage with us. And finally Shyam Sankar is the Chief Technology Officer and Executive Vice President at Palantir. I know many people know Palantir, a relatively new company that has just come into the Department of Defense and the industrial base with gangbusters and doing fantastic work. So Shyam, it's great to have you on the panel with us. What I'd like to do to start this is let each one of them spend a couple of minutes just talking about some of the things that are of interest to them and then we'll get into some questions that we're going to ask them. So with that, Brian, I'll start with you.

Thank you. Right now I am very focused on software development and specifically continuous delivery out to the field. One thing that I think is interesting is if we start to think about software as a weapon of war, there are some components of our software, not all of our software, certainly not our business software, but some of the mission software, the most important aspect of it is going to be how quickly we can sense and respond to the conflict because software like plans will not survive first contact with the enemy. It's going to be our ability to sense what the enemy is doing and respond with new features, bug fixes, cyber defense. And if you think about it that way, that means that the software we go to war with is irrelevant. We're actually going to war with our software developers. And because of that, I think for those unique mission sets, we really need to be investing in people who are soldiers, sailors, airmen, guardians, marines, because again, we're not going to war with our software. We're going to war with software developers. They should be uniform members, not contractors, including Rise 8. So that's a really an issue I'm really passionate about is the military building up a cadre of people who are software developers. It will take a lot of time. But I think it's critical not only to those edge cases but then also to being more informed buyers of software and so I think it's a huge investment we need to make completely irrespective of my business, something that I'm passionate about.

Thanks Brian. Raj. Thanks. It's great to be here. And as Hawk said, I was the last director of DIU when it still had the X, when thinking about working with the commercial world and technologists was considered experimental. It's amazing to see the change now. It's been a massive sea change. Walking on the showroom floor earlier today, new entrants have some of the largest booths. The core technologies that will be decisive to deter and win on the battlefield of the future are coming from new entrants. So it's no longer an experiment and it's really incredible to see this interaction. I know we'll talk about it more. I think the last piece of that I'll highlight is that at the end of it, this collaboration between technology companies, our men and women in uniform, the Pentagon bureaucracy, at the end it's a human endeavor. It comes down to the people, the people in these organizations, their willingness to be courageous, do things differently, find new ways to build bonds. And I got to give credit to the Army because to my left here is Lieutenant Colonel Shyam Sankar who recently took the oath and is now serving in a reserve capacity. The Air Force has taken significant leadership in this as well through this program called Gig Eagle which allows our reserve component folks to find ways to support. So I think we're still in the early stages but it's so much different than it was a decade ago. Thanks Raj.

Thanks Shyam. Great intro and I appreciate all those comments. Let me start with something that I think it may have a bit of a negative tone or negative connotation and that is the term innovation theater. And what that means, there's a lot of talk out there about being more innovative and how do you get there? But as the term kind of sounds like, innovation theater is a lot of it is just that and it's not a solution. It's not concrete. It doesn't fill a gap the military has. So what we really need is efficient capability. But we also need effective and prompt capability. So how do we get away from innovation theater and get to the heart of producing efficient, effective capability in a time that suits a warfighter need to compete with our adversaries?

Thank you, Hawk. Yeah, as someone who's seen a lot of innovation theater with a non-stop string of senior officers coming to the valley to go ride in a Waymo, I think that the core difference is combat effectiveness and how do we do that at scale. So as we think about rebuilding the defense industrial base, the scale component of that is something that we have forgotten about, as Shyam is talking about in World War II and the Arsenal of Democracy, how Ford and Chrysler stamped out B-17s and B-29s. What's the equivalent today? Well, it's not large aircraft. It's drones. For example, DJI, the Chinese drone manufacturer, will build 10 million drones, mostly for hobbyist use this year. The Ukrainians will build 4.5 million drones. I'm told that the US small drones will build about 100,000. So we're about two orders of magnitude off. So if we think about how do we move from theater and good ideas and that early prototyping to scale, it really comes down to the hard choices. The hard choices are senior leaders and the services have to make and of course the appropriators in Congress. And because this is an Air Force event, I'll pick on the Navy. Everyone loves autonomy and the future unless you have to decide are we going from 11 aircraft carriers to 10, and there's equivalence now in the Air Force. And so these hard tradeoffs of being ready to fight tomorrow yet preparing for the future are the ones that need to be made.

What I'll add to that is that by the way, when I was talking about military software developers, not to suggest that our entire software development should be run by uniform soldiers and airmen and marines and guardians, but that you need that knowledge to effectively lead in this space. And one of the problems I see in innovation theater is defining what you want, which should be a mission outcome, not some IT metric. That's what gets you efficacy. So are you actually making a difference for the mission and you have to measure that relentlessly. And then measuring efficiency would be those software delivery performance or IT metrics. If you go onto any competent manufacturing floor, everything is measured to the nth degree. But if you go into a DoD software factory, there's almost no measurement. And so I think that's a real problem and that's what allows the lies to happen and they go unchecked. There's no accountability. So we need the expertise in our leadership to define what the innovation is supposed to achieve from a mission perspective and then turn that around into delivery metrics to make sure that we are getting good value for our money. But I also agree that we prematurely optimize and try to go for efficiency. We should accept some amount of even double or triple or quadruple spend in the exploration phase before we try to exploit. Peter Drucker said there's nothing as useless as doing something with efficiency that shouldn't be done at all.

Okay. So, having spent a good portion of my life being a general officer that did self-flagellation, I kind of feel like I have to ask this follow-on question. So, yeah, it is taxpayer money. And although we talk about pay double or triple or failures, at the end of the day, there's 535 board of directors that sit on Capitol Hill. And so, how do you translate accepting something that might fail, probably will fail, or you're paying too much to get that effectiveness in a timely manner? How does that square with it is taxpayer money and there is a reason that you have to be somewhat judicious with those dollars to reach the goal that you're trying to reach?

Yeah, it's a great question because being good stewards of taxpayer dollars is important. But I think it's a reframing of risk that's really needed. And I'd say on two dimensions. One is understanding that when you're creating and inventing something new that's never been built, you're going to get a lot of things wrong and it's better to get those things wrong fast and at a lower dollar value. So back to Shyam's point, if we were building a new system and we contract with four folks, some have done it before, some haven't, but you move quickly and you be very demanding as a customer. Okay, in 6 months we know that these three did not work and so we're going to let those fail. That's okay. It's actually success because in a multi-billion dollar program, losing a couple 10 million early on to learn from, it works. And this is exactly how the valley and venture capital approaches its investments. We make a slew of investments knowing that some won't make it and the ones that do will be successful. So I think that's one, it's a reframing of risk and more importantly, it's helping the appropriators understand that reframing of risk and having a good partnership with the senior leaders. And I think the second though is understanding where should that risk lie. Who should absorb some of that risk? And right now by having these very long programs, particularly in software, that create subpar equipment but they're a little bit too big to fail, we're transferring risk from the cubicles of the Pentagon where a paper cut is your worst risk to our men and women in the field who then are going to take whatever you deliver to them and fight with it. And so I think we need to again reframe where should we take risk and who needs to be courageous.

Yeah, I think this is an old stat now but when I was starting Kessel Run, this was back in 2017, USDS put out a report that over 80% of federal IT projects were over schedule or budget and 40% failed to deliver period. So the way we're doing business today is the risk. And I think that the fundamental risk is that we have fake competition. So every acquisition strategy or contracting strategy, the competitions that are being run from what you would normally consider a competition are fake. And I've heard you say what I'm about to say before. There is a better way to approach this, but it requires being willing to double or triple spend or maybe shape your acquisition strategy for burn in a way that allows you to do that with the same budget you have. But the real competition should be between two program managers running two different contracts. That's a competition. And I would love to see three or four or five. So I would love if any of you out there would be brave enough to do that. I think that's the future is forcing program offices to compete with each other rather than a fake competition on paper to get on contract.

Yeah. That goes back though if you have two or three programs but only one's going to go to full production. Again, that goes back to the cost of that which I'm not saying is not right but it is one that you do have to go back to your board of directors and try to make that happen. So the next part of the question that goes with that, because I agree the last place and the place that we most definitely don't want to place risk is those young women and men that raise their right hand and swear to protect our Constitution. So, are we growing in today's technology world the folks that can make those decisions on what the risk is and how to evaluate programs?

I would agree with Shyam. We have incredible talent in the military and enabling them and letting them learn also what's happening in the outside world is important. So as we think about the defense industrial base and the modernized industrial base with a lot of software and low-cost hardware, it's just different than the base that we've been used to. And I think the good news is that I've seen out in the valley, I have not seen more engineers, technologists, and entrepreneurs that want to build companies serving national security. They watch the Ukraine invasion and the conflict in Israel on TV. They understand what the risks of authoritarianism are. And they know that they can build a real business now. An unbelievable number of them. And I think it's now the responsibility of us and the folks in the room to say how do we take advantage of that? How do we encourage these individuals to come in and serve either as performers as companies building important kit or even finding ways to come in and spend some time in the reserves or even the civilian reserve after the announcement of Shyam. I mean I must have got a hundred phone calls from others that said how do I go and do this, brilliant technologists. And I think this is the American way of civil-military fusion that we need to embrace through the free market economy and through the power of the purse rather than CCP-style central government and that's the advantage that we have.

It's become a bit of a meme but I think we need to look at embracing founder mode. If anybody ever came across that, it has become a meme but if you go back and look at Chesky's original comments around founder mode and why it's important, which Silicon Valley is starting to look at more, where for a while there was a buy and replace the management structure but now keeping on that founder is precisely because of what Shyam said, those years of institutional knowledge and failures and lessons learned that you can apply towards the future. And so when you take programs like Kessel Run and you swap out the C-suite every three years, how do you expect that to succeed? No software company in the world could survive swapping out their entire C-suite every three years. So it's not like the normal military operations that we have SOPs for and you can just have an infantry mentality where the person in the front drops and the next person picks it up. It's just not going to work for innovation. And so I think revisiting the assignment cycle. And for what it's worth, not only did I not get to stay at Kessel Run, but they looked at me, so I was an intelligence officer on an acquisition intel exchange and I was leading a 1500 person organization with a $250 million budget. And they said, Brian, you really need to go back and get a squadron command, a DO, right? That was my next step. So I had to go back to Intel and be a squadron DO for a smaller organization than what I was already effectively a DO for. And I know many such examples. The person that started the Section 31 software factory, they told him he needed to go do a logistics tour even though he was probably one of the leading software experts in the Space Force. And so we really need to revisit that.

I think that's a good point. I do believe that all the A1s in the world hopefully are listening. But I do think there's something to be said for the fact that we need to build people in positions and allow them to really become professionals at that position. But there is a counterargument of you need that kind of experience in other places as well so that they can help in the decision-making process at other levels and up. So again, taking good notes now that we solved the effective capability problem, and I hope everybody has good notes. The next challenge that I think would be great to hear these three brilliant gentlemen about, and that is speed. We know we have a potential adversary out there that operates in an authoritarian world that they don't have DAFs. As far as I know, they've never had a continuing resolution. They can throw resources where they want when they want at the drop of a hat based on one individual's input. So how do we, one of the challenges I think I faced when I was a customer was I need this capability now, next year, 6 months, 18 months, I don't need it in 5 years. And if you look at the acquisition process, the cycle, and we're trying, I mean there's OTA, there's mid-tier acquisition, there's DIU, there's Raider, there's programs, but how do we speed up and how do we get through the bureaucratic challenges that we face in acquisition in the Department of Defense?

Sure, I'll start with that. Having spent so much time having read the 5000 series, in fact I think we should classify the 5000 series so the Chinese will steal it and maybe adopt it. That could be the most damaging thing we could do to them. I think what you said at the end, Hawk, is spot on, meaning we have a lot of these authorities. There could be more. It'd be wonderful that we didn't have CRs, but that's probably not the audience in this room that's going to have a play in. There are some new authorities that the NDAA will come and every year they have it. But I would say by and large, the department has 90% of the authorities that it needs. It really comes down to incentives. Do we give our KOs the incentives to move fast? And in most organizations, if you're a contracting officer and you take five years to make a decision or a process, no one's going to fire you for it. But if you move really quickly and god forbid you make one small mistake, you're in trouble and your career is in jeopardy. And so I think it's the incentives that we put for our teams that are most important. And again, this comes down to the senior leadership saying we want to move quickly. Two years to do this selection is not satisfactory for the problem set and we're going to go at this rate and people willing to take risk. And we have a long history of secretaries in the services and in the department that have done it, just need to do more of it.

Yeah, work feels time allotted, right? And so you have to have really high standards at the top for what's in the realm of possible. So if you're starting an unclassified software program, the first valuable delivery, which I mean the MVP term gets thrown around a lot, I mean it actually has mission impact. Anybody should be able to get that done in the first 6 months or less. I've seen it done in as little as 60 days. And I'm talking full ATO in production in the hands of warfighters. So if you're willing to say, well, we're willing to spend two years on the source selection, you're already that's longer than it takes to develop software. It's crazy. And so I think those high standards are really important. And then I think there's also this piece that we overlook which is slack versus utilization. So in the human capital realm we optimize for utilization because it's efficiency. We need to be optimizing more for slack. And what I mean by that, I'll use the ATO as an example. We've done value stream mapping on organizations' ATO processes and the average that we're sitting at right now is just over 80%. I don't know the exact number, but about 80% of the time to get an ATO is time spent waiting in queue status. And you're waiting in queue status because we've optimized for utilization of assessors. We could delete some of the assessment process too, I think, safely, but also there's just a human capital problem where we're so focused on efficiency, we actually lose sight of effectiveness. 80% of the time just waiting in queue, nothing getting done.

Maybe I'll add something here too, which is there's not one size fits all. People talk about acquisition reforms, we got to change the whole system. I think the system, if you want to buy a piece of equipment that we're going to keep for 50 years, I think it actually works fairly well. It's designed for that, but we need a different system for software. The idea of sustainment and operations and procurement of software, by definition, you're doing it wrong. Pull out your iPhone, there's probably 50 undone updates that you need to do. So it's constantly changing. And so if you make a mistake in software, you can fix it. And same as we think about low-cost drones, the idea again of risk and quality. Warfighters, we'd love for it to take off and come back every time, it's really important. If we have small drones and we take off a thousand of them and 50 of them crash on takeoff, maybe that doesn't matter because they're $500 a piece and the next version is going to come out 3 months later. And so it's a different mindset. It's not a one-size-fits-all. And I think we need to understand in this future of war, it's not either or. We're going to need the exquisite high-end that has to be done really well and the low-cost more tradable systems where speed is more important than near-term quality.

Yeah, that was a great discussion. I will tell you I do believe from my perspective the United States Air Force, United States Department of the Air Force acquisition process is getting better. We always, again dating myself back before everything was any PEO and any contract was evaluated on cost, performance, and schedule. So what did you do? Made the cost as high as you could, the performance low, and the schedule long. So you would always be successful because that's what was incentivized. But I think folks like Duke Richardson and Dale White and Wahoo Edwards have really been making a change there. I would like to ask, we got about 9 minutes left, but I'd like to ask one question and you're all brilliant when it comes to this, but when I was a customer, I always felt like we needed to make software consumable because the standard process was too long. Part of that is open architecture. I think that's part of it. But the idea that given cyber defense, given the changes in technology, given the ability to turn software, you can't take two years to fix software. You can't take six months to fix software. It ought to be, as was said earlier, it ought to be a continuous process. So on the software side, how do we get to the point where we're making those changes at a pace that keeps up with the technology and the challenges our adversaries are throwing at us?

Yeah, I think there is progress being made and this NDAA, if it passes, the color of money is supposed to disappear for software which would be phenomenal. But there's a joke that I've been making for a long time and I'll continue to make it, which is if the department wanted to do one thing to make its service members and the entire staff more efficient by I would say 50%, it is to change DTS, put in a modern system. And I used to make this joke like 8, 10 years ago and it was really a joke and now it's sad that we still have, for those of us in reserves I don't know if you have to do it but it's a discouragement and a waste of time. Yet this is not a unique mission-critical piece of software, using advanced AI to target underground bunkers, this is travel and every other large major corporation has it. So let's take what's already existing and adapt it. And I highlight this because again, it's not a technology problem. I don't think it's a process problem. It is a problem of courage of saying, hey, what we do right now isn't working. And it is inflicting a lot of cost on our service folks and let's make a change. And that's hard to do.

Yeah. I talk about this a lot in terms of the value line. You have to know where the value line is in your organization and it can change over time as your organizational competencies change. But your value line, what's above it is what differentiates you. So like you have a unique mission, space force, air force, those are the kinds of things that you should think about building. Everything else you should buy or better yet rent. And so if we're talking about travel systems, buy or rent them like a SaaS model would be perfect there. Your finance systems. And yet what we end up doing is we try to buy the mission software and build all of the business system software. Another example of this though in software in particular is platforms. Most modern applications run on some cloud platform and there's an incentive for services providers, and I'm speaking as one but you won't hear many people say this, that if the government is just going to incentivize cost, schedule, performance in the way that you described, there's a huge incentive for me to go build the platform every time. It's way more headcount. I can bill you a lot more. I can make a lot of margin on that stuff too. But if you start optimizing for value, because cost, schedule, and performance are one side of a value equation. So value equals performance divided by the product of cost and schedule. And if you're putting me on the hook for value now, and I have to optimize for that, I want to get to performance as cheaply and quickly as possible. So in that scenario, instead of building my own platform, I'm going to call Shyam and I'm going to say, hey, can I use Apollo? Because I can get my software into the hands of warfighters quickly and all of my resources from an efficiency perspective are going towards building mission differentiating software instead of rebuilding all of the plumbing and foundation. And so I think that if you start talking about value, it fundamentally shifts the incentive structure for all of the people who are building unique software. And again, you should only build what is mission differentiating. Everything else, I would agree, just buy it.

Great. Thanks. I appreciate that. Great comments. I think what I'll do first is I'll allow a minute or so of closing comments from the three of you and Shyam, we can start with you.

You know, I'm very optimistic about the future and all these challenges that we talk about because again, I think that the American system, the free flow of capital, the free flow of people, this is what makes us strong. And so, a little bit like what Shyam says, we shouldn't shy away from what has enabled us to overcome significant challenges in our history. We still have the most innovative people, the most innovative engineers, and the most risk-taking entrepreneurs both inside the government and outside and I think we need to celebrate them, champion them, and double down on them. And I think again, as we've seen in our history, we will overcome. Thanks Raj.

Brian. I want to go back to the topic of courage to close out, which is John Boyd's famous roll call, to be or to do. I think is really appropriate at moments like this. He famously said that you can either choose to be somebody or to do something. If you choose to be somebody, you'll get promotions, you'll be the favorite of your superiors, etc. And if you choose to do something, you might lose all of those things but you'll do something great for your nation. Everybody wants to get up. In fact, you probably might have been surprised with this group. It was good. But usually these conversations turn into acquisition reform and ATO reform because we talk a lot about software. You don't need reform in any of those. People have successfully navigated those without breaking policy or needing waivers and have been able to get software and hardware into the hands of warfighters quickly, safely, and sustainably. What's missing is actually courage. You have to choose. Do you want to be somebody or do you want to do something? And it might make you not get promoted. It might make you not a favorite of your superiors, but you'll do something great for your nation. And so don't wait for reform. Just do something.

So folks, let me close this out by saying Brian, Raj, Shyam, thank you very much. Great insight. Truly appreciate it. It was an honor to be on the stage with you and I appreciate the opportunity to hear your answers and hopefully for the crowd we provided some insight. Let me just close with kind of in line with what our panelists said. We have the greatest Air Force and the greatest Space Force in the world bar none. And it is an asymmetric advantage for this nation. We outpace our potential adversaries in those two domains by an order of magnitude. But our adversaries know that. And the only way we're going to stay where we're at is if we have the incredible professionals we have wearing the uniform, incredible professionals in the civilian corps, and the incredible professionals and patriots in industry that continue to provide our warfighters with the capability we need to deter and if necessary defeat any adversary in the world. Thank you very much. Enjoyed the time today and have a great afternoon. Thank you.