Inside the Department of Defense and its Vision for the Future

Host: In real dollars, DoD's 2024 R&D budget request of $145 billion is triple what it was during the Cold War. It's more than what Alphabet, Apple, Meta, and Microsoft combined spent on R&D last year. Defense is not a small market, likely one reason behind the recent surge in U.S. defense tech-focused startups, scale-ups, and private and venture capital. Some 2,000 deals investing over $100 billion since 2021 per PitchBook in the Wall Street Journal. The issue is, are we able to buy what we need and are we thinking far enough ahead to what the shape of warfare needs to be? What we did in five months normally takes DoD two to three years. If you're not sure what is more mind-blowing, how fast we did it or how long it normally takes, I don't blame you.

The origins of Silicon Valley are deeply rooted in government and defense. From air-defining radar or electronic systems that help the Allies win World War II to the modern semiconductor, emerging from the need for better missile and satellite capabilities during the Cold War. Now, in today's episode, we get the opportunity to revisit that very relationship. Together with one of the most critical people within the Department of Defense. That is the United States Deputy Secretary of Defense, Kathleen Hicks. Kathleen has spent decades of her career focused on the evolving, complex topic of defense policy. And today, we get the pleasure of hearing her keynote address from A16Z's second annual American Dynamism Summit, recorded this January in the heart of Washington, D.C. This event was not only a platform to celebrate our national interest but also a forum for engaging in the tough conversations necessary for America's growth. So in this slightly abridged address, you'll get an inside look into the priorities of the DoD and how it sees its past, present, and future. Make sure to stick around for after her address, where you'll get to hear highlights from Deputy Secretary's fireside chat with Wall Street Journal, National Security and Foreign Policy editor Sharon Weinberger, where they dive into the numerous issues facing our country today, from Ukraine to the use of AI on the battlefield. But first, let's go straight to the source and hear from Deputy Secretary of Defense Kathleen Hicks.

[As a reminder, the content here is for informational purposes only, should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security, and is not directed at any investors or potential investors in any A16Z fund. Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see a16z.com/disclosures.]

Kathleen Hicks: The story of Silicon Valley and the sector it symbolizes shares much with the story of America. It's a story of millions going west to find their fortune, of immigrants pursuing their own American dream, of not all people getting to participate equally, especially at first. It's a story of free enterprise and entrepreneurial spirit, a story of dreamers and builders who dare to pioneer new frontiers, reimagine the future, and change the status quo. But it's not just a story of rugged individualists, founders, and funders. It's also a story of the institutions that supported them and the talented communities around them. It's a story of collaboration, of teams and teachers, formal and informal. Many are not household names. They may not have been wealthy, but they were indispensable to every disruptive innovation. And it's a story of what government, of the people, by the people, for the people, can make possible.

Think of NASA and the Pentagon buying so many silicon microchips in the 1960s to guide moon rockets and Minuteman missiles that the price per chip fell from $1,000 to $25 in just a few years, low enough for commercial applications. Think of DoD creating GPS decades before smartphone apps and startup business models relied on the geolocation services it provides. Or the National Science Foundation funding the Supercomputing Center where Mark Andreessen worked on his first web browser. Think of DARPA funding early AI research at Stanford, the labs where Vince Cerf and Bob Kahn fathered the internet, and the grand challenges that road-tested technologies used in today's self-driving cars.

Of course, none of that reduces the audacity, the brilliance, the creativity, and determination of all the innovators past and present throughout America's commercial technology sector. The point is, our histories are bound up together more deeply than we sometimes admit, and our fates are too.

Over the decades, the ties between the Pentagon and the tech community have waxed and waned. Even after tech-minded defense leaders like Harold Brown, Bill Perry, and Ash Carter sought to rekindle flames, rebuild bridges, and rewire the Pentagon to be a better partner, some in DoD still undervalue nontraditional and commercial innovators in Silicon Valley and beyond. Perhaps they longed for when federal money dominated. The early 1950s through the late 1970s, when government and especially DoD contributed between one-half and two-thirds of all American research and development dollars. That same era birthed DoD's much-loved multi-year budgeting processes. In funding such innovation, it was easy, and Washington got comfortable believing our processes and pays could dictate everything that followed.

But since the 1980s, the script has been flipped, with companies contributing 50 to 70 percent of America's R&D pie. Yet DoD's rules of the game have not caught up. Not only are we seldom innovation pace setters, too often we can struggle just to keep pace with a dynamic U.S. private sector that continues to out-innovate the world. Of course, it's a much larger pie, and we still invest a lot. In real dollars, DoD's 2024 R&D budget request of $145 billion is triple what it was during the Cold War. It's more than what Alphabet, Apple, Meta, and Microsoft combined spent on R&D last year. And our requested 2024 procurement budget, what's converted into fielded capabilities and services, is even more, $170 billion.

Defense is not a small market, likely one reason behind the recent surge in U.S. defense tech-focused startups, scale-ups, and private and venture capital. Some 2,000 deals investing over $100 billion since 2021 per PitchBook in the Wall Street Journal. For three years now, we have taken a comprehensive, iterative, warfighter-centric approach to innovation, recognizing we face an accumulation of challenges and barriers, and there is no silver bullet that will lower them all. Along the way, we've never wavered from our ultimate objective: delivering safe and reliable combat-credible capabilities at speed and scale to America's warfighters so they can deter aggression and win if called to fight.

Why the urgency? Because the PRC has spent decades building a modern military design primarily to do one thing: overmatch us. But the one advantage they can never outmatch, steal, or copy because it's embedded in our people is American ingenuity. Our ability to think freely, innovate, change the game, and in the military sphere to imagine, create, and master the future character of warfare. Our starting position is stronger as a free and open society of blue sky inventors, doers, and problem solvers. We don't seek to control innovation or make it toe the party line. Instead, we aim to seed, spark, and stoke the flames of innovation. And with so much happening outside of DoD, that requires better adopting innovations wherever they add the most military value.

An important way we do that is by bridging the much-discussed valleys of death. Valleys, because there are several. Lab to prototype, prototype to product, product to scale. There are still others, and the valleys can be especially treacherous to cross if you're new to working with DoD. Now, these valleys are necessary for any healthy innovation ecosystem. Not every idea or prototype should scale, because not all of them work for the warfighter. And we need more creative prospects coming in that are likely to make it out, so that we push the innovation edge and avoid groupthink. It's similar in Silicon Valley. There's a reason why "fail fast and iterate" is a mantra. Not every founder who pitches a VC gets funded. Not every company with a series A round makes it to series B and C and so on. Only some get bought or go public. We've all met plenty of good idea fairies, but I don't know anyone that's always right. So our goal isn't for everything to cross the valleys of death. Instead, it is to get the right things across them to the right people at the right time. That requires a fast-moving cycle from identifying key capability needs to effective solutions, a warfighter-defined investment funnel, if you will, comprising novel operational concepts, prototyping and experimentation, expeditious acquisition pathways, open doors for newcomers to enter, and a more level playing field. Every piece should be constantly iterating, moving rapidly, responsibly, and securely, because if something doesn't work or isn't secure, it's not useful for the warfighter. It loses its value. And when something does prove valuable, we need to be able to produce and deploy it at scale and speed.

Let me tell you some of what we've done in the last three years to get at this problem, and I'll do it in three minutes. We've been repeatedly iterating on novel operational concepts for joint warfighting. We have new processes to incentivize and accelerate promising joint capabilities and experiments to advance those operational concepts, shrinking their time crossing the valleys of death by as much as two years. To bridge the lab to prototype and prototype to scale valleys of death, we're using more flexible acquisition pathways for rapid prototyping, rapid fielding, and software development. Over $35 billion have gone through these pathways since 2021 across nearly 200 programs. They've shaved up to six years off transition and delivery timelines for warfighter priorities. To work better with commercial and nontraditional firms, we're embracing innovative contracting tools that can be easier for them to navigate. We also sped how we transitioned the most promising capabilities to scale by overcoming bureaucratic and cultural barriers that slow us down. We knew all that was necessary because of our work to map and then methodically debug the whole DoD innovation ecosystem, a nose-to-grindstone effort that, like software development, never ends. As one of the world's largest organizations, we must always look for ways to get out of our own way.

Beyond that, we issued data decrees that all DoD data must comply with because data interoperability, access, and trustworthiness are critical for AI, for doing command and control across all domains, and for being a modern defense department. No excuses. We also affirmed ethical AI principles, updated our decade-old but still world-leading policy on responsible use of autonomous weapon systems, and issued new strategies and implementation plans on data, analytics, AI adoption, and responsible AI. Simultaneously, we focused on talent and personnel, from the ADA initiative that deploys data scientists to every combatant command, to continuously upskilling our acquisition workforce. We also brought focused leadership to critical organizations and stood up new ones. We elevated the Defense Innovation Unit to report directly to the Secretary so it can help focus and accelerate how we leverage the best of commercially derived tech. DIU is now led by a former direct report to Apple's CEO. We integrated disparate digital AI and data analytics teams under an empowered chief digital and artificial intelligence officer who previously ran machine learning at Lyft. He also reports directly to Secretary Austin. And we created the Office of Strategic Capital to partner with private capital markets and catalyze investment in technologies critical to national security. Across the board, we're making smart, serious investments. In terms of real dollar value, our 2024 budget request would make the second largest investment in defense R&D plus procurement since 1952. We'd send industry clear demand signals with things like multi-year procurement for key munitions, funding to expand industrial-based facilities, workforces, sub-tier suppliers, and more secure supply chains. Meanwhile, throughout, we're ensuring our defense dollars deliver by providing effective oversight for the taxpayer.

Our three minutes are up. And I get it if your eyes glazed over. Not all of this is headline-grabbing, but remember, there are no silver bullets. All of it is absolutely necessary to drag DoD into the modern era. Our efforts are fundamentally resetting behavior for defense innovators, program managers, resource leaders, and decision-makers. And even though it's collaborative, that kind of disruption can still be uncomfortable. Take Replicator, one of our most recent initiatives. Replicator's goal has always been simple and straightforward, to field thousands of attributable autonomous systems in multiple domains within 18 to 24 months, and to prove ways to burn down risk and rapidly and safely overcome barriers to scaling. It's about showing ourselves and our adversaries that DoD can move fast to shape the battle space and equip our warfighters with what they need. What we did in five months normally takes DoD two to three years. If you're not sure what is more mind-blowing, how fast we did it or how long it normally takes, I don't blame you. Honestly, the length of our NOAA process should blow your mind. It's because our system for buying new capabilities was designed in that 1960s era of DoD innovation dominance and reinforced after the Cold War when our leading conventional military capabilities seemed unbeatable. So most things still start with that inevitable, inexorable, multi-year budget process and take five to 10 years or more to field at scale. To go from start to fielding inside the two-year cycle is not normal. It's disruptive. But we plan to make it normal because more speed is essential. This is not the Cold War nor the post-Cold War era. With the PRC, we are in a persistent generational competition for advantage, and we have to double down with urgency and confidence.

Thinking differently does not come easy to large organizations, and the Pentagon is no exception. We still have many more pain points to address across DoD's innovation ecosystem. Institutional and cultural change takes time, constant tending, and consistent leadership. We've got to keep listening, learning, and iterating to continually become better customers and collaborators with the tech sector. So make no mistake, more deliberate discomfort will be required. More collaborative disruption will be necessary. The future of our nation depends on it, and it depends on you. Because American dynamism and American democracy are inexorably intertwined. Enforcing contracts and protecting IP depends on upholding the rule of law. Hiring a talented workforce depends on having good schools, universities, and immigration. The flourishing of all Americans depends on ensuring equal rights and equal opportunity for all. And starting a business, investing in others, inventing a product, and taking it to market depends on safeguarding the institutions that provide the blanket of liberty under which you do so. If we want to keep changing the world, then we have to strengthen the democratic principles that make this nation so worth defending and make changing the world even possible.

Our country is not immune from the authoritarian winds that sweep the globe. We have seen America routinely tested. While she has withstood, we cannot take that for granted. Institutions can be degraded. Belief in institutions can fade. Both endanger the health of our nation and your success. We want private sector innovation to succeed. We want American dynamism. We need you to feel the same about American democracy because neither can thrive unless the other succeeds. Many years ago, one of Silicon Valley's original garage founders, Dave Packard, spoke about his company's obligation to make some contributions to the defense effort in times of peace as well as in times of war. He said it was part of business institutions having a responsibility to the society in which they exist to do something more than simply make a profit. Why? Because in his words, we have freedom of action, which is the direct result of the American type of government. Packard said that nine months before the Soviets launched Sputnik, bringing tech competition to the forefront of America's Cold War consciousness. And he said it 12 years before he came to Washington to serve as Deputy Secretary of Defense. Like I said, our histories are bound together more than we sometimes admit. That binding enabled Silicon Valley to thrive, to drive successive eras of technological evolution and revolution, to change the world again and again. That's worth defending. America's defending. An American democracy is worth defending. And both are worth investing in. Thank you.

Host: All right, now let's pivot to Kathleen Hicks' fireside chat with Sharon Weinberger as they discuss collaborative disruption at the DoD, including lessons from Ukraine, the evolving role of primes, defense acquisition reform, and more. But first, let's start with where the dollars are flowing. This tremendous amount of private investment, venture capital going into defense. We recently wrote about the $100 billion that's gone in since 2021. But one of the concerns from those companies and from those investors is that it's not being matched yet by acquisition dollars. Is that a concern for you? And how do we actually change that?

Sharon Weinberger: Yeah. So first of all, I think that I really would hit that our procurement budget request for '24 is the second highest in the peacetime era in 35 years. So it's not that we don't have money going into acquisition. The question is, what are we acquiring? And are we acquiring what the warfighter needs? And are we acquiring everything we think the warfighter needs? Well, we've always had to make hard choices. There's no surprise that we will continue to have to make trade-offs. I think what I would stress is, as we innovate and move through the changing nature of warfare, again, shaping that changing nature of warfare, we know so much of what is in that shift is software-defined, and it also could be produced, manufactured much more rapidly through advanced manufacturing. There are a lot of areas, biotech, I could think of others, that are driven right now by the commercial sector. So as we shift in the nature of warfare, there's more and more opportunity for non-traditionals and service providers to be a part of that procurement story. So lots of dollars going to procurement. That's not the issue. The issue is, are we able to buy what we need? And are we thinking far enough ahead to what the shape of warfare needs to be?

Sharon Weinberger: So what in that view does the defense industrial base look like? And we could say five years, 10 years time, are there more than five primes? Are there a dozen primes? How does that actually look in terms of a successful change in strategy?

Kathleen Hicks: I don't think there's a pre-definition of number of primes. I think it is important to have competition in the defense industrial base. There's no doubt about it. And we know small business is a huge driver. So we work very hard on trying to drive up the percent of our procurement and investment that is going to those small businesses. I would try to guess here today how many primes there would be. I think the whole nature of what constitutes the defense industrial base, we need to shift that mental model. The American industrial base is the defense industrial base. We obviously have traditional providers we greatly value. We need them. We need them to stay at the cutting edge and we continue to work hard with them. But it's very clear to everyone, including them, that teaming with other parts of the economy, understanding the new areas of innovation, compute others, is incredibly vital to growing where we need to go in defense.

Sharon Weinberger: You talked about this long history of cooperation that goes back decades between Silicon Valley and the Pentagon. There have also been periods where there's been a rift, most notably after the Vietnam War period, and then there was a big effort to heal that rift. We've clearly come a long way. Are there still issues there?

Kathleen Hicks: I think that's a constant refrain for us, and we focus very much on, as a department across administrations, I think there has been a strong focus on making sure that companies want to work with us. We know part of that is making sure we represent the values of the nation. That's always been important for the United States in terms of bringing the commercial sector into work with government. So we continue to do that today. I don't think we are in a crisis phase, to use your Vietnam analogy. And I do think that the PRC did a lot of that work by comparison. And both their military activities and the types of coercive strategies they use working with companies have done a lot to turn others to looking to how you can work with us. I will stress one area that we've been very clear about is our responsible AI approach and a responsible autonomy approach. That's decades long, but we continue to evolve it as the technology evolves. We have Project Lima right now, which is looking at large language models. So we look at how to adopt innovations in a way that's safe for our warfighters and, again, reflective of American values. And we'll continue to do that.

Sharon Weinberger: You actually preempted my next question, which is on Lima, on that task force looking at generative AI. Where is it at? What do you see coming out of that? What are the issues it's looking at?

Kathleen Hicks: Sure. So they are looking at hundreds of use cases to identify what's the right framework for the department. This spring is when that information comes forward to the CDAO. And then to me and the secretary thereafter. So I anticipate getting some feedback on where there are use cases that are ready to go today, if there are those, and other areas where we need to do more work. We work closely with all the major generative AI providers and would-be providers to make sure we're lashed up. We're very enthusiastic to work with them. We just have to do it in a way that actually delivers for the warfighter.

Sharon Weinberger: There have been a lot of new technologies deployed in Ukraine. In some ways, I think some people have compared it to a test lab for things like UAVs, for drones. What are some of the lessons learned out of that from your perspective of the technologies that have been deployed there?

Kathleen Hicks: Sure. I think what I would stress for the U.S. really has been the proliferation of space, low-Earth orbit satellites in particular, and communications that the commercial sector really drove. That's been a clear, that ability to communicate and to have mass that's distributed has paid off. I think when you look at autonomy, it's a more mixed picture in terms of exactly how it has evolved, where it has provided benefit. And we take all of those lessons away. I think one piece that the United States has an advantage on is that when innovation happens down at the unit level, we know how to bring that up and scale it. We also need to make sure we do that exceptionally well and fast so that tactical innovations, TTP innovations that happen, software innovations that some very clever sergeant comes up with, for instance, which you see in Ukraine all the time, that sort of is lifted up and promulgated quickly across the force as we go through rapid iteration. So I think that's another lesson I would add. And then the last is intel in general. The United States clearly has an advantage in intelligence that's paid off substantially. And to the extent that we can help our partners with that, in this case, Ukraine, that's an area we know is a strength to build on. That iterative process that you talked about, particularly coming down from the tactical level. How do you get that into the U.S. system? It's an all-hands-on-deck. It comes at all kinds of different levels. We've really emphasized with the joint morphine concept, making sure that there's rapid concept to experimentation to fielding. Our radar initiative is one of those areas where we've really tried to highlight how do you quickly take a concept, make sure you have the right testing environment, experimentation environment, the lab, if you will, the substitute for Ukraine for us, the experiment and exercise realm for us out with our co-coms. And you see a lot of that innovation going on today, Southcom, Centcom, Indopacacom. Others are doing a lot of that work. That's how we do it, really make sure we're testing it in the field with the warfighter closely linked to the technologist. So the operator and the technologist work together to do that rapid iteration. A lot of this too is, again, concept of employment. So it's not just technology. It's about how do we use that technology best? And that's where we really can excel. We have the most incredible military service members, individuals who are incented to bring forward their best. That is definitely a different model than authoritarian states.

Sharon Weinberger: You talked about American ingenuity is where we have an advantage over China. It still very much feels like our acquisition process is the 300-pound gorilla. Can we get to where we need to be without fundamentally reforming Defense Department acquisition?

Kathleen Hicks: I think we have to fundamentally reform defense acquisition, but we are down that path. We did have the alternative acquisition pathways, middle tier acquisition pathways, for example. We have to prove those out. We have to show that those pathways, software is one, for example, we've already put billions of dollars through that software acquisition pathway. Now we have to show our oversight committees that we can deliver through those alternate pathways because there is a different oversight model built into those, and it's really about trust between Congress and the executive branch to prove out that we can do good things with these tools that they've given us. So that's why we've been so focused in this administration on taking all those authority changes, showing we can advance against them. Replicator is an example of that. And then where we can build that trust, I think there's more opportunity to expand our authorities. So, yes, we need acquisition reform. We're in acquisition reform now. And we need to build trust as we go through that.

Sharon Weinberger: Are you surprised where we are now with tech in Silicon Valley compared to five years ago?

Kathleen Hicks: I am. I think it is a marker for what's happened, I think, in broader world events. I think there's more ability to see how you as an individual innovator or investor can contribute to the American experience, to American dynamism, dare I say, and more concern about what the alternative might hold. We've seen that in Hong Kong. We've seen that elsewhere. And I think that more than anything has brought to mind what the risks are for falling behind. It's a great way to wrap up. Thank you so much.

Host: All right, you can find all of the above at a16z.com/ADSummit. And we'll include a link in the show notes.