Pod Insights a16z Podcast a16z Podcast | What's Next for Technology and National Security? About this episode We live in very interesting times, to say the least -- whether it's a shift in how technology is built and adopted today compared to the past; a changing international landscape with leapfrogging players; or an increased cyberattack surface as computing and networking touch everything.
So what's next for technology and national security? This episode of the a16z Podcast is based on a conversation that took place last month between Marc Andreessen and Michèle Flournoy -- former Under Secretary of Defense for Policy and co-founder of the Center for a New American Security -- moderated by Matt Spence, partner on the a16z Policy and Regulatory Affairs team (and a former Deputy Assistant Secretary of Defense). It covers everything from technology procurement and the tyranny of the inbox, to the politics of industrial policy and ethics debates around use-cases for new technologies.
But... do we really want innovation, not just in the abstract but in the specifics? If so, how do we think about the future? And how can both policymakers and technologists work together in different ways to help the U. S. keep its competitive edge and "give the future a seat at the table"?
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Sonal? This episode of the A6C podcast covers technology, national security, and the future, and is based on a conversation that was recorded last month between Marc Andreessen and Michelle Flornoy, former undersecretary of Defense for Policy and co-founder of the Center for a New American Security. The conversation, which follows a couple of minutes after the introductions and is moderated by A6C partner Matt Spence, covers everything from technology procurement to the politics of industrial policy and ethics debates around use cases for new technologies to the changing international landscape for security.
0:33 Matt Spence? Before coming to the firm, when I was working at the White House of the National Security Council and as deputy assistant secretary of defense, I had the great privilege to work with Michelle Flornoy. We have a term in formal policy language that we use for someone like Michelle. It's called a total badass. She really is one of the most inspiring and impressive public servants I've had the chance to work with, and it's really an honor to have her come today. I want to introduce my friend from Washington, Richard Fontaine. Richard is the president of the Center for New American Security. He was a top foreign policy advisor to Senator John McCain, and he was one of the most respected and thoughtful commentators on national security today.
1:13 Speaker 02 Well, thanks, Matt, and thanks to all of you for being here. To introduce Michelle, as Matt said, badass is her formal title, but her informal title is chief executive officer and co-founder of CNAS, the Center for New American Security, a national security think tank in Washington, DC, where the only national security think tank that is co-led by a Democrat and a Republican. We have a strong emphasis on building the next generation of national security and foreign policy leaders and developing innovative solutions to some of the problems and challenges that face the world. Prior to that, Michelle was the undersecretary of Defense for Policy under Secretary Gates and then Secretary Panetta at the Pentagon, and has had a long and distinguished career in
1:58 Speaker 02 government and in the intellectual world on matters pertaining to national security. So it's a pleasure to be able to introduce Michelle, and it's a pleasure for us to be here this morning.
2:11 Matt Spence? I now like to introduce Ted Ullyot. Ted was the general counsel of Facebook. He served in senior roles in the Bush administration of the White House and Justice Department. Now he heads up the Policy and Regulatory Affairs Team here at Andreessen Horowitz. Ted, we'll come into talk a little bit Andreessen Horowitz and introduce Mark.
2:27 Speaker 03 Welcome, everyone, and introducing Marc Andreessen at an Andreessen Horowitz event. So, Mark was born in Cedar Falls, Iowa. He was raised in New Lisbon, Wisconsin. Kidding. He needs no introduction here at Andreessen Horowitz, the firm he co-founded in 2009 with Ben Horowitz. Mark and Ben have been a leader in the growing movement these days to get Washington and Silicon Valley more engaged with each other, including but not limited to forming a group here at Andreessen Horowitz to focus on precisely this with respect to the firm and our portfolio companies, namely the policy group. Given Mark's background and Michelle's background, this should be a very interesting discussion
3:12 Speaker 03 about issues of great relevance to both Washington and Silicon Valley.
3:16 Matt Spence? Again, Michelle and Mark, thank you for doing this. I think it would be a profound understatement to say we live in interesting times. Michelle, one of your main responsibilities at the Defense Department was preparing and anticipating future threats to the US national security. In the hundreds of hours you spent in the Situation Room, you dealt with Afghanistan, you dealt with cyber efforts, you dealt with Asia. So, I'm wondering, right now, to kick us off, what keeps you up at night most? What are you most worried about?
3:42 Speaker 04 You know, I actually, the thing that worries me most is just the sheer complexity and dynamism and volatility of the international security environment. You know, we have, there are so many different threats and challenges to worry about, whether it's a resurgent Russia, a rising China that may or may not play by the rules, evolving threats of terrorism, cyber attacks, the list goes on and on. And some of these actors are very opportunistic. Vladimir Putin is probably the best example, Kim Jong-un is another. They actually look for opportunities when we are, when they perceive us to be distracted, otherwise engaged, and so forth. And if you look historically...
4:25 Speaker 04 Every single presidency since World War II, some major policy crisis has happened in the first six to 12 months of administration, usually unpredicted, not the thing you expected. So I worry about the potential for opportunistic actors to take advantage of a transition period, a period where an administration is getting its feet underneath itself, present us with some kind of crisis.
4:52 Matt Spence? One of the opportunistic things is Russia, of course.
4:56 Speaker 04 Yeah.
4:57 Matt Spence? One of the profound, unique things that a state actor potentially trying to interfere with our own political system. Could Russia actually influence our elections?
5:04 Speaker 04 Well, I think it's a possibility. I think what we've seen in the cyber attacks that have already taken place is a broader effort to try to see embarrassing or, you know, controversial content released in a way that might influence the outcome more indirectly. That's what we're seeing with the WikiLeaks. It's pretty accepted now from the intelligence community that these leaks were directed by Russia.
5:33 Matt Spence? Mark, as Michelle said, this is an example where America's technological inefficiency may actually help us, what we should be worried about with some of these national security threats right now.
5:42 Marc Andreessen? Question that just nags at me a lot. So, you know, there's, we talk about, long conversations about cyber, online security. Certainly, the issues are rising in importance. They're now rising to levels of ever greater importance. We've talked about things like tampering with elections, and we talk about things like electrical grid. From the valley standpoint, and this is a question, by the way, not a statement, it appears that in the world of military affairs and policy in Washington, if there's kind of real-world security, and if somebody physically attacks one of our ships, we respond with physical force. And then it feels like cyber is this whole separate thing. So if somebody cyber hacks us, the retaliation, the deterrence of the retaliation is going to be a return cyber attack. The question I've always curious about is, when do they combine? What level of cyber attack, in either direction, right, rises to the level where it would
6:26 Marc Andreessen? be considered an act of war, and you would actually retaliate with physical force? Or conversely, what level of physical activity would result in offensive cyber operations? How do these two kind of sides come together? Because at least it feels right now they're pretty separate.
6:38 Speaker 04 I think they are very separate. In my view, the policy and concept development.
6:44 Speaker 04 Side of the conversation is not kept pace with the technology and what we're actually experiencing in terms of cyber activity. And so, we really don't have a conceptual frame for thinking about how to respond. There's also the added problem that most of the policymakers don't fully understand the technology in detail. And so, they're tempted to apply previous frameworks in a way that doesn't really fit. So, they'll harken back to the nuclear domain and, you know, concepts of escalation dominance. And even offense-defense gets pretty complicated in the cyber domain. First, there's a gap to be bridged in really educating people who are going to be making
7:29 Speaker 04 policy in this area on the technology and vice versa. Make sure that there are some technologists who can engage in the policy discussion. Secondly, we've got to catch up with our conceptual thinking. I think what's slowed down the response thus far is worrying about not the first move, but the second and third and fourth, the response. How do you avoid getting into an escalatory situation? If we respond in certain ways, then Russia responds and we respond, we end up making matters actually worse. So, this is an area where a lot of work needs to be done. I agree with you that you don't always want to be thinking and responding in symmetric terms.
8:11 Speaker 04 There are areas where the best response will be some integration of what we're doing in cyber with what we're doing in the physical world.
8:19 Marc Andreessen? In the wake of the Sony attacks, presumably for the moment that they were from North Korea, from the commercial world wondered if the response to the Sony attacks was to take down the North Korean internet, that didn't seem like a very strong response.
8:30 Speaker 04 Yeah, right.
8:31 Marc Andreessen? North Korean internet is not a hotbed of information exchange.
8:36 Speaker 04 Exactly.
8:38 Matt Spence? Mark, you spend your days looking at what's new and it's not just cyber, but it's autonomous technology, it's things like that. So as you look at the new technological landscape, what do you see as the most important tech trends for some of these national security issues that Michelle's talking about?
8:52 Marc Andreessen? It's just this on-rushing, continuous impact of all these new network technologies. It's the continuous spread of chips being put in everything. It's everything being connected onto the network. It's information being gathered in unprecedented ways. It's information technology being woven into every part of our lives. And all of these technologies are like tremendous leverage for us as individuals. We're much better off as people with a supercomputer in our pocket, which also happens to be a universal surveillance device. And so the benefits are gigantic, but they do introduce all kinds of new risks. If you use this term, like the surface of the things that have to be defended keeps expanding. And then the form of attacks keep expanding. One of the things I think about is these new technologies historically arrived in a way that was relatively friendly to the US, in particular to large government organizations like the Defense Department.
9:35 Marc Andreessen? In fact, many of these innovations were first funded by the Defense Department originally. But new technologies, 50 years ago, or even 30 years ago, or even 10 or 15 years ago, would arrive... When they would first arrive, they would be extraordinarily expensive and difficult and complex and challenging to deal with. And so the largest organizations in the world had the easiest time implementing them. DOD was a prime example of this over many years. When computers came out, they cost an enormous amount of money and literally only the DOD and a small handful of big companies could afford them. When the internet first came out 20 years ago, the military and the national labs were networked way before private citizens ever had access. And so you had this kind of top-down approach to new technology adoption that really favored large centralized institutions. These days, more often, it seems like new technology is coming bottoms up.
10:19 Marc Andreessen? And so who gets the next new cutting-edge computer is not MetLife or the Army. It's you. It's when you go to the AT&T or Verizon store and buy the new smartphone. That's the new piece of cutting-edge technology. Who gets the next new piece of state-of-the-art encryption software? You, because you download it, it's open sourcing, you download it off GitHub. These technologies are now arriving bottoms up. That's extremely empowering to all of us around the world as individuals. But it does mean that in many cases, smaller organizations can now adopt new technologies faster than large organizations. I think it was Napoleon who said that in military affairs, God is on the side of the big battalions. There's historically this giant advantage in defense towards mass, right? Towards mass and applied force at scale, which meant money and material and men in large numbers.
11:01 Marc Andreessen? To the extent that we're shifting into a world in which the attacks are taking place by a rogue North Korean intelligence unit operating in Japan and able to launch cyber attacks. By the time we figure out where they are, they move somewhere else and they're launching something else. And we feel 10,000 people to go try to figure it out. By the time the 10,000 people get organized and figure out how to procure the new technologies and get everything implemented, the attacks have evolved four times. It seems like the asymmetry of scale being an advantage might be flipping now to speed being the advantage in these new technological areas. And if so, is that, to what extent does that actually empower smaller adversaries in a historically new way?
11:38 Speaker 04 It absolutely empowers not only smaller adversaries, but adversaries who might not be able to take us head on and win in a conventional fight. There's huge incentives to find asymmetric means of going after US vulnerabilities that undermine our strength. As you said, in many technology areas, the cutting edge is coming out of the commercial domain, from the bubbling ferment of places like Silicon Valley and elsewhere. And so the race is now about access to that technology and integration. A lot of the same technologies will be available to not only potentially individuals and terrorist groups or hackers, but also other countries.
12:20 Speaker 04 The name of the game now is how fast can we integrate it and leverage it and keep up with it? This requires wholesale change in the way a place like the Department of Defense thinks about acquisition, thinks about concept development, thinks about technology integration. Frankly, this is a project that I think Secretary Carter should get great credit for shining the light on and trying to start down this road. But it's really going to mean the difference about whether we keep our military technological edge in the future or not. To me, this is probably one of the most Important challenges we're going to have to deal with. And you can't do it without some kind of partnership with the tech world.
13:01 Matt Spence? One of the challenges that we talked about of proliferation of new actors, new areas of conflict is China. Michelle, you recently described the rise of an increasingly powerful, capable, and confident China that appears bent on becoming a dominant power in Asia. As China is rising and trying to do more, what are the advantages the United States has as we face China undergoing a pretty dramatic change from the national security front?
13:27 Speaker 04 I think the principal advantage we have in Asia is that we're seen as a force for good. It's not true in every region of the world, but in Asia it is. I mean, in the post-World War II period, we architected the rules of the road that provided a huge degree of stability, a foundation of stability, on which tremendous economic growth and development occurred, great deal of political development occurred. And I just came from South Korea, where they literally credit the United States with having given them the foundation, first, to become a first-world economy and, second, become a vibrant democracy. And so there's a lot of goodwill.
14:07 Speaker 04 We have also played our cards well across multiple administrations, Democratic and Republican, to build a real network of strategic alliances. You go to Korea, Korea pays more than 50% of the cost of our troops there. It would be more expensive to bring them home and house them in the United States. But it's missing the point. The real value of the alliances is the fact that we have a set of countries across the region who come with us when there are crises to be managed, who buy into the rules of the road that we're all trying to develop and enforce. And it's a huge, huge strategic advantage. I think China right now is its own worst enemy.
14:52 Speaker 04 The mask has come off. There used to be a hide and bite strategy. Now that China is feeling more empowered economically, it's starting to be more assertive in everything from the maritime domain to the cyber domain. And it's scaring other countries in the region. What's happening, with a couple of exceptions, is countries are coming to the United States and saying, we understand our economic dependence on China. They're our number one trading partner in the region, but we are afraid of how they're going to use that power, and we're afraid of being coerced, we're afraid of being manipulated. We want a closer security relationship with the United States. That's a tremendous advantage. The name of the game with China is going to be trying to incent them to buy into the rules-based order, to adapt it in appropriate places where we need to, but not create a
15:37 Speaker 04 completely alternative regional architecture and structure that competes with the system that has served us very well for several decades.
15:45 Matt Spence? Yeah, it's that duality you mentioned. China is an economic trading partner and superpower, yet rising security threat. And of course, one part of that security threat, Mark, would be the technological piece. As we look at China, they're trying to advance in quantum computing, they're trying to advance in virtual reality, they're building an airstrip in disputed open seas. As you look at the technological piece, how do we compete with that? We look at what we're doing here with what China's trying to do in the state-directed way to leapfrog the technological disadvantage.
16:15 Marc Andreessen? So I think there are two big concerns that Americans have about China when it comes to technology and economics. And I really strongly disagree with one of them, and I really strongly agree with the others. The one I disagree with, you'll hear this a lot from people in the valley, in the business community, that China has a very specific top-down industrial policy. They have decided that there are technologies that are extremely important to their national interest. The government is extremely hands-on in setting up frameworks and structures to develop those technologies, up to the point of subsidizing, directly subsidizing the companies that build the new things that are in favor. The government greases the skids for companies doing these new things, whether it's in bio or chips, or many of these other areas, sensors, quantum computing. And it's sort of this top-down organized effort. And this is where people start getting, it might be a little extreme, of like, you don't
16:58 Marc Andreessen? understand, these Chinese leaders are all engineers, they're all really smart, they're just going to organize everything, it's just going to all happen, and if they have to move a city of a million people to create a new testing strip for drones, they'll just do it. And in the US, we're a giant mess, and we can't coordinate anything, and our government doesn't run an industrial policy, and so we're therefore fated to fall behind. That part I don't agree with, and the reason is because industrial policies have a long history of failure, because the problem is they get politicized. Twenty-five, thirty years ago, you heard these exact same arguments about Japan. Japan had a centralized government policy about innovation, and Japanese executives, I think, would agree with this, led the Japanese economy right off the cliff in the early 90s, because they doubled down on the wrong technologies at precisely the time when everything changed with the arrival of digital, and the Japanese economy is still struggling to recover 20 years later. So I'm not as worried.
17:43 Marc Andreessen? I think in the long run, it counts as an advantage to us that the Chinese are so state-directed on some of these things. The other concern I'm much more concerned about, which is related, goes back to the idea of, in this case, we're our own worst enemy. Do we want innovation anymore? Do we in the US want these things to happen? Whether the government orders it to happen or not, do we want self-driving cars? Do we want autonomous drones? Do we want unbreakable encryption? Do we want ride-sharing? Do we want Airbnb? Do we want search engines? Presumably, we want search engines, except there's a lot of political pressure for this so-called right to forget, where if somebody says something bad about you, you get taken out of the search engine. So do we want there to actually be a way to find information, or do we want it to be able to be dropped on the memory hole? And so, and everyone in the abstract says, oh, yeah, we want innovation. In the specifics, more and more people are saying, oh, no, wait, all this stuff was all fine and good. We're glad you guys came up with the wheel. That was good. Fire was good.
18:28 Marc Andreessen? We'll grudgingly concede that PCs were probably good, but, you know, this whole Snapchat, Tesla, Uber, like, I don't know. Our companies hit this, and, you know, it's a big part of the regulatory work that we do here is to try to help our companies navigate through this. But I think it's also a broader social-cultural question, which is, how do we think about the future? That's where I worry, because I think there's no question that modern Chinese culture is completely enthusiastic about leaping into the future. The Luddite fallacy is back. The idea that technology destroys jobs, the core fallacy that's been wrong for hundreds of years, even that's back. But I think it's a social-cultural question that will end up having a big impact on how these issues play out on the international stage.
19:01 Speaker 04 I also think, in this country, we are likely to have important ethical debates about things like fully autonomous military systems, or the use of genetic modification to create sort of superhuman capabilities in soldiers or what have you. In a way that may be truncated or altogether bypassed by some of our potential future adversaries. So it's a positive thing that we'll have those ethical discussions in this country, but we can't count on others to have them or to end up in the same place. That worries me.
19:36 Marc Andreessen? Well, we also can't count on them to happen. We also can't count on them to make sense. So
19:40 Speaker 04 That's true, too.
19:41 Marc Andreessen? There's this argument in ethics and science and technology, this argument that like scientists and technologists need to take more moral responsibility for the implications of their inventions. If you go back in history and you look at when new inventions happen, it's precisely the scientists and engineers working on those inventions who are the least able to predict what was actually going to happen. Not that anybody else was either, right? As an example, famously, Thomas Edison invented the phonograph, which was a fairly good idea. I would like to believe we would have funded that. His use case for the phonograph, it was 1888, something like that, was that everybody would be able to have a library of recorded religious sermons in the home. So you get home at night and you want to unwind and relax, and so you put on a phonograph of a preacher screaming at you for 20 minutes. That was the use case, religious environment. Music was nowhere on the list. Even just the implications of something as straightforward as that,
20:24 Marc Andreessen? right, really uncertain. Would you have strangled the phonograph in its crib because you would have had this debate about the pros and cons of religious instruction versus music? And that's true of many of these, you know, many of these new technologies. By the way, it's not that they're all just positive. They do have negative impacts, but I think it's an open question how well we can anticipate the impact. So the idea that we can somehow have the debate and make logical decisions itself is an interesting question.
20:44 Matt Spence? So what would you like to see from each other? Michelle, as we talk about both like the ethical principles of this new innovation, implications, what would you like to see Silicon Valley doing and thinking about that they're not? And Mark, what would you want to see the Defense Department, the largest IT buyer in the world, 50% of the US government's already, what would you want to see more from this side about these issues that you're talking about?
21:04 Speaker 04 I think the first thing is to really help people in government and policymaking understand What you do know about a technology and its applications and where you think the future path may be and to have an ongoing, interactive dialogue. I think part of the reason that dialogue doesn't take place is that we've grown very litigious in the US system. People are very worried about any government official who may get near an acquisition process, talking to a technologist or an industry rep without a lawyer in the room. One of the things we found doing at CNAS, convening dinners where you bring the service chiefs together with different technology experts and industry people, is they just
21:49 Speaker 04 say, "Thank you." Because this is the first time I've been able to describe my problem set and have someone sit at the table and have an open discussion about what technologies might or might be applicable and what might be possible in the future without having someone looking over my shoulder saying, "You can't say that because that will bias an RFP or that will bias the competition for a contract." We've structured our system in a way that's really disrupting the conversation that needs to be had. I don't envision an approach where you impose a lot of policy constraints on the development of technologies, but I do think, you know, helping the policymaker think through the choices.
22:30 Speaker 04 I think the first places is going to come up in the military context is going to be on autonomous systems and how fully autonomous you want them to be versus where the person in the loop needs to be inserted and under what conditions and situations.
22:47 Marc Andreessen? If I could answer the question in one word, it would be experiments. It goes back to how I started, which is very consistent with what you just said, that the new technologies, the new technologies used to drive top-down, they were major decisions about implementing new systems, major budget implications, major systemic change implications, organizational implications. The new technologies arriving bottoms up means that, first of all, you can run experiments. Like, you can run point experiments with a small group of people, or with a small military unit, or with a small intelligence unit, whatever. Somebody in the field, somebody locally. It doesn't need to be a $50 billion weapons system procurement, it could be a million-dollar beta. The difference in that, the number of experiments you can run is really big. And so, from an adoption standpoint, I think experiments end up being key. And we also see big companies, large organizations that are in the corporate world, also taking
23:31 Marc Andreessen? increasingly the same experimental approach. As you say, the regulatory framework that's in place now makes that very difficult, because the framework in place today anticipated that these are large top-down decisions.
23:39 Speaker 04 Right.
23:40 Marc Andreessen? When in reality, more and more, they're bottoms up.
23:41 Speaker 04 Yeah. This is where creating an acquisition stream that actually uses existing authorities that aren't exercised very well, enables you to have a much more diverse portfolio of investments. Not everything that you're going to bring to full production, but that you're going to prototype, you're going to experiment, you're going to try things out, you may acquire small batches of one generation and then move to the next. That allows that agility and that learning process, rather than a massive top-down 20-year procurement. We cannot try to procure IT-driven solutions the way we procure aircraft carriers or fifth-generation fighter jets. It just doesn't work. It's a recipe for being obsolete before you even fully feel the capability.
24:26 Matt Spence? This gets to something you would say often in government. Even after you left the Pentagon, people repeat it again, that we need to give the future a seat at the table. You said this even when the US government was undergoing a self-imposed budget-cutting situation known as sequestration. From where you sit right now, where are we not giving the future a seat at the table as we think about either the Pentagon, national security? What more do we need to be doing on that front?
24:49 Speaker 04 I mean, the first aspect of giving the future a seat at the table is intellectual bandwidth. You know, we tend, senior decision-makers, I used to talk about the tyranny of the inbox. You're trying, Matt, you can speak to this, you're trying to survive your daily schedule, and you're going from crisis meeting to crisis meeting. About three or four months into the policy job in the Pentagon, I realized I have yet to have a meeting, a strategic meeting on China. Uh-oh. What's going on there? The vice chairman and I joined up to create a demand on the schedule to have these regular deep dives on where are we going with China. So the first is intellectual bandwidth. But secondly, you have to have the resources to invest in the future, to take your best
25:30 Speaker 04 and brightest and get them involved in experimentation and concept development. I've come to believe that the most important thing the next Congress could do for U.S. national security is negotiate with the new administration, at least a four-year budget deal. That would create a predictable top line, that would end the threat of sequestration, that would allow us to have multi-year investment horizon to start placing bets about the future. When you're living from continuing resolution to continuing resolution and you don't know what your budget horizon is beyond a nine, 12-month period, it's terrible. If we continue on that road, we will not keep our military technological edge.
26:13 Speaker 04 The next 10 to 15 years is the critical period that will determine whether the U.S. maintains that as an unbelievably valuable asset or whether we lose it in some areas.
26:25 Matt Spence? Mark, everyone talks about what Washington can learn from Silicon Valley. Is there anything that Silicon Valley can learn from Washington?
26:32 Marc Andreessen? There is a critique of Silicon Valley that I at least somewhat agree with. We are relatively insular community. We live in a special place and special time and socioeconomic kind of bubbles of words I don't usually like, but there's a little bit of that that's true. One of the things that's just so striking about the U.S. government and the U.S. military intelligence community is just engagement all around the world and a pretty deep understanding of how things actually are, where most of the people actually live. I've always just found it instructive when I work with senior folks in either defense or intelligence, how much they think about what happens outside our borders, how people react, how people perceive, how people are going to make decisions who aren't here. I think that's something that the government can really help with the value with.
27:10 Matt Spence? So what do you think? Is there anything from the time you spent out here you think Silicon Valley can learn from Washington?
27:14 Speaker 04 Probably the most positive thing that you find consistently across people who work in Washington, from the military to the civil servants to the appointees who come in, is a real mission-driven focus. Almost everybody could... Make better money and make a better living doing something else. But people serve because they feel they want to make a difference for their country. They want to make a difference in the world. And actually, I find a fair amount of that mission orientation out here. It may not be as specific in the way it's articulated. But one of the things that I've found encouraging is there are a lot of people here that, once they get to a certain point, actually want to figure out how they can contribute in the public sector.
27:58 Speaker 04 What you really want is a superhighway where people from the tech world can come in, do some period of service, make a difference. There's the US Digital Service. You need to really grow that. And then you need people, similarly, from government, to have a chance to come out, work here for a couple of years, understand what innovation really looks like, what agility and experimentation really look like, and then bring that knowledge back in. I think that cross-fertilization opportunities are huge and would benefit both sides.
28:26 Marc Andreessen? And there are very small programs. We had an Air Force officer on staff at my company, Netscape, as part of a sort of exchange effort.
28:32 Speaker 04 That's one guy.
28:33 Marc Andreessen? Right, very small.
28:34 Speaker 04 That will be your only one you get for the next 10 years.
28:36 Matt Spence? You're still talking about.
28:37 Marc Andreessen? I assume that program continues. I should know this, but I literally haven't encountered it since. We would welcome more people coming out, either from civilian or from the military side, spending time out here.
28:45 Speaker 04 Actually, you know, when you get to know operators, they're incredible entrepreneurs that have been trained up but thrown into battlefield situations where it's like my training didn't tell me about that. And they are constantly problem-solving and creating new things and using what they have, the toolbox they haven't been given, in creative ways. I actually think culturally, there's a lot more in common there than might be assumed.
29:13 Matt Spence? I want to leave some time for a few questions from the audience. Folks have a question, raise your hand, and we'll go. Right here.
29:21 Speaker 06 One of the things you mentioned was that in Asia, we have good reputation versus some other parts of the world. I presume one of those parts of the world is the Middle East, where America is not seen as a source of strength. So what are we doing long term to change that, if nothing else? And what are we doing strategically, as well as bringing technology, innovation, and entrepreneurship as sources of good? Is there hope in the next 10, 20, 30 years of that changing? Or are we doomed for the short time period?
29:51 Speaker 04 Yeah, I think the US has been guilty of great sins of commission and great sins of omission in the Middle East. Times where we have done things that really weren't very well thought out, and times where we should have been there and we haven't been adequately there. I think the next administration needs to start with some serious engagement with key allies and partners in the region to first really listen to what their concerns are, and that we want to try to work with them to get on the same page in terms of dealing with some of the challenges we share, whether it's terrorism, or destabilizing
30:34 Speaker 04 activities in the region, or what have you. Doesn't mean we're going to instantly agree, doesn't mean that our interests will always align, but I think what they're looking for is assurance that the US can be relied on as a partner, and that we have done our homework, and we actually have a strategy beyond trying to manage the eaches of where we're trying to go. I am not suggesting that the US can write the history of the Middle East. We do not have the pen on this. But our influence and what we do matters, and can impact things in very important ways. So I think there's a lot of work to be done there.
31:14 Marc Andreessen? Mark, are there any opportunities there? One of the really promising things, I think, is there is a spread of entrepreneurship culture around the world. Silicon Valley used to be a black box. The whole approach to what we do out here, I mean, when I was growing up in Wisconsin, I might as well have been on Mars. Like, I had no idea what into this stuff was. And now with all these new technologies, social networks, information flow that takes place, online courses. People all over the world are learning how to do what we do here and learning how to participate and either working with Silicon Valley companies from other locations or just starting their own companies and their own ecosystems. A good friend of mine, Chris Schroeder, is a former State Department official from years back, wrote a book that's become kind of the definitive book on this topic called Startup Rising, which is about the rise of tech startups in the Middle East. Everybody's familiar with the Israel story. He wrote the book about everything outside of Israel.
31:59 Marc Andreessen? There's lots and lots and lots of fascinating stories. There's a whole ecosystem that's building up that's extremely exciting. He actually, twice in the last two years, has traveled to Tehran as part of the very limited number of trips that people have been able to make. And it turns out there is a small but thriving tech startup ecosystem seen on the ground in Tehran. Very excited, fired up, actually, interestingly, heavily female tech entrepreneurs, starting companies and tech companies in Iran that are doing everything, e-commerce and social networking and instant messaging and all of these things. The optimism in me says, you know, that's a kernel of hope and optimism and a positive view of the future. And economic development, economic interlinkage, learning, growing, value of education. If we can foster that kind of activity and do things that support it, it seems like something we can build on.
32:41 Matt Spence? That's great. I think we'll have one final question.
32:43 Speaker 07 Thanks for having us. Mike McFaul from Stanford. It's striking how few people from the Valley work in the government. And my Stanford students, one of them's sitting with you right now. There are very few Stanford students who go to work in the government, right? They got so many other opportunities here, why would they go deal with the bureaucracy that gets nothing done? My challenge to you two, there's this really subversive program that the Council on Foreign Relations runs, where they choose the people, and then they pay for the people, and then they implant them into the US government. When I first learned about it, I was like, it was crazy. It's like, you don't work for us, but you're sitting here in the State Department? And they're like, yeah, I'm on a Council on Foreign Relations fellowship. Why don't the two of you guys do that?
33:27 Speaker 07 CNAS and Andreas and Horowitz? We're game if you're game.
33:34 Speaker 07 One-year fellowships.
33:34 Speaker 04 Thank you, Mike, and that was not a plan to comment.
33:38 Speaker 07 One-year fellowships, send some of your people to go spend a year in the government, you know, and really like technologists types. So, and then they come back. That's my challenge to you all.
33:47 Marc Andreessen? Without committing on the spot.
33:50 Marc Andreessen? It is interesting, and Michelle mentioned the US digital service, like something is happening there. And I think it's certainly worth exploring, doing more.
33:56 Matt Spence? Great, thanks. Well, before we get any more proposals like that, I think as we wrap up, I just want to again, thank Mark, thank Michelle, thank you for spending time with us. Thank you, everyone, for joining us. I look around, there are at least three former ambassadors, other senior officials from the US governments, founders of tech startups, executives from companies, and there's a lot of dialogue that would happen if we bring these two worlds together.
Previous Company Building in Crypto: A Conversation between Brian Armstrong and Chris Dixon Next What's Next for Education Startups Page updated October 26, 2026 · transcript and summary are machine-generated