Blog

Frontier Tech and the Geopolitical Future

GILES: Welcome to The Interconnect, a new podcast series from the Council on Foreign Relations and the Stanford Emerging Technology Review. Each episode brings together experts from critical fields of emerging technology to explore groundbreaking developments, what’s coming over the horizon, and how the implications for American innovation leadership interconnect with the fast-changing geopolitical environment. I’m Martin Giles, and I’m the managing editor of the Stanford Emerging Technology Review. 

Now, in previous episodes of The Interconnect, we focused on a specific field such as biotechnology, robotics or semiconductors. But in this episode, we’re going to take a different approach. Instead of diving deep into one area, we’re going to look across the entire landscape of emerging tech and explore critical drivers, or what the review calls cross-cutting themes, that impact all these fields. We’re going to discuss the key factors that influence the evolution of emerging technologies, the relative advantages of democracies and autocracies in developing frontier tech, and the central importance of talent and public and private investments in driving America’s innovation ecosystem. To explore these and other timely topics, I’m delighted to be joined by four terrific experts. From Stanford, we have Dr. Amy Zegart, who’s a senior fellow at the Hoover Institution and a co-chair of the Emerging Technology Review. And Dr. Herb Lin, who’s also a Hoover Fellow, and is Director and Editor-in-Chief of The Review. And from the Council on Foreign Relations, we have Adam Segal, who’s the Chair in Emerging Technologies and National Security, and the Director of CFR’s Digital and Cyberspace Policy Program. And Kat Duffy, who’s a senior fellow for Digital and Cyberspace policy at CFR. 

We’ve got a lot of ground to cover, so let’s dive straight in here. I think it’s fair to say that what makes setting policy for emerging technologies so challenging is that it’s particularly hard to anticipate how fast they’re going to evolve. Amy, I’ve heard you say that the nature of change in emerging tech is changing. What exactly do you mean by that? Can you unpack it for us?

ZEGART: Sure. So Martin, as I’m sure our listeners know, technology is always changing and it’s always affecting economics and geopolitics from the time of Roman aqueducts to nuclear weapons. But this time it’s really different, and it’s different I think, in three key respects. The first is there’s a convergence of technologies. So it’s not just AI and it’s not just biotechnology and it’s not just commercial satellites, it’s all of these things happening at once. The second element of change that’s different is the speed and scale of change. Gutenberg invented movable type in the 1440s. It took a century before we got printed newspapers after that. But look at the rapid adoption of technologies like AI, so that speed and scale is fundamentally different than it’s been in the past. And the third piece, and you alluded to it, Martin, is the difficulty even of experts at the forefront of these fields to anticipate the speed, scale and direction of change. It’s just much harder for us to know what’s coming next.

GILES: So that’s all of these technologies coming together at once and kind of it all evolving at a very rapid rate and then the fact that it’s really hard for experts to agree and predict exactly what’s going to happen. Adam, those three things, do they resonate with you?

SEGAL: They do, and I think we probably could add some more characteristics. One, often the breakthroughs happen in the private sector, so sometimes we don’t necessarily know where that’s going to happen. DeepSeek, for example, the breakthrough on AI, might’ve been known to a small number of China watchers and AI watchers, but was not widely known to people who might’ve generally understood the larger technology landscape. And second, I’d add where that innovation is going to come from geographically, that there are more places that have agglomerations of technology that are pushing the edge beyond just what we used to think about Northern Europe, the United States, Japan. Now we have just a whole new range of actors out there.

GILES: Great. And Kat, do you want to pick up anything else for the list?

DUFFY: One of the things I really appreciate about the report, and which it emphasizes, is also the interdependent nature of so many different technologies and how that impacts the degree to which one might speed or scale. And so you can look at a rise in artificial intelligence, for example. That shift in artificial intelligence may be transformative for material sciences, it may be transformative for robots, shifts in quantum may be transformative in other respects. And so I think the interdependent nature of those technologies is something that, in the policy space in particular, we tend to lose sight of and should probably be looking at a little more carefully.

GILES: And Herb, from your perspective?

LIN: Well, I should point out that the interdependence of technologies goes in multiple directions. As Kat said, AI is driving important advances in material science and robotics and biology and so on, no question about that. On the other hand, the reason that we have AI working in such a large scale at this point is because of the semiconductor industry, which provides important chips. AI people would be out of business if they didn’t have the chips. And of course, you could say, “Well, is chips the most fundamental technology?” Well, chips are made out of semiconductors and better semiconductors come out of material science. You see a kind of self-reinforcing loop there in which A helps B, and B can help C, and C can help A. It’s this kind of interlocking technological interdependence that prevents any one technology from being, quote, the most fundamental.

GILES: One of the other crosscutting themes that you pick up in the report is the risk of what I refers to as frontier bias. What exactly is that and why does it matter, Amy?

ZEGART: Frontier bias is the assumption that the most transformational technologies always exist on the frontier. Certainly, the frontier is critically important in geopolitical competition today, but there are all sorts of examples of transformational technologies that aren’t on the frontier. Herb likes to use a great example of shipping containers. Not particularly technologically advanced, but steel shipping containers revolutionized global shipping. That’s just one example of how lower tech innovations can lead to massive change.

GILES: Got it. Kat or Adam, any thoughts on that?

SEGAL: Well, I think it also biases where we think the international competition is, that we tend to focus on which country we think is at the technological edge or has made the most important breakthrough. But it may be that for long-term economic competitiveness, for social welfare, for job employment, that in fact, deploying, innovating on becoming better at manufacturing technologies that are one or two generations behind have a longer term effect.

LIN: Let me amplify that. The assembly line is arguably one of the most important, quote, inventions, unquote, of the industrial age. It enabled large scale manufacturing to happen relatively inexpensively. And yet, when the assembly line came along, it wasn’t any new gizmo that was invented, it was a different way of organizing work that created the assembly line. Not a new technology at all in the usual sense of the term. And if you want a more contemporary example, we have SpaceX promising to lower launch costs by a factor of 10, maybe even 100 in the future. SpaceX reusable rockets are the key to that, but there is no one technological breakthrough that gives you SpaceX reusable rockets. And in fact, the mathematics and the technology of that has been around for a very, very long time. It’s just small incremental advances in a variety of technologies have all come together now so that we can now integrate all these technologies into a functional booster that returns itself to the earth safely and can be reused. And that’s of course, the key to reducing costs by so much. But the idea that you would use a reusable booster to reduce costs, that’s been around. Everybody has known that for 50 years, that that was the way to do it.

ZEGART: And just to pick up on something that Adam said, which I think is really important, this difference between manufacturing and innovation. Just because a particular country, a particular company is first with something, doesn’t mean that’s going to be the lead because it’s about deployment of that technology as well. So we tend to think about competition as who’s winning the race, like there’s a destination, and that’s not really what’s happening today. You can be first to innovate, but slow on the deployment side, and it’s the deployment that may make the difference.

DUFFY: Amy, I would add on or I would say the way that I think about that is that there is innovation and there is transformation, and deployment is really where transformation sits. And again, in order for transformation to occur, you have to have social implications, political implications, policy implications that allow for that deployment. And so this is something that I’ve thought about a lot in our innovation economy, is even if we are coming up with some of the greatest technologies first, are we coming up with them and exporting them in a way that is actually going to allow American technologies to win a deployment race and be the technologies that are transformative on the ground, as opposed to, for example, a country like China? And this, Adam, goes right back to your point around competition. Precisely what is it that we are trying to win? And it may be that one thing we’re trying to win is early deployment, but the other thing we may be trying to win is a world in which American technologies that are based democratic norms and principles are the prevailing societal infrastructure, and not technologies that are going to be based more on things like surveillance capabilities.

GILES: Right. Kat brought up a really interesting point, and it’s kind of a good segue to shift gears here and look at the issues of the relative advantages and challenges that different regime types have in accelerating development, deployment, diffusion of frontier tech. Now, Adam, you’ve spent a great deal of time studying China. How would you characterize its approach here and has it changed much in recent years?

SEGAL: I think the report does a good job of capturing the kind of common sense understanding of the U.S. versus China. China is top down, great at setting goals, very effective in mobilizing resources and being able to lower costs over time as more and more Chinese firms enter the same space and compete intensely. And the U.S. side is more bottoms up, more innovative, more driven by creative individuals and the government plays mostly a supporting role. I mean, it funds federal R&D, but its larger role is to get the incentives right for entrepreneurship and for being able to commercialize that. I think in many ways, China was poised to actually have the best of both worlds. They were at one point, trying to marry both the top down and the bottom up approach together. You can look at some strategies, for example, the 2005 mid to long-term plan. At least the first half of that report is very top down focused and talks about indigenous innovation and the need for the state to help China reduce its dependence on foreign suppliers, the US and Japan in particular, and has a lot of big science and big tech goals. But the second half of that report reads like a pamphlet for Silicon Valley. It’s all about protecting IPR, encouraging spinoffs from universities. I mean, how do you create an environment that supports this kind of technological entrepreneurship? What we’ve seen under Xi Jinping is that that first half of the story has really dominated. They have tried to balance that still. We saw this month, Xi Jinping meeting with tech entrepreneurs after two years of control, but China has really shifted the pendulum much more towards the top down, state driven innovative model.

GILES: Amy, how do you think about this and the sort of contrast between U.S. and China models?

ZEGART: I think Adam put it really well. I often worry about the myth of innovation. And the myth is that it requires freedom to innovate, and that’s true only to a limited extent. When we think about China is going head-to-head with the United States, it is not a free society. But we have lots of examples throughout history where innovation occurred in very restrictive, very repressive regimes. Think about the Soviets during the Cold War and their innovations in nuclear weapons and other related technologies. Think about classified research in the United States, that’s done under pretty restrictive conditions. So I think we need to be careful about not making too much of the fact that free societies are necessary in order to innovate. I couldn’t agree more with Kat when she said it does matter whether a democracy is the leading country in technological innovation. It matters for scientific collaboration, it matters for standards, it matters for human rights, it matters for geopolitical leadership. But the source of innovation is just as robust in a repressive regime like the Chinese Communist Party, as it is in the United States.

GILES: I want to just explore some of the kind of non-technical factors that play here. I mean, I’ve often heard it said, “Oh, in China, it’s great for them because they don’t have real privacy laws. They can take all the data that they want and train all their models on it. They have a massive advantage here because of that societal attitude towards data.” That’s just one example. Kat, what’s your take on how these kind of non-technical factors play out across the democratic, autocratic spectrum?

DUFFY: Well, I’m going to lean into one that I think is being under-indexed or under-reported at the moment, which is the role that I think this current administration’s attacks on or reduction of, depending on how you frame it, foreign assistance and diplomatic footprints globally may really impact U.S. ability to be leading in transformation around the world in terms of tech. Between 2013 and 2022, China invested $679 billion in infrastructure across more than 150 countries through its Belt Road initiative. And its digital Belt and Road Initiative put in incredible dependencies on 5G, on internet connectivity, on AI. And the United States and China are the only two countries in the world that have comparable reach in terms of their diplomatic footprint and their foreign assistance footprint. So when this administration came in, we had a diplomatic infrastructure that spans 271 embassies, consulates and missions around the world, and a foreign assistance infrastructure that reached more than 170 countries with about $68 billion in foreign assistance funding in 2024 alone. On March 10th, Marco Rubio announced that after an extensive six week review, USAID would be cutting, they would be cutting 83 percent of USAID’s programs and canceling more than 5,000 contracts around the globe. It’s one thing to say foreign assistance programs don’t have much to do with the export of technology. It’s another thing to look at it and say, “Huh, we were doing foreign assistance contracts in 170 countries around the world and we just broke 83 percent of them.”

GILES: But does it have much to do with technology? Because I would say, “Well, so what?” I mean, that doesn’t affect the way that American technology is diffused versus Chinese technology, does it? The Belt and Road, I give you that. If it’s China that builds the 6G networks, you own those networks, you own that critical infrastructure. Is that more the risk here than something like foreign aid?

DUFFY: No, I think it impacts it in two ways. I think the first way is that when you break that many contracts on key issues to countries around the world, you are telling them that America is no longer a government or a country where they can rely on solid business practices, where people can trust that the United States will stay true to its word. And I think that the reputational impacts of that go beyond the US government, potentially to American business writ large, because you start to question the basic legalities and fundamentals of doing business. But the second, and I think more critical point, is that China also now has a significant entry in which to go in and fill a lot of those gaps, and they have already been demonstrating their capacity to do that. And I think there is a peril to these decisions that will impact American technological exports over the next three to five years, that I’m concerned that we may not be putting those dots together adequately in terms of the impacts.

GILES: Other views on this?

ZEGART: I guess I would just say I really have a different view, Kat. I mean, foreign assistance is a separate conversation. The question more broadly about the role of allies and partners in geo-economic competition is the central question. We should be asking questions like, how can the US and its allies and partners work together on digital free trade agreements? How can we work together in AI safety testing? How can we work together in ensuring that innovation is led by like-minded countries? Those are all vitally important, but USAID is not the long pole in that tent. You may have concerns for other reasons, but I do not think it is centrally related to technological innovation. And I think when there are so many challenges that the United States and its allies and partners are facing, we need to be really careful and targeted about what are the most important things that our government should be doing today to ensure that the United States and like-minded countries sustain leadership and technological competition? And USAID is just not, for me, at the top of that list.

DUFFY: Well, Amy, the one thing I would push back on there is that when I say foreign assistance, I’m talking more broadly than USAID. We’re also talking about bilateral programs for things like rule of law, for supporting law enforcement, for supporting human rights, for supporting free and fair elections. There’s a lot that goes into foreign assistance that isn’t just USAID programming. But I think to your point, I agree with you, the broader geopolitical competition or geopolitical relationship issue is the more critical one. Foreign assistance to me, is one sub-element of that. I just worry that it’s an unforced error to be violating those norms so rapidly and so completely because I think it diminishes trust and our ability to operate with our partners, with our allies, but also with potential partners and allies.

GILES: Adam, Herb, any thoughts on this?

SEGAL: Well, I think trying to kind of bring your original point together with the larger one about the international competition, I think it’s somewhat not true that Chinese firms have access to all the data that they want. There are privacy laws in China. They actually do a pretty good job of protecting the individual from a company, not from the state, but from the companies, Chinese companies. And when you read the Chinese press or you read Chinese journals about data and data collection, they all complain about their access to data. They complain about there’s lots of islands of data, they don’t have access to the data that they want. And we know that for one of the reasons why, besides compute and algorithms that US AI companies have jumped ahead of their Chinese competitors is that Chinese firms have great data about Chinese consumers, but not really anybody else. And the US companies have access to the rest of the world’s data, as well as the United States data, and they can use it for training for lots of other different reasons. So I think to the larger point that both Kat and Amy made, which is that one of the US’ great strengths in competing with China is being able to work with its allies and partners, both on hard tech and soft tech. And as Kat was saying, the norms around the usage of tech, as well as building things together and data collection.

GILES: If I go back to the beginning, we were talking about the accelerating pace of diffusion of these emerging technologies. There’s a clear trend there, and to try and hold that back, we’ve seen efforts like imposing export controls on the export of advanced semiconductors. How should policymakers be thinking about tools like that and setting policies that try to restrict, at least for some time, the diffusion of advanced tech? Herb, what’s your take on these?

LIN: Well, I’ve been actually involved in the export control issue for, I guess 30 years now or so, something like that. What we see widely across the entire panoply of export controls is that when they are applied to a major power that may not be friendly like China, they certainly have short-term effects of slowing down the progress of the other guy, but it spurs a number of efforts on their part, so they learn to work around the export control. If you set a limit, they’ll try to use stuff that’s below the limit. And Western manufacturers collude in that. We’ve had a discussion in this country, I think between Nvidia and the commerce department, in which the CEO of Nvidia said, “Well, if you put an export control level at X, I’m going to make a chip that’s X minus epsilon, and I can go ahead and it’s worth that,” and it becomes a cat and mouse game. But it also stimulates the development of indigenous production and expertise and so on to grow your own. And that has important effects too, because now what you’re doing is you’re creating a situation in which over the long term, the other guy has a capability that is comparable to yours on the stuff that’s been controlled. And there’s also suppressed a lot of computational technology exports to Russia, the Soviet Union then, and as a result, the Russians actually learned the mathematics and physics of the things that we were using compute for. And as a result of that, their understanding of the science behind the things that they were computing wound up being more advanced than ours. So export controls have all kinds of long-term effects that aren’t necessarily in our interest. I’m not saying this to say that export controls should never happen, I’m not saying that at all. But the point is that somebody ought to be understanding these long-term effects and making at least an informed trade-off about them.

GILES: Other views on the diffusion issue and how to think about that as a policymaker?

ZEGART: Martin, I’ll just say coming back from Washington, speaking to people on both sides of the aisle, I’ve heard with respect to export controls of semiconductors, export controls went too far, export controls didn’t go far enough, you hear every which way to Sunday. And I think the one thing that people would agree on is we don’t have enough data to know. One thing that Herb relayed is the time horizon issue. The other is, what’s the objective of export controls? Is it denying capabilities to another government? Is it delaying capabilities to another government? And if we’re delaying those capabilities, what are we doing with the head start? And so I think there’s a lot of, and rightly so, a lot of debate about export controls. And I would just say one other thing, which is particularly when you’re talking about AI technology, you cannot export control your way to victory. Because once an algorithm is out of the gate, you can’t claw it back. So one of the fundamental differences about a lot of the technologies that we talk about in the report today compared to technologies of the past, like fissile material for nuclear weapons, is you cannot restrict them. What they’re born unclassified, as Adam alluded to, they’re often invented in the private sector and they naturally diffuse. And so the idea that export controls will keep that genie in the bottle doesn’t work today like it did with the Cold War analogs like fissile material.

LIN: And just to foot stomp on that, that’s true, not just for algorithms, but for any kind of knowledge, will diffuse out. That doesn’t mean you can’t slow it down, but it will eventually diffuse out.

SEGAL: I’ll just add also that there’s third parties involved too, because the US uses foreign direct restrictions on other countries that are using US technology to then export to China. And so the incentives for them long-term are to design US technology out of their technology. We saw that with export controls around space, European manufacturers started advertising that they had no US tech in it so they could do deals with China. And so the unintended outcomes are, as Herb stressed, in the target country, but also in third countries as well.

DUFFY: One of the things I liked with the report’s focus on frontier bias was how I think clearly, it plays into the conversation right now around export controls as well, that when the focus on export controls is really based in that frontier bias, you can really lose the forest for the trees. In terms of what the broader effects might be, I think DeepSeek was a really great example of how innovation is going to happen no matter what, even if China was still using chips that the export controls would have addressed. But that question of how we work with policymakers to really think through export controls more broadly so that frontier bias isn’t a kind of prevailing dynamic in those decisions. And Amy, I think your point is such an important one, that if frontier bias is informing export controls because you are very strategically trying to buy time for one particular reason or on one particular use case, I can understand it as a strategic tactic. But you have to be really clear on that strategy and really clear on exactly what ground you’re gaining. And I would not say that that’s been the prevailing dynamic to date, I think in a lot of the conversations I’ve heard around why export controls were occurring.

LIN: Right, buying time often becomes an end in itself, and what you do with the time is a secondary consideration.

GILES: And I guess one of the things you’re trying to buy time for is to come up with even better ideas for the future. And to come up with those, you need money and you need talented people. Let’s kind of zoom in now on both of those themes in relation to the American innovation ecosystem. The new administration has made some cuts, some significant cuts in funding for organizations like the National Institutes of Health, National Science Foundation that fund basic research. There’s a view around that say, “Well, look at the private sector. It’s super dynamic, it’s investing all these things.” Herb cited SpaceX earlier. “Great, we have this dynamic private sector. Maybe we don’t need to do as much in the basic research area because the private sector sector’s taken this on.” Amy, is that right? What’s your take on this?

ZEGART: I think, Martin, one of the main thrusts of the cross-cutting themes of the report is that all research and development is not created equal, and all sources of funding aren’t equal either. So if we look at what is the innovation ecosystem that has led the United States since 1945 to be the world’s greatest innovation power, it’s a specific model. And that model involves first, the federal government is the world’s best patient, long-term investor that can make big bets on technologies that are foundational. And so the federal government for decades has funded fundamental research in universities, research that has no foreseeable commercial application. Research that looks at, what are the laws of physics or how does the human immune system work? And then universities publish that research openly, which enables the private sector to commercialize and capitalize on all of that fundamental knowledge to develop new applications. Our colleague, Mark Horowitz, who was on the interconnect earlier, who’s an electrical engineering professor and chairman of the department at Stanford likes to say, “Nobody remembers that the federal government funded digital library research in universities for years, but everybody knows and uses Google. And that was the result of that patient investment in fundamental research. So I think we hollow out the fundamental research part of that innovation model at our peril. We are harvesting today, the seeds of fundamental research funded by the federal government that were planted 25, 30, 40, 50 years ago. And if we don’t plant those seeds today, we won’t be able to harvest them tomorrow. Venture capital is wonderful, venture capital does not substitute for federal funding of fundamental research.

GILES: Adam, you mentioned earlier that China has reverted more to a kind of top-down approach. I mean, it seems to be taking the American model and running with it right the way down the field. It’s investing, I think six times faster and in basic R&D. We still outspend in dollars, but boy, are they racing down the field over there. What’s driving that, do you think?

SEGAL: Yeah, I mean, I think that’s exactly right. I think the Chinese looked at our system and the successes that it generated and said, “We want that.” So for most of the ’80s and ’90s and early ’00s, most Chinese government funding went to applied and development R&D, and very, very little on basic R&D. Chinese universities got very little support for that. And then we saw in the 2000s and 2010s, China set up its National Science Foundation and began funding big science, very ambitious projects on the things that, as Amy mentioned, that at the short term, doesn’t seem to have a lot of commercial applications, but long-term, could set the grounds for big breakthroughs in material and other sciences. So the Chinese have been increasing the amount that they spend on basic science. And the other thing I think, Martin, you alluded to in your first comment is on talent. And the Chinese are basically saying to an entire generation of Chinese scientists who’ve gone to the English-speaking world and saying, “Facing budget restraints? Come back, we have a beautiful new lab. We have lots of eager graduate students and you can do all the work that you want to do at home and have your parents and your spouse’s parents take care of your kids as well, so it’s really the best of all worlds for you.”

DUFFY: And I think that question of talent, that for so long the United States has been the place where everyone wants to come to build, right? And our universities have been among the most sought after in the world. The US for the most part, hasn’t had to fight hard to bring talent in. Talented people have fought to be able to come to the United States. I think we are seeing a shift. I think we’re going to see much greater strides by India, you’re certainly seeing a more aggressive stance by China. But understanding how important it is for the United States to stay hospitable to the world’s talent, to bring in that range of not just expertise and curiosity, but also the range of problems that people coming from so many different environments want to solve, that’s a huge strategic advantage. And I worry that we’ve taken it for granted and that we can no longer take it for granted, that we have to start actively fighting for it a little bit more.

ZEGART: I think DeepSeek is a real warning shot with respect to talent. I have a research assistant who did this fantastic analysis of all 211 authors of the recently released DeepSeek paper. She looked, actually, at every paper that DeepSeek published, and she looked at what was publicly available about the backgrounds of those researchers in all of those papers. And what she found was a homegrown talent story. 50 percent of the authors on that recent breakthrough paper, 50 percent of them were educated and trained nowhere outside of China. So to your point, Kat and Adam, people around the world, including people in China, have lots of options for world-class education, world-class private sector opportunities. They don’t have to come to the United States to be trained in the way they might have had to do in the 1960s. So we have to compete for the world’s talent in a way that we have not been used to doing.

GILES: And how do we do that concretely? Because often, skilled immigration gets sort of wrapped up with all kinds of other issues around immigration and doesn’t go anywhere. How can we break out of that cycle?

DUFFY: I would say one strategy is returning to an earlier part of the conversation, is we really do need to double down on making sure that our universities are adequately funded and that our public research opportunities are adequately funded. Because schools, universities, studies, those are incredible areas in which people can enter the country, enter the culture, enter research, and then go on into other opportunities domestically if they want.

LIN: It was striking in a recent visit to Washington, to talk about the report. And when we bring up the talent issue, maybe it was just a function of who was willing to meet with us, but we didn’t find a single voice of opposition on either side of the aisle to the idea that we would want to keep in the United States, foreign science and technology PhDs. Everyone thought that it was a dumb idea to push them out of the country after they completed their degrees. Now, why that doesn’t translate it to some bipartisan action to do that, I don’t know. But that it was really striking.

ZEGART: I would just add that that’s only half the solution. The other half is we have to grow our own. And so K-12 education isn’t a domestic issue alone, it’s a national security issue, and our K-12 education, particularly in math, is bad and getting worse. The United States ranks 34th in the world in international tests and math scores, 34th. We’re trying to compete with Slovenia and Vietnam. Massachusetts is the highest scoring state in the United States in terms of math scores. If Massachusetts were a country, it would rank 16th. We are not educating our own talent to compete for the future. So making the United States more of a welcoming place and competing and funding universities to attract that talent is incredibly important. But we have got to do a much better job at K-12 education in this country.

GILES: Doing my own math, I think that’s about all the time we have left for this wonderful conversation. I would love it to continue, it was so great. Thank you all very much for being here. Adam, Kat, Amy, Herb, terrific teamwork.

LIN: Thank you.

ZEGART: Thanks for having us, Martin.

DUFFY: Thanks, Martin.

SEGAL: It was delightful.