John and Misha are joined by Eric Schmidt, former Chairman and CEO of Google/Alphabet and chairman of the National Security Commission on Artificial Intelligence. Eric discusses the ways in which the US can win the tech competition with China, why AI is the crucial technology of the future, the proper role of government policy, the divide between Washington and Silicon Valley, where the US fell short in 5G, building talent at home, and how to work with techno-democracies.
Read the transcript of this conversation below:
John Yoo: Welcome, everybody, to another episode of The Pacific Century, the Hoover Institution's podcast on China, America, and the struggle for the 21st century. I'm John Yoo, visiting fellow at the Hoover Institution and a professor at UC Berkeley School of Law; and my partner in crime, Misha Auslin, a fellow at the Hoover Institution, well known expert on all things Asian history and politics.
John Yoo: Misha, say hello, and introduce our guest.
Misha Auslin: Hello John. Well, if this podcast is about the struggle for the 21st Century, there's really no better symbol of that than the battle over artificial intelligence, 5G, and the leading commanding heights of technology. And so, we're thrilled today to welcome someone who really needs no introduction, who is at the forefront of all of that, and that is our guest for today, Eric Schmidt. For those of you who may not have heard of Eric, he is an accomplished technologist, entrepreneur, and philanthropist. He joined Google in 2001, and he helped grow the company from a Silicon Valley startup to a global leader in technology alongside its founders, Sergey Brin and Larry Page.
Misha Auslin: Eric was Google's Chief Executive Officer and its chairman from 2001 to 2011. He was also the executive chairman and technical advisor. In 2017, he cofounded Schmidt Futures, a philanthropic initiative that bets early on exceptional people who are making the world better. He has his own podcast, Reimagine with Eric Schmidt. He converses with leaders on exploring how society can build a brighter future after the global coronavirus pandemic, and very importantly for our purposes, he's been deeply involved with the United States Government, with the Obama, Trump, and now Biden administrations. He served as the chairman of the Defense Intelligence Board and was also the chairman of the recently concluded National Security Commission on Artificial Intelligence and concluded by releasing their report, though I know that they'll be continuing to do work.
Misha Auslin: So we are thrilled that Eric is taking time to join us. Eric, welcome to the Pacific Century.
Eric Schmidt: Thank you for having me on your show.
Misha Auslin: Well, we're thrilled to have you to talk about AI today, the challenge from China, the broader question of American tech competitiveness. But I think it's important maybe to start off with the commission that you chaired that recently just a week ago or so released its report. Now this was the National Security Commission on Artificial Intelligence. For those of our listeners who are well aware of China and Asia issues but may not be looking quite as much at the tech side, can you just briefly tell us what the commission was, who was on it, what you did, and then we want to talk about some of the conclusions that you came up with.
Eric Schmidt: The commission was set up a bit more than two years ago by Congress as a bipartisan commission to provide recommendations to the federal government on national security issues involving AI. We were given a pretty broad set of questions to ask. We were given plenty of money and a staff, surprisingly, that produced a report that is now available to the public in nscai.gov. I encourage people to read it, or at least read the first parts of it, and find what you find most interesting.
Eric Schmidt: We concluded a number of things, but the most important thing we concluded was that America is not yet ready to win the upcoming AI competition, that we are not prepared for the challenges that lay ahead, that China is organized to win, is following us closely, is likely to dominate certain areas, and there are areas where we have a chance of maintaining our dominance and competitive strengths, but only if we act now and in a fairly resolute way.
Misha Auslin: So, I was going to ask you exactly about that quote, which is out in the press when the report was released, which you state, America is not prepared, remains unprepared for the coming era. Before we can get to the solutions, I think part of the question is why, the fundamental questions of why the commission thinks, after all these years, after everything that we poured into these technologies early on, why is it that we're unprepared? Is it a cultural problem? Is it a business problem? Are we short-sighted? Do we not take it seriously enough? Because otherwise, the recommendations aren't going to matter if we don't get to the fundamental reason of why this country, after how much time and how much money, remains in your view unprepared.
Eric Schmidt: I think the new news is that we have a new focused competitor in the form of China and her global aspirations. During the time of the report, China announced its intent to master and eventually dominate artificial intelligence, semiconductors, energy, transportation, quantum, synthetic biology, and software platforms. By the way, that's my whole world. That's all of the engine in front of America for trillions of dollars of new businesses, new global markets, and new platforms.
Eric Schmidt: So China's rise has brought with it a new competitor, an able competitor that is organized under different values, runs things differently, and is executing well against its goals.
Misha Auslin: The report, and I agree, we'll probably put up a link to it when we post our podcast, because people should go and read it. It is not a short report. It's 700 pages. There are 300 pages of recommendations in the report. But do the recommendations in your view begin somewhere? Meaning, is there a fundamental starting point for America to take this seriously and to reorient itself to rise to the challenge? Is it talent development in basic and advanced math? Is it intellectual property protection? What is it where we have to start at or otherwise we're not going to be able to get down the road?
Eric Schmidt: We need a plan to win, or at least not lose, and we need ownership in the government. The American system, which is phenomenal, is relatively decentralized. There's an awful lot of people doing many different things that they find interesting. It's typically not coherently planned. The Chinese model is centrally planned. The goals are set out. We call for a technology competitiveness council that would be headquartered under the vice president to serve as a steering commission to assemble the activities. There're enormous activities in the federal government and in the research institutions that are consistent with the direction that we're talking about, but they're not coordinated and they're not leveraging each other.
Eric Schmidt: There are huge issues around talent. Much of the talent is misapplied, and basic talent is not being used in the right way. There are issues around funding, in particular research funding and funding for transformational research and especially on the semiconductor problem. There are partner problems. We won't succeed without having partnerships with like instituitions and democracies, the so-called techno-democracies who share our values. And by the way, this must include Japan, South Korea, perhaps India in order to do this.
Eric Schmidt: So if you look at it systemically, you need a top down leadership focus, you need a plan, you need the people, you need some money and platforms, and you need some serious partners. And we also emphasize that we want to win using America's values. It's easy to imagine that this new technology could be misused in ways that are not consistent with American values of liberty, freedom, nondiscrimination, privacy, et cetera. Those all have to be maintained in these tools and outcomes.
Misha Auslin: So you've been talking about this race, whether it's AI, specifically the broader race in advanced computing, for a long time, and we've actually been able to talk about it. You've been at Hoover a number of times. You've been quite involved with us and with George Shultz before he passed on talking about these issues. One thing I always wanted to ask you because you would talk about it, and everyone would get very worried, and then we'd all try to come up with how would we begin to approach it, very much like what the commission did, but I always wanted to ask you if there was, in your mind, a red line that would make clear to you that we had fallen behind?
Misha Auslin: Fallen behind obviously, it's a way of assessing comparatively across a whole range of different technologies and productive capacities and the like. But is there something in your mind that says, okay, it's now fundamentally different? We are no longer the leader. We are now in a position of racing to catch up. What is it, and what should we be looking for?
Eric Schmidt: We know what that was for China, and we know it was the day, and I know because I was there, that the Chinese Go Champion was beaten by AlphaGo. And we know among other things because the whole match was televised throughout the country of a billion people. Once he started losing, they shut down the television show,but we know that that spurred them to act. For us, I would suggest our 5G situation is the first red line. To review, China has 720,000 5G stations, adding 600,000 more towers. America has about 50,000 such towers. The telcos in China have unfettered access to a total of 600 megahertz for very high speed 5G connectivity in the right spectrum. We just auctioned about 100 megahertz of that spectrum for 90 billion dollars and further indebted our own telcos, which makes it even harder for them to compete with China.
Eric Schmidt: I estimate that we are a factor 10 behind China, and it's unlikely that we can catch up even with changes in our strategy because of the delays in our action, which are complicated. Why does this matter? Well, 5G matters, and you'd like your phone to work better and so forth. But 5G is a core component of autonomy, which is a core component of military advantage. 5G will be a core component of new applications. If those applications are first invented in China and become popular globally, then they will cut off markets for our high tech startups here in America that hope to be the next Facebook and Apple and Google and so forth.
Eric Schmidt: So there are consequences is to losing. Now, none of us five years ago thought that this was remotely possible. I'll give you another example of TikTok. TikTok has taken the world by storm and American teenagers by storm, such a great concern that President Trump tried to essentially halt their success in the United States. That work does not seem to have halted and in fact, they seem to be doing well, because the strategy of containment wasn't the right one.
Eric Schmidt: For our geopolitical relations with China, we have to understand something new. The fight is no longer over aircraft carriers and nuclear bombs, important fights from years ago, it's over global platform dominance. The country that controls the global platform is the one that will get all the returns, both in national security and economics. I'll give you an example. This morning, I was doing a review of a quantum business which I'm associated with. In the last few months, China has become the undisputed leader in quantum communications, and they have begun a large new effort around quantum computing, which is thought to be, this the rumor, that their goal is to break all of the encryption that we're currently using for our secure communications. That's not a good thing.
Eric Schmidt: So my point is that each of these battles is underway now, and we need a strategy to win. I call this the global competitiveness strategy. And I want the West to win these. I want it to win with our Pacific partners, and I want our country to get organized around focusing on this.
John Yoo: Eric, I'm coming to you from the Berkeley Law School.
Eric Schmidt: Yes, hi.
John Yoo: The rumor is that you lived in international house as a student right across the street as a PhD student, and that, in fact, you met your wife there. Is that all true? Confirm that for all of us at the campus before we turn to serious questions.
Eric Schmidt: That's completely true, and I remember the law school very, very well.
John Yoo: Just because we block your view of the bay-
Eric Schmidt: That's right.
John Yoo: ... I guess from your college dorm.
Eric Schmidt: Berkeley was and remains a fantastic institution, and I remain associated with Berkeley along with Stanford. I cannot say enough good things about the impact that you both have.
John Yoo: I was going to ask you because one of the things you mentioned was, of course, research and development. Suppose you were... I guess this is one of those things you haven't yet done, but there's still time for you. If you ever run a university, suppose you were the president of a university, and you wanted to move forward in the ways that you've been describing on AI, these other platforms, what would you do? What kind of programs would you build? How would you try to direct students and faculty research in these important ways?
Eric Schmidt: I think universities have met this challenge. If you look at Berkeley, for example, because I know the numbers, there has been an explosion in computer science and AI to the point where they're producing thousands of graduates of both liberal arts as well as out of their engineering school. Across the United States, it's also true at Stanford, the number one major is computer science, and the most popular submajor within computer science is machine learning and AI. So there's good news, and the good news is that the young men and women who are in our universities today are going to come out with world beating knowledge and world beating ideas if we can get the rest of the world organized to take advantage of that.
Eric Schmidt: Universities also have done a fantastic job of moving the research in the areas that are important, in particular in the area of ethics, scale, and so forth. But universities need some things. If I were a university president, aside from dealing with all the usual problems that a campus president has now, it's a hard job.
John Yoo: You wouldn't have to worry about fundraising, for example.
Eric Schmidt: Yes, it's a tough role. I would try to do a number of things. I would try to get the university professors as a group to operate with the government to create a national research AI facility. Much of this AI research, and we talk about this in our report at great length, requires very large super computing calculations, much more than is the normal amount of research that a university has. Our report calls for a natural research network of open source data and training material that would allow for research to proceed more quickly at the graduate level and also for those grad students to leave and form startups.
John Yoo: The ones living in international house right now in those little-
Eric Schmidt: Exactly.
John Yoo: ... 500 square foot apartments.
Eric Schmidt: And it was fine at the time. But let me tell you, you get the raw talent that's in our universities. The next thing that I would do is I would remind everybody how dependent we are on foreign smart people. One of the questions we asked was, who is driving this industry forward? The answer is, the stereotype is true, the top universities, the top graduate students, and the top graduate students are usually 20 to 40% international, often with a Chinese researcher graduate student.
Eric Schmidt: So if you were to ban those people from America, which various people have suggested, it would hurt America. Instead, we should encourage the top people to come here, and then we should keep them here by giving them the appropriate visas with the appropriate security checks. It seems to me that the current research enterprise in universities, if you combine with the necessary infrastructure that they need, can move us aggressively forward. The Chinese model, by the way, is pouring money into their graduate programs in the model that they use, which is called Civil Military Fusion. So we know they're doing it. We need to do the same.
Misha Auslin: The second set of questions is, I don't think in America we would be as worried about another country becoming dominant in a platform if it were which you called them similar democracies with similar values, so if it were France or the UK, would we be as worried? I wanted to ask you about your perceptions of China and its intentions. For example, tell us about your first trip to China and your last trip to China. Did you see a change in the intentions of their leadership, in the culture? Has this been a change or has it been going on all along and we just didn't see it? What did you see as a business executive when you were there on the ground visiting?
Eric Schmidt: My first trip to China, I was at Sun, and we were setting up the internet of China.
John Yoo: We should tell everyone that's Sun Microsystems, which doesn't exist anymore, but I remember it.
Eric Schmidt: I had just graduated from Berkeley and so forth. It was a country of bicycles. There were a small amount of connectivity, and the first connection was set up at Tsinghua University. I was done using Sun computers. I remember it because I was talking to the faculty who never got their PhDs in physics and so forth because they were victims of the Cultural Revolution, but they had returned with great courage and had helped to build a great country. But I remember the bicycles more than anything else, and I have the pictures to prove it.
Eric Schmidt: My last visit was with Dr. Kissinger, who is a good friend of mine, and I'm writing a book with him.
John Yoo: Tell us about that later in the show because that'd be really interesting.
Eric Schmidt: It's a book about AI. But with Dr. Kissinger, we went and toured the usual spots in Beijing. The most interesting thing were the private dinners with people who are executives, not unlike the executives that we deal with in the tech industry, very wealthy, very smart, very successful in their hopes and fears. They are proud of what their country has achieved. They're proud of the safety. They're proud of the security. They're proud of the growth. They're proud of the products. They're proud of its global ambition . But they all wanted to send their children to America for education.
Eric Schmidt: I thought that the secret to dealing with China, in my view, is that there are globalists in all of these countries who see the world a better place where we communicate with each other, where we understand how to co-exist, where we compete when it's appropriate and we collaborate when it's not, when it's appropriate. And we obviously don't want any sort of combat.
Eric Schmidt: I was struck by the hopes and fears were much more similar than I expected. China today is a very, very modern country, at least in the cities. If you grow six percent a year for 30 years, which is roughly what the CCP has delivered, the fastest growing country in the world. I have taken the view, which is different from other China scholars, that I'm just going to take China for what they say. I'm just going to believe what they say. Here's what they say. We're going to have a peaceful rise. We're going to limit our ambitions to our local area, which that's code for Taiwan. We are going to build the kingdom on the hill, in terms of our essential kingdom. The conception that the Chinese have is that they were in a kingdom where everyone paid them credit, and then 150 years ago because of various political errors they lost their way, and they want to come back.
Eric Schmidt: I take them at their word that they're organized to do that, that they have the economics and the control and the laws that will allow them to do that. I'm not endorsing it, I'm just saying I take it. The reason I'm saying that is, if you take them at their word, then you have to believe that they are going to be our primary economic and national security competitor for the next 20 or 30 years. The reason is because that's what they said they're doing. We don't have to sort of investigate deeply in the intelligence apparatus. They actually published it.
Misha Auslin: Eric, actually to pick up off of that, this is Misha again, you mentioned a few minutes ago the questions of cooperation and collaboration and then competition on the other side. I think as you exactly point out correctly they've made no secret of the competitive elements of their strategy. Do you find as you talk to them that you actually see there are areas of cooperation that benefit us? Or is it simply cooperation that benefits the Chinese, meaning ultimately it's really zero game? It's a zero-sum game that we lose through what we think is cooperation because the Chinese see cooperation as a way to continue to build the national strength of China, whether it's made in China 2025 indigenous innovation, civil military fusion, whatever it is.
Eric Schmidt: I'm an amateur in foreign policy, but I have opinions now, so I'll just state them.
Misha Auslin: Please.
Eric Schmidt: I spent an awful lot of time with people who make pronouncements, which sound to me like Cold War pronouncements. We love them, we hate them, we block them, we fight them. I think that that is a very unnuanced view. We're not at war with China, we are competing with China. If you were at war with China, you would do different things. But since we're in a competition with China, we want to make sure that we're always acting in our own best interest.
Eric Schmidt: For example, full decoupling of trade with China, literally the goods that flow back and forth between the two countries, would be crazy because all it would do is drive up our prices, and it would hurt them as well. I'll give you a simple example of how this works. Look at Apple. I used to be on the board there, and it's an impressive company. Their products are made in Shenzhen. They have huge government relation issues with respect to iPhones and apps and so forth in China. Don't you think the Chinese government would like to get rid of them? Of course they would, but they can't because they need them as an indigenous partner. They need the economy, they need the training, they need the income, the knowledge transfer, they need the modernization.
Eric Schmidt: Don't you think Apple would prefer to build those products in a country that wasn't hassling them over all of these other issues? Don't you think they'd prefer to have a nice democratic country that could actually build everything at the same quality and cost as China? Of course, they would, but they don't have an alternative. So they're forced to work together, and they're not best friends. But they found a way to make it work. They have a lot of meetings. They have a lot of yelling, but they have a mutual interest in the success of their product platform. That, I think, is the way this will evolve.
Eric Schmidt: In our report, we specifically say that areas of software are going to be very difficult to withhold from China and vice versa. The reason is that software leaks, the ideas are flowing. There's very little advantage of one over another. China, the moment they get something from us, they improve it, and then somebody from us gets their ideas and leaks back. The software industry is very leaky in that sense.
Eric Schmidt: However, our report says that it's really important for us to find a way to maintain two generations of semiconductor leadership ahead of China. Now, the history here is important. In the 1980s, we created a group called SEMATECH. We had a bunch of semiconductor manufacturing in America. Eventually that all moved to East Asia, primarily Singapore, and then South Korea and now Taiwan through TSMC. The most important chips are made in Samsung and TSMC, South Korea, and Taiwan. China has had over 30 years to plan to try to catch up. It's really difficult.
Eric Schmidt: We don't want them to catch up. We want to stay ahead. We call for all sorts of techniques to try to make sure that we rebuild a domestic semiconductor and semiconductor manufacturing facility within the United States. This is important, by the way, for our commercial industry as well as for national security for obvious reasons. By the way, chips, I'm not just referring to CPU chips, there's a whole new generation, I'll give you an example, of sensor chips that sense things. It's really important that those be built in America.
Misha Auslin: You mentioned SEMATECH , which I assume in schools is probably taught as an unsuccessful effort in industrial policy. That's what I want to get into. What's the right line between an industrial policy but also letting the free market work, which has created companies like Google?
Eric Schmidt: I think that what I've learned, I've been doing this now for 15 years with the Obama White House and then with Trump and now with the Biden folks. If you propose an industrial policy, you are dead in the water. Whatever I'm proposing is not industrial policy.
Misha Auslin: Just by definition.
Eric Schmidt: It's not industrial policy. Whatever you think it is, I'm not doing that. The reason I say this is that there's this reflexive view in America that government cannot do this kind of planning. Let's just review, our competitor China has a central policy. It's very well-articulated, by the way. They have huge and rough domestic competition. They pick the winner out of that brutal internal competition. They make that winner a national champion, and because of civil military fusion, you must assume that there is a military component of each of these companies that we don't know about.
Eric Schmidt: That's their model. Now, the American model is laissez-faire. Two people create a company, lots of wealth is created, everyone is happy. This works until when it doesn't. It has not worked in semiconductors. It didn't work in 5G for sure. It may or may not work in synthetic bio. It may or may not work in some of these new AI areas. I think we need to be a little bit more nuanced again, and we need to say that there is a rule for government in a couple of areas. One is publishing a plan. Second is printing money, which is what the government does these days, to help fund the precompetitive work as it's called, the research work, strengthen our American research institutes, et cetera.
Eric Schmidt: Research spending in America is at the lowest percentage it’s been since Sputnik in 1957, which is 0.7 percent. Our incredible economic growth, which I'm very, very pleased to say I'm the beneficiary of and I'm really lucky, was due to things that occurred in the 50s, 60s and 70s when we were awash in money in those areas, because of Sputnik and so forth. We need to repeat that.
Eric Schmidt: So I'm not suggesting that we need to have industrial policy because, as I said, anything I say is not industrial policy. What I am saying is we've got to be a little smarter. For example, let's use 5G, just to hammer on that, the Chinese gave the frequencies to their telcos with a build out requirement. We auctioned off one-sixth of that frequency for 80 billion dollars, further indebting the telcos who don't have any money to build out the same infrastructure. It's madness. Now, that's not industrial policy, that's just bad policy.
Eric Schmidt: To put another way, whatever we're doing now in these areas is producing the wrong outcome. Just to hammer on it a bit more, we talk a lot about the federal government's technical capabilities. The government is full of very smart generalists. There are very few people who can delve into and manage the technology that I'm describing. We call for a federal training program, which includes a four year university with a five year give back working for the federal government, to bring in technically skilled people who can help administer these programs.
Eric Schmidt: One of the problems is that the customer on the federal side doesn't understand the implications of these things in the way that they are. We're going to face enormous issues. In cyber attacks, for example, you're going to have very sophisticated third party actors that will be able to use software to do all sorts of misinformation attacks or other kinds of attacks. I want our government to have people who can protect us. So getting the policy right, which allows for the competition to occur, is important, and getting the right people to understand it is important.
John Yoo: The argument is our classic claim that price signals in the market can't function here because consumers are not fully capturing the benefits of these products. Society is misdirecting resources by relying solely on the market. The counter argument is, what if the Chinese make a mistake? They're so directed in picking technology's champions. We've seen that with other countries. They may end up like France picking Minitel as their work. We can fall into that same problem too.
Eric Schmidt: Exactly, right.
Misha Auslin: So what happens-
Eric Schmidt: Well, in the first place I think that praying that your competitor is going to fail is not your best strategy. That's a strategy when you only have a losing strategy ahead of you. In other words, we don't have any good options, so let's just hope our competitor screws up. Of course, they could screw up.
Misha Auslin: That's the American way.
Eric Schmidt: Yeah. It's okay. Of course, they could screw up. By the way, they're busy building all sorts of powerful missiles. They are ahead of us allegedly in hypersonics. They're ahead of us in drones. If they're going to screw up, they're going to screw up in some things other than the areas they're already leading in. So we've got that point. The way you do this in America is you build a consensus on what winning looks like, and you have to get people to say this doesn't make any sense.
Eric Schmidt: In the competition with China, it doesn't make any sense to take the smartest people in America and send them to China. Americans are intelligent people. Everyone would say that's kind of like a stupid plan. Don't do that. Similarly, for key technologies in the semiconductor area, we should probably limit their access. I think that you can come to a consensus on these. I think that there are always intellectual arguments about platforms and failure and can the government make it right, and there's a presumption of government incompetence.
Eric Schmidt: You have plenty of successful groups, the National Science Foundation has largely been successful in what it's done. It founded under Vannevar Bush that basically created what we have today. The industries that I work in all have largely come through NSF funding and other associated funding. You had DARPA and ARPA, which are government funded but civilian controlled activities. You have plenty of examples of this. We just need more of them. By the way, we also propose a national technology foundation that would attempt to bridge the valley of death between some technologies and their commercialization.
Eric Schmidt: We think that we just need more of everything. You're going to have some failures. You're not going to get great, and I don't want the government to pick winners. I do want the government to pick winning categories, the category of AI, the category of semiconductors, et cetera.
Misha Auslin: I'd actually like to pick up on that. You mentioned you’ve worked with the Obama White House, you've worked with the Trump White House and you're now working with the Biden Administration. But there was an attempt to take a different route with China and in relation particularly to some of these tech questions by the Trump administration compared to administrations before it, whether it was Huawei and ZTE, it was looking at access to chips. What did they get right? What did they get wrong, and what do we need to do to make it right under the Biden Administration?
Eric Schmidt: I'm not a fan of the Trump Administration, so put that in context. I think that their overall instinct was correct, that a change of strategy was needed, and I think the tactics did not work very well. Let's give them credit for saying the world has changed, we need a different way to approach. But, I'll give you an example of TikTok. The president and his team attempted to block TikTok from operating in the United States. They ultimately brokered a deal that would force the cloud hosting of TikTok to be 85% controlled by Oracle. That deal did not go through either, so basically nothing has happened.
Eric Schmidt: After all of the Sturm und Drang and all of the excitement and all of the politics and all of the deadlines, nothing really happened. The correct strategy for TikTok, if you're concerned about the Chinese government taking private information from Americans, is to have TikTok put its data in a US regulated cloud provider of which there's at least three well known ones, Amazon, Microsoft, Google and maybe some others, maybe even the Oracle platform could do it. The important point is that if you can define the problem, there is probably a technology solution that goes right at the issue, so there's a good one.
Eric Schmidt: In 5G, the simple answer is we've got to come up, and, again, President Trump, I spoke about this a number of times with the White House, they understood it, but they couldn't figure out a solution. We need a solution. I proposed one involving sharing frequencies, but you can debate that. We need a solution that is competitive with China within two to three years. At the moment, there is not one on the table. So that's a work item.
Eric Schmidt: I'll give you another example. TSMC Taiwanese company, well run, important partner of the tech industry, is producing five nanometer chips, moving next year to three nanometer chips. Nanometers refer to the spacing between the wires, if you will. And this defines everything in terms of performance and value. The US players have largely not been able to get to that threshold. How do you want to solve that problem? The Trump Administration had a good idea, which was to try to get domestic fabs in the United States. It's not enough because you're not going to get the leading edge stuff, so we need more.
Eric Schmidt: Our report says that we need a national semiconductor strategy, which addresses the commercial needs as well as the national security issues. We think it's going to cost 30 to 40 billion dollars because we don't think that the private sector, and this goes back to John's point, the private sector is not going to raise 30 to 40 billion dollars on a hope and a prayer for a semiconductor plant if it's owned in the US in Arizona, which is where they all seem to be.
Eric Schmidt: You're going to have to have some kind of federal guarantee, some kind of federal purchase cooperative or some mechanism that causes that thing to get built. Those are actions where you have a precise goal, a precise trade goal, a precise competitive goal, which the government can do, and we recommend those.
Misha Auslin: We've been talking about this divide between the US and China, but you've also been intimating as you've talked about what government can and can't do and what it should and shouldn't go about another divide that's just as significant, which is the divide between Silicon Valley and Washington. Different perceptions, you said a few minutes ago you're not a foreign policy specialist, and John and I are not business people and tech specialists, and we have different views.
Misha Auslin: We look at China often from a national security perspective. Silicon Valley looks at it from a business perspective. You straddle both worlds. You're one of the few who does that. You're sort of the Bernard Baruch of our age. You're straddling both of those worlds, which is vital and we need that. I think you were part of one of the Hoover initiatives that our colleague H.R. McMaster and Amy Zegart ran on trying to bridge that divide. How big is that divide, and do you have confidence that the two sides first of all even understand each other and understand each other's concerns? And then we can actually work to repair it because the ultimate benefactor of that is the country. Or do the companies simply see themselves as global and not particularly American?
Eric Schmidt: Let me say it first that I thought that the Hoover work that Amy and H.R. did with you, I think, were helping, was really phenomenal work. It really helped me a lot, so thank you for inviting me for that. It's so easy to do this in sound bites, but it's much more interesting to do this with real knowledge of what's going on in the country. I think that's what your scholarship involves. I think the same needs to be said for the tech versus the government narrative. Because of what happened to Google, there's this presumption that tech doesn't want to work with the government, which is certainly not true.
Eric Schmidt: There are hundreds and hundreds of companies including all of them, including Google, that are trying to work with the government, and they find it difficult because the government has byzantine and bizarre ways in which it works with the private sector. They have a comfortable relationship with the primes, the prime contractors, because they exist to serve the government, so they've all sort of adapted. But a commercial company that has important technology that the government needs finds it very, very difficult to work with the government in a procurement cycle for many reasons. I was the head of the Defense Innovation Board and working for the secretary of defense for five years, and I saw this firsthand.
Eric Schmidt: To give you an example, if you have an idea of a new product, there's something called the Program of Record, and this so-called POM process is a plan for two years from now. If you have an idea, you plan now for two years from now to start the award of the contract to begin building the product. To say that this is slow is an understatement. We're facing an aggressive global competitor. So this is a well understood problem, but the solution is not obvious. People have struggled with this for a long time.
Eric Schmidt: I would tell you that all of the tech companies are now trying to work with the government. All of the tech companies have essentially AI ethics boards, which try to determine if something crosses their own ethics policies, which I support. Each company is different, but they are similar. As part of our military work, we got the government to announce a military AI ethics rule, which I was also very satisfied with. I think the process, which was tumultuous, ended up with players who understand what their goals are, and now we have to get them to work closer together.
Eric Schmidt: There are quite a few proposals within the government as to how to do this, but think of it as there needs to be an innovation group, a technology insertion group, that's much bigger than what they have today or these technologies would be built in the commercial sector and not used in the military.
Misha Auslin: Eric, a final question before we let you go though we'd love to keep talking. You talked a little bit earlier, and I was very happy, you actually anticipated this question a little bit in the beginning. You talked about a tech alliance of democracies. We have to work with countries that share our values. But it's so much... It appears to be really a China-US competition. It's really a China-US heavyweight bout, and other countries are in different weight classes. They're middleweights, some are flyweights, maybe we've got a light heavy in there. Is it really realistic that we can find other countries that can make material contributions to what we need to do in order to maintain our level of competitiveness, and, in fact, as you said to keep ahead in certain areas, let's say two generations ahead? Or let's be honest, are we really alone? Is it really just all about us? If we mess it up, there's no one that's going to come to our aid, and China is going to sweep the board.
Eric Schmidt: I think I have a clear answer to that, which is let's think about how good Korea is in some things. If we had a really close relationship with Korea, as well as Taiwan, we would be in a stronger position. My friend Jared Cohen talks about the techno T-12, the 12 countries that have... They're at a certain size. They have big industries. They have a lot of technology capability, and they have a lot of money, and they have a lot at stake. I think it's obvious that unifying them will make us stronger.
Eric Schmidt: We should be able to compete and win in the majority of the areas. One way to think about this is we're now in a competitive race, and the stakes are huge. We estimated that the AI market was 15 trillion dollars, that the quantum market was 10 trillion dollars, and this is in global revenue. So these are very large industries that we are competing for. What happens is, I'm sorry to bash the traditional dialogue, but the traditional dialogue is all around traditional bombs, traditional tanks, traditional steel, traditional wheat. Those are static objects. We're talking about technologies that learn, that grow, that change their platforms. They provide the basis of whole new industries. We need the United States with its allies to be the winner globally for those.
Misha Auslin: That's great. It's also an optimistic point at which to end because often our discussions, they get in depth, but they really don't seem to lead us to a way out. Again, it's the National Security Commission on Artificial Intelligence. NS-
Eric Schmidt: NSCAI.
Misha Auslin: NSCAI.gov. People should go and read the report. There're sections on the intelligence community. There're sections on how this interfaces with obviously the civilian economy. But it really is about ultimately ensuring the defense of the homeland. There are lots of other questions we would've loved to have gotten to, but this was extraordinary. We really appreciate your time, what you're doing, and obviously we hope that you'll stay engaged not only with us at Hoover but be able to come back to the Pacific Century and talk to us about how you're doing, maybe a report card update in a few years.
Misha Auslin: On behalf of John, Eric Schmidt, thank you so much for joining us on The Pacific Century.
Eric Schmidt: Thank you, John. Thank you, Misha. I'll see you soon.
Speaker 4: This podcast has been a production of the Hoover Institution where we advance ideas that define a free society. For more information about our work and to hear more of our podcasts or see our video content, please visit Hoover.org.
