Global power, rivalry, and A.I.; a preview of the U.S./China summit

This is Fresh Air. I'm Terry Gross. Some combustible topics are likely to be addressed at tomorrow's summit between President Donald Trump and China's President Xi Jinping, like controls on AI safety, the war in Iran, tariffs, the trade war, and China's export of rare earth metals and rare earth magnets. A subject sure to be on the minds of the Chinese and the Americans is the possibility of China invading Taiwan, and whether the U.S. would arm and defend the island nation, possibly getting us into an armed conflict with China. Whether or not anything is resolved at the summit, what's said, how it's interpreted, and what's left unsaid will matter. Beijing is trying to take over the position the U.S. has long held as the global leader, and they may be getting closer to achieving that goal. To help us understand all this and set the stage for the summit, I'm Terry Gross. Our guest is Rush Doshi, a senior fellow at the Council on Foreign Relations and the director of its China Strategy Initiative. He's the author of the book The Long Game, China's Grand Strategy to Displace American Order. He's also an assistant professor in the School of Foreign Service at Georgetown University. During the Biden administration, Doshi was the deputy senior director for China and Taiwan affairs at the National Security Council. In that capacity, he was the deputy senior director for China and Taiwan affairs at the National Security Council. In that capacity, he negotiated with his Chinese counterpart and helped arrange summits with China. We recorded our interview yesterday. Rush Doshi, welcome back to Fresh Air. Thanks. It's great to be with you. So Trump likes to say that President Xi likes him, that they get along. How do you think Xi would describe Trump? Well, what's interesting is these two leaders have had dealings going back all the way to 2016, 2017, really, is when that starts. And I think President Xi has always seen President Doshi as a leader. He's always been a leader. He's always been as unpredictable and also solicitous at times. President Trump is quite prone to flattering President Xi, just as he's prone to flattering Vladimir Putin. He has a tendency to flatter autocrats. So I think President Xi sees him as potentially dangerous in his unpredictability, but also manageable if President Xi is able to kind of appeal to his ego. And what we see between the two leaders now is a changed dynamic. You know, in the first term, President Trump felt very confident in pushing China. But at the same time, he's been very confident in pushing China. And he's been very confident in pushing China. He always believed in stability in the relationship and wouldn't always push too far. This time around, I think President Xi feels like he has the upper hand. And President Trump feels like he has maybe a weaker hand than President Xi. Why? Well, the big reason why is because of what happened last year. You know, President Trump took tariffs on China up above 140 percent, unprecedented, maybe even reckless. And in doing so, he actually pushed President Xi into a corner. And President Xi had to make a decision. Does he reach for the break glass tool that China has spent decades building? And that tool, of course, is its control over the rare earth minerals and magnets that go into every manufacturing supply chain in America. When President Xi reached for that tool and said, we make more than 90 percent of the world's supply and you're not going to get any of it, that was a gun to the head of U.S. manufacturing. Now, President Trump could have seen that coming. Certainly past presidents knew that that threat could always be deployed. But they didn't push China into a position of having to use the tool. And here's the thing. President Xi said, we're not going to get any of it. If President Xi thinks he has the advantage with that tool and President Trump thinks that President Xi has the advantage with that tool, then no matter what the reality is, perception becomes the reality. And that's why right now I think Beijing feels like it's tamed President Trump. Did President Trump fold on anything related to that? Oh, very much. So right after President Xi used that tool, President Trump took those tariffs way back down to something closer to 30 percent. And after that, essentially did not pursue a number of actions. That were in the American economic and technology interest, took a slightly softer line on issues related to Taiwan and human rights. Basically, we saw a complete shift in American policy. And Terry, this is a big deal. I mean, this is a moment, I think, where we can feel the tectonic shifts beneath us. What we saw was China received the full force of American economic coercion and for the first time could completely turn that force back against the U.S. and get the U.S. out of the way. And that's why I think it's so important that we think about the U.S. and the U.S. as a whole. as a whole. I think it's so important that we think about the U.S. as a whole. And that's why I think it's so important to go after those smaller banks until you feel real pain. You know, China is the manufacturer of the world. It's the world's factory. So it's got great leverage when it comes to being able to make a thing that you need. But the U.S. is still in many ways the world's banker. We have great leverage when it comes to finance and currency and transactions in the financial sector. And so that was one area of leverage. The other is that the U.S. remains a technology superpower. And so a lot of the technology that China relies on to build chips, to design chips, to run its companies is a American software. And there are other choke points that the U.S. can play with in that front. So the U.S. did have these tools, but reaching for them would have been really escalatory. And so in the end, the president opted to back down because the risk of escalation was too great. What would you like to see the outcome of the summit be? And what do you think it will be? What I'd like to see is probably not what we're going to see. What I'd like to see is an effort to resolve some of the economic and tech issues between us, cooperation on AI safety, some de-escalation on Taiwan, and of course, some kind of path to ending some of the global conflicts that implicate both the U.S. and China, whether that's Russia-Ukraine or, of course, the current Iran war. But I think if I'm being realistic, this is a summit that's much more about pageantry and much more about the appearance of stability rather than actual substance. We're not going to see major breakthroughs or deliverables. At most, we'll see China make some purchases of agriculture. We'll see China make some purchases of agriculture. We'll see China make some purchases of agricultural products. There'll be a dialogue announced on AI. There'll be some photo ops. The first ladies will meet. And that'll be that. Why are you so skeptical about the summit? Well, I'm skeptical for a lot of different reasons. I think the first one is that there's really a difference of interest between the U.S. and China on many of these issues. And the second, there's a lack of trust. I mean, President Trump, from President Xi's perspective, isn't the most reliable interlocutor. He makes promises, then he breaks promises. And I think the risk that he's going to break promises also disincentivizes China from making its own promises. But the final point here is that the U.S. and China are just in a fierce rivalry. It's true that there's been some de-escalation ever since the rare earth situation last year. But the reality remains that the U.S. and China will probably be fierce competitors for some time to come. So AI and cybersecurity are issues that are likely to come up at the summit. You just co-authored a report on cybersecurity. For the Council on Foreign Relations. And I want to read a paragraph from that report, a paragraph that is really quite alarming. You write, China has built a sophisticated hacking apparatus to gain leverage in peacetime and wreak havoc during conflict. Currently, the U.S. lacks the defenses or the credible deterrent to combat that threat. Over the past decade, China has built a sophisticated hacking apparatus to penetrate the water systems and energy grids, telecommunication networks, and transportation systems that underpin American society. The goal is not simply espionage. China is pre-positioning malware and maintaining dormant access to critical systems so it can gain leverage in peacetime and wreak havoc during a conflict. That sounds pretty bad. What has China already done in terms of hacking into American and into government agencies? Well, maybe I can start off by offering some personal perspective here. When I entered government in the Biden administration at the National Security Council in 2021, you know, I was first briefed about the extent of these penetrations into what we call critical infrastructure in the U.S., infrastructure upon which tens of millions of people rely. And it was alarming. And you saw over the course of the subsequent years, many U.S. officials say what the paragraph effectively that you just read, that China has pre-published a program to hack into the U.S. infrastructure. And I think that's a very pre-positioned for destructive cyber attack on the power, water, gas, telecom, and transportation infrastructure of this country. And what they would want to do is potentially shut down some of those systems, potentially in wartime, but could even be in peacetime to gain leverage. So the U.S. has to adjust its policy. And if you think about it, what China has done is put tens, if not hundreds of thousands of engineers on this problem set to build these offensive capabilities. And the U.S. has sort of dropped the ball. We don't have regular regulation of our cybersecurity the way we should because companies often oppose it. It's an extra cost. And the result is we have enormous vulnerabilities that nation states or even private actors can take advantage of to the detriment of all of us. Now that we know that AI can go rogue and team up with a gang of AI-equipped computers and hack into another company, we have a hint of how powerful a weapon AI can be. What is China and what is the U.S.? Each worried about it because I think they're worrying about different things. It's true. China tends to worry a little bit more about questions of political security, where the Chinese Communist Party wants to stay in power and control over information is an important part of that equation. So historically, when they regulated technology, they thought about it in terms of information security. Can I make sure the content that's circulating on the Internet is consistent with the party's views? It's the right propaganda. It's appropriately censored. It doesn't make fun of President Xi Jinping, et cetera. When AI became a major factor in the technology landscape, China began to think about it in similar terms. They wanted to control the content that AI produced, and that's where they began with political stability. In the U.S., so much of the conversation has not been about questions of speech, although that's there, the deep fakes, et cetera. The questions have really been about will there be existential risk or other kinds of catastrophic risk that can come out of the loss of control of this technology, whether that's AI agents that hack. Critical infrastructure, the same infrastructure that China is actually hacking today, but perhaps far more competently, or whether AI combined with things like biosynthesis can create artificial viruses that are lethal and transmissible and can be catastrophic for the human race. Those are just a handful of the concerns people have about AI. But you're right. There's two different conversations. For China, initially, it's been about political security, maybe social displacement from job loss. In the U.S., those things are there, but there's this bigger existential layer. Right. One thing they have in common, China and the U.S., is they both want to win the AI race. And Trump wants no restrictions. He has said about AI, we will cherish it, help it, and watch over it. It kind of sounds more like marriage vows than an AI policy. Compare that to China. Yeah, I think to have and to hold your AI models will not be the right regulatory approach, if I'm being honest. But look, the Chinese Communist Party has regulated AI. They've just done it for their social purposes, right? They're worried about AI giving you the answer to Tiananmen Square or coming after an official in ways that will basically weaken the Chinese Communist Party's hold on information. In the U.S., the regulatory questions have been about existential risks. What President Trump is basically saying is, I'm not really worried about those risks. I think those are all made up. I think they're basically designed to sell AI companies. That the risk of anything catastrophic happening is minimal. Let's just go. And so there's a huge disconnect there. And actually, Terry, it creates challenges. Because at this state visit, if the two leaders are going to talk about AI, you know, it's hard for the U.S. side to say, hey, China, we should work together to make AI more safe and secure when China's view of us is that we're regulatory cowboys. So we've established that both the U.S. and China want to win the AI race. How far ahead are we? From China? And what is China trying to do to catch up? You know, American models are maybe six months to a year more advanced than Chinese models. But the question is, why is China so close? China has great talent. They have great power. But what they don't often have is the compute capacity that the U.S. has. They don't have the chips, if you will, that are so critical to building AI technology, to being able to train the models and then being able to run them, what we call inference. So the U.S. has that. China has a huge gap in compute. And in the last administration and in the first Trump administration, there were efforts to extend the American lead by limiting China's access to that compute. But here's the thing. China did find some ways to catch up despite that limitation. And often that's through what we call model distillation. That's sort of, you know, where they query our models to find more efficient pads to training their own models. So we've done the first draft of the homework and they're able to do a lot, you know, without having to start from scratch. That distillation. That distillation has basically kept the U.S. and China closer than people might have otherwise expected, given the gaps in compute. And I think what you're seeing is many are saying, well, if China is able to distill our models, that means they have to have access to our models. And if our AI labs say, well, these models, we're going to delay the rollout, then China will have fewer opportunities to be able to access the models and fewer opportunities to learn from them, which would extend the American lead. Why does it matter so much who, quote, like wins the race or is in the lead? Well, there's a sense, I think, that China has that history proceeds through a series of industrial revolutions. We have to remember China feels like they were a victim of what they call a century of humiliation at the hands of Western powers and Japan. And the origins of the century of humiliation in their mind is falling behind in technology, that the West mastered the steam engine and the railroad and China didn't. And those things made the difference when the West became dominant globally. And China fell behind. So China's vision is every time there's a major industrial revolution, they need to catch up. The first one, again, steam power, Great Britain won that, became a world power. The second one, electrification, maybe the third one, mass manufacturing, the U.S. won those and became a world power. China wants to win the fourth industrial revolution. That's AI, that's biotechnology, that's robotics, that's quantum computing. And if they're able to win it, they don't think it'll just bring them prosperity. They think it'll bring them power. It'll give them. It'll give them the chance to leap ahead of the United States when for so long they've been behind. So that's why AI matters to China. It's not just a fancy tool. It's not just a revenue stream. It is this essential input in the rise and fall of great powers, which from their perspective is the central story in China's history with the West. And for the U.S., you know, AI is a market story. You know, Wall Street is heavily betting on it. It's a technology story. We want to invent and innovate the future. But it's also a market story. But it's also a competitive story. There's a hope that when the U.S. has lost so many industries to China, succeeding in AI, being ahead in AI, can help us bring back some of those jobs, some of that prosperity and perhaps some of that know-how, and also catalyze a new era of growth in the country. One of the things the U.S. can do to slow down China's advancement in AI is to withhold the sale of high-powered chips that you need for AI. Where do we stand on that now? So the beginning of this policy really dates back to the first Trump term, but it picks up in a big way in the Biden administration, where there was a decision not to prevent China from acquiring all chips, but just perhaps less than 1% of chips, the most high-value, most advanced, what we call leading-edge chips that are useful for training AI. If you don't have these chips, you basically cannot create or train an AI model. And so the U.S.— The U.S. restricted a lot of that to China and then updated those restrictions. The Trump administration initially maintained some of that policy, but then relaxed it in a few areas with, you know, NVIDIA's CEO, Jensen Huang, and President Trump initially saying, let's just sell all the chips to China because if we sell them, China will become dependent on them. That was the wrong thing to say from a Chinese perspective. They don't want to be dependent on anybody. They have a history of dependency. They don't like that. And in addition, they've been trying for over a decade. They've been trying for over a decade to avoid that dependency entirely. Even before Trump was president, they had a plan to indigenize semiconductor production at the leading edge. It's really hard to do. It's probably harder to build these chips than it is to go to the moon. And yet China will keep trying. So those controls are there, and they give the U.S. an enormous advantage in being able to, you know, win the race in AI. But there are many, many people in the U.S. and in some in China who are saying it's time to end those controls. And so far, those voices have not won the debate. So we're in a position. We're in a position where, you know, the U.S. could withhold chips from China, which would slow down China's AI growth. China has the rare earth minerals, which can slow down America's tech growth in general. And so it's kind of like one of those movie shots where two people have a gun pointed at each other's head. So where are we with that? That's right. There's actually a Chinese term for it right now, which is strategic stalemate. So China says now, and they've really been saying this in think tank. They've been saying this in government commentary for five years and in government commentary and documents for the last two or three years, that this is a strategic stalemate with the U.S., that they have some cards and we have some cards. And that's true. Although as important as rare earths are, you know, in some ways, it's not the most advanced technology. It's just very dirty to refine them. And we decided to give away that industry. We used to be the place that processed rare earths. We invented the refining process for that particular commodity. China holds that industry now. But AI chips is different. That's a really hard thing to do. To be able to make these AI chips, these advanced leading edge chips, that is the very frontier of human technology. It is far beyond the advancement we have in almost any other sector. It is the most complex supply chain in the world. And so it isn't quite a one for one vulnerability on that issue. But there is definitely an effort in China to link the export controls to the chips. That being said, it's important to note that I think China decided to use this break glass tool over tariffs. It was really because President Trump pushed them into a corner on tariffs, 145 percent, that they had no choice but to reach for the tool. They didn't ever use this tool in the same way when President Biden did export controls in the last administration. And the tool you're talking about is withholding rare earth minerals. They withheld a few, one or two, for a little bit. But they didn't do this across the board massive push to stop the flow of those minerals to the U.S. that they did under President Trump. My guest is Rush Doshi, a senior fellow at the Council on Foreign Relations. and the director of its China Strategy Initiative. conversation after a break. And if you'd like to see what our interviews look like, please check out our YouTube channel called This is Fresh Air and consider subscribing. Thanks. I'm Terry Gross and this is Fresh Air. This is Fresh Air. I'm Terry Gross. To help us understand the significance of tomorrow's summit between President Trump and China's President Xi and get a better understanding of the current and precarious state of U.S.-China relations, we're talking with Rush Doshi, a senior fellow at the Council on Foreign Relations and director of its China Strategy Initiative. During the Biden administration, Doshi was the deputy senior director for China and Taiwan affairs on the National Security Council. He helped organize summits between China and the U.S. And he's also an assistant professor in the School of Foreign Service at Georgetown University. So you just co-authored a report on cybersecurity. What kind of treaty do you think might be possible on AI where both countries have some kind of mutual understanding of necessary restraints to prevent AI from going rogue? In the U.S., you have enormous alarm right now about AI models and what they can do and whether or not they can be secured and be made safe. And China is the only other major AI superpower. So it makes sense that these two countries would get together and find a way in their mutual self-interest to restrict elements of this technology so there aren't catastrophes. There are a few problems. I think first, China is skeptical that if they slow down, we'll slow down. And we are skeptical that if we slow down, that China will slow down. And so there's not yet a break on this race that we can put in place. The second challenge is that I think China is going to be the only country that's going to be able to China believes that the U.S. isn't really all that serious about regulation anyway. There's a sense in which the U.S. and President Trump doesn't take any of these risks seriously. And even if some of the researchers at these AI labs do, if the president doesn't, why should China trust us that we're going to take regulation seriously? Why should they do it? But there's a third part as well. I think China, again, is much more focused on AI-related political risks, not as so much on the existential risks or the loss of control. For example, if we're talking about rogue AI swarms hacking critical infrastructure, that's not the thing that they're the most worried about yet. Now, that could change. But for now, there's a mismatch in what we focus on. And then finally, there's a mismatch in what we're seeing. I think our frontier labs, because they're withholding some of the most advanced models, have a better sense of how dangerous this technology could actually be than China, which isn't at the frontier in the same way. If they're drafting behind our leading companies and our leading companies stop basically sharing the most critical information, then China is going to be the only country that's going to have a cutting edge technology with them, then China has less of a sense of what risk is coming. And those four things make it much harder for us to make progress. Now, I had some experience with the AI dialogue question. When we were planning for summits between President Biden and President Xi, part of my responsibility was negotiating the AI deliverables for that summit. That meant, could we set up an AI dialogue? And could we get an agreement that AI shouldn't be connected to nuclear weapons, specifically nuclear command and control? Even those deliverables, which would sound like low-hanging fruit, took a long time to achieve. And when we had that dialogue with the Chinese, they didn't really send the right people. They didn't take it seriously as a technical discussion. As for AI and nuclear weapons, eventually, the Chinese did agree that we shouldn't connect AI to nuclear weapons. But that's kind of low-hanging fruit. There's got to be a lot more we can agree on besides that. So we've been talking about AI. Taiwan is important in this scenario because they supply essential components to nuclear weapons. They supply essential components to nuclear weapons. And they manufacture essential components for tech, including AI. Tell us about that. That's right. The entire AI build-out in America, something like a trillion dollars in data center construction, really owes itself to the supply chain that is rooted in Taiwan. Taiwan's fabs, as we call them, operated by TSMC, one of the big Taiwan companies that makes chips, they supply something like 90 percent of all the leading AI chips. And so people think about how oil was so important to the U.S. economy. And the Gulf War was fought, perhaps, to maintain a free flow of energy in the rest of the world. And people remember that. But AI is even more concentrated risk than energy. There's more concentration on the supply chain in Taiwan than perhaps any other critical input, aside from rare earths, in the modern economy. So since there's the threat of China invading Taiwan and taking it over, what would it mean for U.S. development of AI and other tech if China does take over Taiwan or try to? China claims Taiwan is really Chinese territory, not an independent island nation. Is all this tech-related stuff one of the reasons why China really wants Taiwan? Well, first of all, I think that China wants Taiwan for a lot of reasons. One is that China believes that Taiwan truly is part of China. But another is that Taiwan represents the unfinished Chinese civil war. You know, you had a civil war between the Nationalist Party in China and the Communist Party in China that ended in 1949, when the nationalists who were defeated by the communists fled to Taiwan. Now, the nationalists ultimately governed Taiwan for a long time until Taiwan democratized in the 1990s. But from the Chinese perspective, when they fled to Taiwan, it split China. And they believe that recovering Taiwan, unifying with Taiwan, is a foremost priority for the great rejuvenation, as they call it. It is a key political objective. And so they have an emotional tie to Taiwan, a historical tie that is probably even more relevant than the technology relationship. In fact, the Chinese have said the chips are not the reason we care about Taiwan. The reason we care is because of history and because it's ours. Now, if there were to be a conflict over Taiwan, the implication for the U.S. and for the global economy is enormous. If you look at the Bloomberg economic estimates, a conflict over Taiwan could cause something like a nuclear war. $10 trillion in global economic damage. That's 10% of global GDP. That's a bigger hit in some cases than the Great Depression, certainly bigger than the financial crisis of 2009. At the same time, even if it's not an invasion, even if it's a blockade or a quarantine of the island by Chinese vessels and aircraft, that too could cost us $5 trillion in global economic activity. And as for the AI buildup, there's almost no scenario where an invasion or a blockade doesn't damage. Those facilities that I just mentioned, which are concentrated in, you know, about three places in Taiwan. And if those facilities go down, the entire AI industry basically is set back. There will be no more leading AI chips that will be used to train the models. And even the chips that are used to run the models won't necessarily be exported from Taiwan. So it'll be the end of the AI moment for some time to come. Both for China and the U.S. For the world. For the world. And that's apart from, in some ways, the macroeconomic damage. So the Taiwan scenario is the thing that kept a lot of us up at night. I think it keeps a lot of businesses up at night and the risks are growing. We sell arms to Taiwan. Trump put a pause on that. It's something China wanted, but China would prefer that we just freeze weapons sales to Taiwan. Right. So where are we on that now? So there's a long history here. The U.S. has sold weapons to Taiwan for decades. And China has objected for decades. In 1982, there was a communique between Reagan and the Chinese, the third communique in the U.S.-China relationship, which basically suggested that we might eventually one day wind down arms sales. But as we indicated privately and publicly and to Congress and to the Chinese, the quantity and quality of the pace of our arms sales to Taiwan was going to be dependent on the threat that Taiwan faced from China. In other words, it was going to be dependent on our sales to what China was doing in the region vis-a-vis Taiwan. And we thought that was a reasonable position. These were primarily weapons, of course, of a defensive nature. We wanted to make sure Taiwan could resist basically being attacked by China. Now, coming to the summit, the Chinese position is we don't want any arms sales to go to Taiwan before the summit or immediately after. That position is a very strongly stated position. And the reason they're doing that, I think, is because President Trump in May, in Beijing, said that we don't want any arms sales to go to Taiwan before the summit. Beijing said, this is a bargaining chip. Now, we have never treated our arms sales to Taiwan as a bargaining chip before. But when the president said this is a bargaining chip, it opened the space up for more forceful negotiation from China, which is exactly what's happening. The Trump administration did sell weapons to Taiwan at the beginning of the year, but they've paused the most recent package because China has indicated that if the sales go through, the summit may not happen. And President Trump wants the summit to happen. Meanwhile, there is a law in the U.S. saying that the U.S. has to supply Taiwan with defensive weapons to maintain sufficient self-defense capability. So does that law have any teeth? It does. And I'm glad you brought it up. It's called the Taiwan Relations Act. And this piece of legislation is a huge part of American policy towards Taiwan. This legislation was signed largely because Congress was unhappy that the U.S. broke off ties with Taiwan in 1979 recognized China in that same year. And by breaking off ties with Taiwan and ending what used to be a U.S.-Taiwan alliance, Congress thought, well, we have to create some backstop to make sure that we don't leave Taiwan essentially unsupported. Now, at that time, there were a variety of reasons for it. But as time has gone on, the American position on Taiwan has been buttressed by the fact that Taiwan has become a free democracy, a free society, and one of the most important parts of global supply chains. And so as a result, the logic that ties the U.S. to Taiwan has evolved. But the legislation, well, of course, dates back almost 50 years. Well, we need to take a short break here. So let me reintroduce you. My guest is Rush Doshi. He's now a senior fellow at the Council on Foreign Relations and the director of its China Strategy Initiative. We'll be right back after a break. This is Fresh Air. Let's get back to the interview I recorded yesterday with Rush Doshi. During the Biden administration, he was the deputy senior director for China and Taiwan affairs on the National Security Council. He's now a senior fellow at the Council on Foreign Relations and director of its China Strategy Initiative. Let's talk about Iran and the war. I think President Trump would like China to use whatever influence it has on Iran. To get Iran to end the blockade of the Strait of Hormuz and to negotiate a treaty that will end the war. How much influence does China have over Iran? China has a lot of influence over Iran, but Iran is also, you know, under attack and is going to do what it thinks is best for itself. Right now, China's influence comes through a few different channels. The first one is military support. A lot of China's drone parts and missile technology is what makes Iran's missile systems and drones capable. And that's primarily what Iran is using to inflict casualties on the U.S. and surrounding neighbors. So in many ways, Iran relies on Chinese military support. The second is, of course, economic and financial. China is a big purchaser of Iranian oil, which gives Iran revenue. It's also a major funder, lender of Iran, too. And those economic and financial support are a big part of Iran's economic and financial support. And those economic and financial support are a big part of because there's an earthquake in China or a natural disaster or because they want to weaponize that supply chain. And the world can't really tolerate those dependencies indefinitely, so now is the time for us to change. Let me reintroduce you again. If you're just joining us, my guest is Rush Doshi, and he is a senior fellow at the Council on Foreign Relations and director of its China Strategy Initiative. We'll be right back. This is Fresh Air. This is Fresh Air. He is a senior fellow at the Council on Foreign Relations and director of its China Strategy Initiative. So I want to talk with you about summits and what the arrangements are like and what some of the symbolism and ceremonial aspects are, because you helped arrange summits when you were in the Biden administration. So what is the Chinese attitude about ceremonies and how they want to appear when they're in Washington? This is so important. Whenever Chinese leaders come to the U.S., which is less and less frequent now, there's a risk. There's a risk that we could take steps inadvertently that might embarrass them. There's a risk from their perspective we might be doing something that's deliberately embarrassing. At a past summit, actually the U.S. played the wrong national anthem for China's president. Instead of playing the Chinese national anthem, we played Taiwan's national anthem. Oh, wow. That's a huge faux pas. That's bad. And there have been other things like that that have occurred across time. You know, protesters getting up in the middle of a speech by a Chinese leader and yelling at that leader that for them are big deals. So when I was working summits, I had the privilege of working one in Indonesia between the two leaders and one in California called the Woodside Summit back in 2023. What struck me about that second experience compared to the first was our Chinese counterparts were very, very focused on protocol. Making sure that everything went smoothly and there was no risk of embarrassment to President Xi because if he were to be embarrassed, their neck was on the line and they did not want that to happen. What are some of the details that are really significant that might seem insignificant to those of us who aren't involved in the summit? For example, I know China has been upset that it didn't get a lot of advance notice about how many people are invited to the state dinner. That doesn't seem so important to me, but apparently it's really important. Yeah, it's almost like planning a wedding. I mean, there's all these elements that one side thinks is more important than the other side that the two sides might negotiate over. And so there's enormous room for miscommunication. We have entire teams of protocol professionals who work all this out well in advance. There's a whole protocol track when doing summit preparation separate from the policy track. But the two often intersect. So, you know, it could be everything from what the flower arrangements are to who's seated at what table to who greets who at the airport to who gets to speak first at a press conference. If there is a press conference, everything can become a question of face, of status, of respect. And so the risks are really enormous if you're not focused on the small details. You've got to sweat the details. Just one more question. Do you find it strange that President Xi is coming to Washington to meet with, with President Trump and have this big ceremonial dinner? All that ceremony is on Thursday. Meanwhile, he's not going to the U.N. meeting. It is a little unusual. And I don't know what the explanation for that is. President Xi has not gone to the U.N. General Assembly for many, many years now. He often sends others, his foreign minister or the premier. But in this case, he has the ability to go if he wants to. My guess, although I have no evidence for this, is that there's no value for him in staying in the U.S. more than he needs to. The more time he's here, the more risk there is of something embarrassing happening or the U.S. potentially sliding him in ways that he'd rather avoid. So you have to understand the Chinese are always nervous coming to the United States. That's always been true. It's especially true in this more tense moment in our relationship. And it's especially, especially true because President Trump is the president. Rush Doshi, thank you so much for, for talking with us today. It's been really informative and helpful. My pleasure. Thanks for having me. Rush Doshi is a senior fellow at the Council on Foreign Relations, the director of its China Strategy Initiative, and an assistant professor at the School of Foreign Service at Georgetown University. Tomorrow on Fresh Air, we talk about Sean Combs, the music mogul known as Diddy, who's now in federal prison. Our guest will be Rolling Stone investigative reporter Cheyenne Roundtree. For her new book, Bad Boy for Life, she interviewed more than 100 people to ask, how did Combs remain powerful for so long, despite decades of warning signs and allegations? I hope you'll join us. To keep up with what's on the show and get highlights of our interviews, follow us on Instagram at NPR Fresh Air. Fresh Air's executive producer is Sam Brigger. Our technical director and engineer is Audrey Bentham. Our engineer today is Adam Staniszewski. Our interviews and reviews are, produced and edited by Phyllis Myers, Anne Reboldonato, Lauren Krenzel, Teresa Madden, Monique Nazareth, Thea Chaloner, Anna Bauman, Susan Yakundi, and Nico Gonzalez-Whistler. Our digital media producer is Molly C.V. Nesper. Roberta Shurock directs the show. Our co-host is Tanya Mosley. I'm Terry Gross. ♪ ♪ © transcript Emily Beynon