China, Chips and Chaos with Chris Miller

0:04Hello, I'm Anthony Scaramucci, and this is Open Book, where I talk with some of the brightest minds out there about everything surrounding the written word, from authors and historians to figures in entertainment, neuroscientists, political activists, and of course, Wall Street. Sorry, I can't resist. Before we get into today's episode, if you haven't already, please hit follow or subscribe wherever you get your podcasts and leave us a review. We all love a review, even the bad ones. I want to hear the parts you're enjoying or how we can do better. You know, I can roll with the punches, so let me know.

0:38Anyways, let's get to it.

1:08and explains how it will define our geopolitical future.

1:19So joining me today on Open Book is Professor Chris Miller. He's an associate professor of international history at the Fletcher School at Tufts University and the author of a best-selling book. I think the FT also said probably the best book of the year, Chip War. I'll hold it up here, Chris, the fight for the world's most critical technology. And before we got started, I was telling Professor Miller that I read this in one session, primarily because I saw my entire future unfolding before me, Chris, okay, because there's a lot in here. And whether people like it or not, what you're writing about, they're relying on in their lives.

1:54So much of this technology now is embedded in every part of our lives, whether it's our medicine and healthcare, our telecommunications, the way we interact with each other in business. But let's go to Tufts for a second. You probably know I'm a jumbo. How long have you been teaching at Tufts? How did you get to Tufts? And considering your previous books, how did you get interested in this field of the chip field? Well, I've been teaching at Tufts for six years now, and my background is in Russian history, which, as you say, is not the most natural background for writing about semiconductors. But I started around six or seven years ago, planning to write a book on the evolution of defense technology during the Cold War.

2:36And the question I want to understand was, why was it that during the Cold War, in the early stages, both the US and the Soviet Union could produce the key weapon systems, nuclear weapons, long range missile systems, by the end of the Cold War, it was only the US that could produce the types of precision strike systems that the U.S. dominates today. And I thought that'd be a story just about missile systems or fighter jets, but it turns out it was a story of computing power, which the U.S. applied to all sorts of defense systems. And I quickly realized that it wasn't just the military that had been transformed by computing power.

3:08It was everything, because today it's hard to find a device that doesn't have at least one and often dozens or hundreds of semiconductors inside. So I want to ask one question about the Russian situation first. Why did you get into Russian history? What fascinated you there? You know, it was just sort of by chance. I studied it as an undergrad, was fascinated by how different the society was, wanted to understand what made the country tick. And it is so different from the U.S. It struck me as an interesting place to study and to better understand the comparisons and the contrast between the U.S.

3:46and Russia. So, you know, I've only been to Russia once. I took a semester away from Tufts and went to the London School of Economics. I spent five days in Russia in March of 1985. This is when Chernenko had just died. Gorbachev was rising to power. And again, I don't have a lot of consequential intersections with the Russians. But one thing that struck me was one of the tourists, I was on an in-tourist, you probably remember that agency that took you around and obviously they spied on you the whole time. But what the woman said to me was that they're halfway in Asia and halfway in Europe and therefore in no place, meaning they don't feel culturally attached to either.

4:26And so even though it's the largest land mass in the world controlled by one sovereign, they sort of feel like they're not connected to either place, which has probably wounded their pride maybe or has done something to affect the culture there. What do you think of that observation by an interest tour guide 35 years ago or 37 years ago? And how do you think that's impacted their culture? You know, I think you're right. I think if you look at the Russian war with Ukraine right now, you see that playing out in action. On the one hand, Russia feels itself not part of Europe in conflict with the West.

5:02On the other hand, if you look at Russia's relations with China, the Russians are, they have to be friends with China today because they have no other option. But they're really deeply concerned about in the long run, are they putting themselves in a position of being a second tier power with China, the more dominant country? And I think that that's a dynamic that's been reflected throughout Russian history, this sort of pendulum swinging back and forth between Russia aligning itself with China, aligning itself with the West back and forth, never feeling itself really at home in either part of the world.

5:30Well, correct me if I'm wrong. There's a 1 ,500 mile or so piece of land mass that the Chinese feel was taken from them by the Russian government in the 1630s. And they feel like that that's their rightful ownership. So there is some tension about a large piece of land bordering China. Is that not right? That's right. There are centuries old territorial disagreements. I think, though, I wouldn't overestimate the territorial question. A lot of Russians are actually concerned about, does China want to take their land? The reality is that there's not much of an economy in most of Siberia. Such economy as there is involves drilling oil or digging for minerals that are already shipped to China.

6:15China has already complete access to Russia's mineral wealth. And so there's actually not a whole lot of logic for China to really try to take back territory because it's got free access right now. And one of the aspects of the war between Russia and Ukraine that has been underappreciated in the West is the extent to which, because of the war and because of Western sanctions, China's got privileged access now to Russian minerals because Russia can't sell them to anyone else. So for China, it's been a pretty good financial deal because they get cut price access to some of the world's greatest mineral deposits.

6:45Okay, let's go to the book. This book is fascinating, The Fight for the World's Most Critical Technology. I have young viewers and listeners on this podcast. So let's start with the most basic question, Chris. What is a microchip? So a microchip or a semiconductor is also known as a piece of silicon, often the size of your fingernails, sometimes larger, with thousands or millions or today often billions of tiny circuits carved into them. And these circuits are turned on and off by a device called a transistor. When a circuit's on, it produces a one. When it's off, it produces a zero. And all of the ones and zeros undergirding all computing are these microscopic circuits flipping on and off.

7:29And so those processes, which are basically operating in a binary system, 0101, those processes are able to do tons of different computations, tons of different algorithms, tons of different software commands, okay, that can light up your refrigerator and dial your refrigerator into the right temperature. It can help you access the internet from your phone, your car. It can uplink you to a satellite so you can get a GPS system in the car and take you precisely to the location that you're looking at. It has the ability to allow computers to talk to each other as well. And so there's a multitude of different things that are going on, including our power grid.

8:15In fact, I don't think I can name anything, Professor Miller, that does not involve at this point a semiconductor chip if it's something to do with the interoperability or the interaction between human beings. Am I right about that? Yeah, that is right. Today, basically everything with an on-off switch has at least one and often hundreds of chips inside. And I think most people struggle to realize this because you've never bought a single chip in your life. In most cases, you buy chips embedded in the products that you acquire. You never see the chips inside of them. And so most people don't really realize just how dependent they are on semiconductors and how they rely on not just hundreds, but thousands and thousands of semiconductors to make their daily lives work.

8:55So to really date myself, because I'm closing in on 100, I mean, I'm closer to 100 than I am to zero. Professor Paul Kennedy was all the rage when I was in college. He'd written a book called The Rise and Fall of the Great Powers. He's a Yale professor. It was a book that scared the Reagan administration and talked about America's decline. And we can debate whether or not America is declining or other nations are rising or some combination of the both. But what I found fascinating about your book is that there's only four or five countries. Let me read them. Okay. Tell me if I have them right.

9:28Okay. Right out of your book, United States, the Netherlands, Germany, the UK, Japan, Taiwan, and Korea are basically the major manufacturing centers for these semiconductors. So weirdly, they're technological superpowers, are they not? I mean, they're controlling the keys to everything. If their foundries or factories are eliminated, it would cause a international crisis and probably a global recession. Do I have that right, sir? Yeah, that's absolutely right. And we talk about globalization and the market for selling chips is global. You can buy a smartphone anywhere in the world, but the production of them is not global.

10:05It's hugely concentrated. And so if you look, for example, at the production of advanced processor chips, the type of chips that go inside of smartphones or PCs or data centers, 90 % of the most advanced processor chips can only be produced by one company on one island, and that's Taiwan. And there's just no other part of the global economy that comes anywhere close to that level of concentration. I like to make the comparison of oil. Saudi Arabia produces less than 15 % of the world's oil. Taiwan produces 90 % of the world's most advanced chips. What do you think happened there in terms of U.S.

10:36policy? We're not stupid people. We led to the rise of the Silicon Valley. We put venture capital taxation in place, long and short-term capital gains to create capital flow into innovation. We did so many smart things in our country. But glaringly, over the last three decades, we're relying on the island of Formosa and the Republic of Taiwan, which is in a tenuous situation. So just to remind all of our listeners, Taiwan for greater China, mainland China, is considered part of China. Of course, the Taiwanese government doesn't believe that, and they want to have their own sovereign nation. But the United States is torn and has to accept this sort of two systems, one country sort of approach, if you will.

11:20I mean, there's a better vernacular for it. You probably know it better than me because you're at the Fletcher School. But the point being, Taiwan seems to be under threat from mainland China. Since Chairman Mao, the founder of Communist Chinese Party, or at least the one that took over in 1948, the mainland, they want that island back. And we have 90 % of the semiconductors being produced out of our Island. So where do we get this wrong from a policy perspective? There's nobody in our national security complex at this point would say, okay, yeah, that was a really good move to have 90 % of the semiconductors, the most important semiconductors manufactured there.

11:58Yeah, I think there are two trends that intersected simultaneously. One was that TSMC, the Taiwanese chipmaker that towers above everyone else in the industry, really grew dramatically over the last 15 years. It was a company that put itself at the center of the smartphone industry. And then now it's at the center of producing the chips that power AI. And so it's grown from a medium-sized company 15 years ago to an absolutely central player in the world's tech ecosystem. And it did that by producing very high quality cost-efficient chips. And so the Silicon Valley companies that produce their chips at TSMC, so Apple, Google, Facebook, and many others, they've been very happy with TSMC playing this central role because they've gotten the components that they needed.

12:39And so there's no one in Silicon Valley that saw this is problematic. As this has happened, the U.S. ability to defend Taiwan has, I think, deteriorated because China's military power has grown and grown and grown each year. If you go back to the 1990s, the Chinese would have been deeply afraid to threaten Taiwan because it was obvious if there was a war, who would win? You can think back to 1996 when there was a crisis in the Taiwan Straits. The Chinese shot some missiles through Taiwanese airspace. And President Clinton at the time ordered an aircraft carrier to sail through the Taiwan Straits and the Chinese stood down.

13:11And today, I think any U.S. president would be very hesitant to take such a move because China's military is so much more powerful today. And there really wasn't anyone putting together the pieces, looking at these two trends simultaneously. Defense officials knew the military balance was changing. People in the tech world knew that Taiwan is going to be more and more important for semiconductors. But no one was looking at these two trends simultaneously. And it wasn't really until the pandemic when most people woke up and said, wow, this is extraordinary concentration and it happens to be in a geopolitical hotspot.

13:38So no one knows the answer to this, Professor Miller, but let me ask you this and get your opinion at least. The fact that 90 % of these semiconductors are being made in Taiwan, does that make it more or less likely that the Chinese will make a move to reacquire the Taiwan or the island of Formosa? Well, I think there are some people who say surely the Chinese would want to attack Taiwan now to get a hold of the chip making facilities. And it's just not that simple, because Taiwan's chip making facilities only work by importing all sorts of chemicals and components and software tools from Japan, from the US, from Europe.

14:15And so if there were a Chinese takeover, the facilities would stop work. And moreover, you need the people involved, too. It's the engineers who work at the facility that make the facilities actually function. And so if they were to leave, again, the facilities wouldn't be worth all that much. So it's not so simple to say China's going to march in and take over the chip making facilities. But it's certainly the case that the Chinese government is unhappy about the fact that over the last couple of years, the U.S. has been leaning on Taiwan to stop selling China cutting edge chips. And today, if you look at U.S.

14:49export controls that ban the transfer of certain types of chips to China, many of these export controls target chips that are designed in the U.S. but manufactured in Taiwan. And so there are plenty of voices in China who blame Taiwan for going along with U.S. export controls and use that as a justification for escalating tensions. All of which is true and fascinating. I guess the ancillary question would be, maybe they would wait until we repatriated those foundries and move them back to the United States, reshore them, if you will, or put them somewhere else, because it is of too much national interest and national consequence for the U.S.

15:26at this point. And if that was no longer the case, it may be easier from a national interest perspective and the balancing of national interest for the Chinese and the American government. What do you say about that? Well, I think over the next decade, there's no chance that we repatriate a substantial share of the cutting edge chipmink capacity in Taiwan. We're going to repatriate some of it. TSMC is building some facilities in Arizona. Samsung is doing the same. Intel is doing the same. But the reality is there's so much capacity in Taiwan and they're investing so much more year after year that they're going to remain an absolutely critical player for many years to come.

16:04so you know you you you have a couple of characters in this book uh the book is fascinating by the way and you're a great writer so i'm going to name these characters i want you to react to their personalities and who who they are in your book uh let's start morris chang well i think morris chang is one of the most underestimated entrepreneurs the last hundred years people know bill gates and steve jobs but it's morris chang that produced the chips that all of us rely on In your iPhone, there's chips that Morris Chang's company makes and the data centers that process your data in your car as well.

16:35He's a fascinating character because he spent most of his life actually in the U.S., most of his professional career in the U.S. and moved to Taiwan to found TSMC only after being passed over for the CEO job of Texas Instruments, where he'd worked for a long time. It's an amazing story. And he took one of my friends, Bill Tai, with him. Bill is very proud of his, I think he's employee number one, two or three of TSM, Taiwan Semiconductor. So Morris is obviously a fascinating guy. Let's go to Gordon Moore. Gordon Moore just passed away earlier this year, was one of the co-founders of Intel, and coined the phrase Moore's Law, which predicted that the number of transistors per semiconductor, and therefore the processing power of each chip, would double every two years.

17:18And that's remained the case for now over half a century. Is it still going to remain the case, or will it eventually get to the end of that replication? Well, eventually we'll get to the end, but we've got a pretty clear line of sight for at least 10 more years of Moore's law doubling. Beyond that, it's impossible to say. So stuff's going to get way faster, way more complex. The artificial intelligence, the ideas behind it will be passing through faster processing chips and faster speeds. And again, if it's doubling every year, it'll be exponential 10 years from now. Is that not fair to say? That's absolutely right.

17:54And compare that to any other sector of the economy. You know, nothing else doubles over many decades. I like to use the analogy of just imagine if airplanes flew twice as fast every two years. At this point, we'd be many times faster than the speed of light. That's actually happened in chip making. They've doubled for so many decades that we've now got a billion fold increase in computing power relative to the first chips that were produced in the late 1950s. The very virtue, the very fact that you and I are talking over this computer to each other like George Jetson would have been doing in the 1960s in a cartoon is an example of that, right?

18:27Yeah, that's right. So let's go to the U.S.-China rivalry for a second. The U.S. is trying to restrain China's technological growth. And the Chinese obviously want to domesticate their chip production. So all of this you spell out in the book. for our listeners sake, tell us how they're trying to, the United States is trying to restrain the Chinese and what materials are no longer supplying to them and how are the Chinese retaliating and who do you think wins that or doesn't win that? What do you think the outcome will be? Well, right now, basically every chip making facility in the world has machine tools from just a couple of countries, the US, Japan, and the Netherlands are the three most important, whether it's a facility in Europe or Japan or China, you basically need to buy tools from these three countries.

19:12And that's been part of China's strategy over the last several decades is buy lots of tools from these companies and use it to build up manufacturing capabilities. And what the US, as well as Japan and the Netherlands have done over the past half a year is restrict China's access to cutting edge chip making tools with the aim of stopping China from producing cutting edge chips. And right now, China's already several generations behind what can be produced in Taiwan, for example. And without access to these tools, which are some of the most precise and complex and expensive tools ever made, China will find itself struggling even more to catch up.

19:44One last character involved here, Secretary William Perry, former Secretary of Defense during the Clinton administration. Tell us about him. Perry's really an extraordinary figure. He was a Silicon Valley entrepreneur, but he was focused on defense technology and intelligence technology. And so he was someone who stood at the intersection of all of the extraordinary changes happening in Silicon Valley, but also thought about what were the national security implications. And he first went to the Pentagon as an undersecretary for research and engineering in the late 1970s, and was the person more than anyone else responsible for convincing the Pentagon to bet very heavily on computerized weapon systems, because he realized that Moore's law was happening, that it would continue for some time, and that it gave the US an extraordinary advantage, not just in commercial technology, but also in military systems.

20:30Back to China, though. Secretary Perry is very worried about the Chinese. He speaks regularly about it on CNN. So I want you to brief me. You're the national security advisor to the United States, and you're going to send a brief memo to the president. And so we're working on it together. What are you going to put in that memo? Well, I think the Chinese want access to advanced semiconductors because they realize that advanced ships are critical for training artificial intelligence systems. And the fact of the matter is that today you can't train an advanced AI system like any of OpenAI's GPT systems without access to a tiny number of chips called GPUs, which are produced almost exclusively by Americans.

21:09These are the chips that China most wants because of their AI relevance. And that's exactly why the U.S. is trying to stop China from acquiring these chips so that the U.S. can keep an advantage vis-a-vis China when it comes to AI. Are we ahead or behind the Chinese when it comes to AI? Well, it's a hard question to measure. It's hard to know what you would measure. You can look, for example, at who's got the most widely cited AI scientists. There, the U.S. is far ahead. You can try to measure who's got the most advanced research in AI, but it's actually hard quantitatively to know who's got more capabilities there.

21:46I think, though, it's not a coincidence that OpenAI was founded in the U.S. and not in China. When you look at the commercialization of AI, what you find is that I think the U.S. is far ahead of what you've got in mainland China. So, again, I'm going back to our brief. You and I are working on this for the American president. How do you feel about the Chinese right now? Are they – let me speak as a person that has some knowledge of China, or at least their point of view. They're not a threat to anybody, they would say. They're focused on their region and they're not a group of people over 5 ,000 years that have generally attacked other people.

22:23I'm not saying there aren't instances where they've done that, but in general, they don't see themselves as hegemonic as much as they see themselves as an inward country building. And so they don't feel that they're a threat to anybody. But do you feel that they're a threat to anybody? Well, I think as you said in the early portion of our conversation, they certainly are a threat to Taiwan. And the Chinese have a law in the books that commits themselves to either, quote, peaceful or, quote, non-peaceful reunification, which is about as explicit a military threat as you can get. And that matters for this AI discussion because militaries are thinking very seriously about how can you apply artificial intelligence and autonomous systems to military and intelligence use cases.

23:03We're already seeing in the Russia-Ukraine war the earliest iterations of this, and there's no doubt that in 10 years' time, the world's militaries will rely more and more on AI. And so the question of who can apply these technologies to defense systems first is going to shape the balance of military power. Okay, let me put my Chinese hat on for a second, though. Okay, that is our territory, and we never gave up the fight after 1948. That is still our territory. And by the way, our patience being quite a virtue, we were able to reacquire Hong Kong peacefully in 1997. And we intend to do the same thing with Taiwan.

23:39Why is it different, Professor Miller? Well, there's hardly anyone in Taiwan that wants to be, quote unquote, reunified with China. So there's no pathway to do it peacefully because the Taiwanese don't want it. So when the Chinese talk about, quote unquote, reunification, and remember, they haven't actually, China, Taiwan has actually historically not been part of China. It's been ruled by all sorts of different empires, and China hasn't ruled Taiwan for over 100 years. But when you talk about, quote unquote, reunification in China, there's no way to do that other than via threatening force because the Taiwanese don't want it.