The innovation, use and experience, and exchange of new and emerging technologies today are influenced by the role that China plays in global politics and economy.
Recorded on April 18, 2025, this Matrix on Point panel brought together experts of the Chinese political economy and law and society in a conversation to discuss the political, economic, security, and social dimensions and complexities of technology in China’s internationalization during times of global tensions. Topics covered included the institutional foundations of China’s technological development, technology governance and industrial policy, global technology competition, and legal technology and societal impacts in today’s China.
The panel featured Mark Dallas, Professor of Political Science and Science, Technology, and Society at Union College; Roselyn Hsueh, Professor of Political Science at Temple University and Visiting Scholar at the Berkeley Economy and Society Initiative; and Rachel E. Stern, Professor of Law and Political Science at UC Berkeley. AnnaLee Saxenian, Professor in the School of Information, served as chair and moderator.
Matrix On Point is a discussion series promoting focused, cross-disciplinary conversations on today’s most pressing issues. Offering opportunities for scholarly exchange and interaction, each Matrix On Point features the perspectives of leading scholars and specialists from different disciplines, followed by an open conversation.
The panel was co-presented by the Berkeley Economy and Society Initiative, and co-sponsored by the Institute of International Studies (IIS), the UC Berkeley School of Information, and the Charles and Louise Travers Department of Political Science. This public panel is a part of the two-day Bringing the Sector Back In conference, also co-sponsored by the Institute of East Asian Studies and the Institute of Slavic, East European, and Eurasian Studies.
Watch the panel above or on YouTube. Or listen to the audio recording via the Matrix Podcast below (or on Apple Podcasts).
Podcast and Transcript
[MUSIC PLAYING]
[WOMAN’S VOICE] The Matrix Podcast is a production of Social Science Matrix, an interdisciplinary research center at the University of California, Berkeley.
ROSELYN HSUEH: Hello, everyone. I want to welcome all of you to here, in person, as well as online. My name is Roselyn Hsueh. I am a visiting scholar at Berkeley Economy Society Initiative, and also a professor of political science at Temple University.
And I’m delighted to welcome you to the technology and China in the new political economy panel. This panel is the culminating panel of a very stimulating, bringing the sector back in workshop, that we just had the last two days.
And I’m delighted to have AnnaLee Saxenian, who is the professor at the School of Information at UC Berkeley and also the former Dean of the School of Information, chairing today’s session, and also moderating us and joining our conversation, so on the social, political, economic, and security implications and consequences of China in today’s technology. So thank you so much. And without much ado, let AnnaLee go. Thanks.
ANNALEE SAXENIAN: Thank you very much, Rosie. I am delighted to introduce this very interesting group of scholars who will speak to you. First, we’re going to, the organization is that we will each, each of the panelists will speak for about 5 to 10 minutes about their research. And then we’ll have a discussion of some contemporary issues, and then we’ll open up to questions from you guys.
But first, I want to introduce Roselyn Hsueh, professor of political science at Berkeley. Not at Berkeley, but she has a PhD from Berkeley. She’s at Temple University. Yes, to my far left is Rachel Stern, who is a professor of law at the Berkeley Law School.
And to my right is Mark Dallas, professor at Union College, who is currently on leave at the US Department of Commerce. So I am going to turn to Roselyn to start a conversation about the institutional foundations of state business relations and China’s technological development.
ROSELYN HSUEH: Yeah, so I want to just provide a little background on what I call the institutional foundations of technological development, and also the consequences of technological development in China and the role that China now plays in technology, how state business relations and market governance in China and the role of the Chinese state plays a role in that.
I want to make three points in this time I have. And point number one– and I’ll go into greater detail for each point. Point number one I wanted to make is that the interconnectedness between China technology and China’s distinctive global economic integration, particularly since China’s entry into the World Trade Organization in 2001, is critical for the role that we see China today, that China plays in technological development, advancement, as well as even the responses globally on China’s growth and technological development. So that interconnectedness, that distinct global economic integration, I’m going to talk about.
Number two, I am going to then also emphasize that there has been sectoral variation in that dominant patterns of market governance, that economic integration into the global economy, and that then affects the industrial policy that has shaped technological development in China.
That sectoral variation in the market, governance that comes along with it, will then lead me to the third point, which is that throughout the last two decades, from WTO accession into the global financial crisis and the rise of Xi Jinping, and then also, the trade war that we see and where we are today, the new political economy, what we see is that, because of that sexual variation, we do, indeed, see the preponderance or the exponential growth of private initiatives.
And, at the same time, we also see political consolidation of the state. So it’s both. And that bifurcation is very much to do with that distinctive economic integration. And so I’ll begin with there. But those are the three points I would like to make.
And so in my own work in the first book that China’s Regulatory State, a New Strategy for Globalization, which took me to China to right around the time of WTO accession, what was really, what stuck out to me, empirically and very different from, intellectually within political science, what was the story about the developmental state was that during this height of neoliberalism, the height of the global, when we could call the liberal international order, we saw that China was, indeed, liberalizing, that it became a member of the WTO and, indeed, made these commitments to liberalize.
But what I also saw empirically was that, as I traveled across the country when I first went to China, was that we saw not just some subnational regional variation, but a lot of sectoral variation. And that what I, then, called the liberalization 2-step. China liberalized on the macro level and then reregulate at the sectoral level. And that reregulation at the sectoral level really then dictated the industrial policies. That was the institutional foundation of technological development.
And so what we had, and this is where the sectoral variation comes in, is that in sectors that are perceived to be of strategic value to the state, capital intensive, value-added sectors that have application for the national technology base that contribute, I mean, contributes to the national technology base to Indigenous development, industrial development, and have applications for national technology.
Those are the sectors, then, the state will have more involvement in, will participate more, will intervene, and will partake in industrial policy that, in many ways, is what many of us might think is the typical pattern, which is that state capitalism.
It must be because of state-owned enterprises and the role that state owned enterprises play in, and the role that the state plays in mobilizing research and development, and the different ministries coming in, particularly at the central level, at the state council level, all the different working groups mobilizing for that technological development. That’s the dominant view. And that, indeed, was true in these sectors that contributes to national technology base and national security.
But then there’s also the other view. There’s a bifurcated picture, which is that, in the sectors that are labor-intensive, less value-added, and have less applications for the national technology base and implications for national security, but nonetheless is important to regions and to some national governments for economic development and economic competitiveness. That’s where we have decentralized market governance.
And that decentralized market governance, we will see some national governments institute industrial policy, and we will see that there are going to be that growth of private sector in China and that, indeed, that distribution of property rights that we see in China today is very much part of that liberalization 2-step.
And how do we make sense of it? Well, it’s that bifurcated capitalism. It’s that bifurcated technological security developmentalism that allows us to see that sectoral variation in those dominant patterns. And those dominant patterns have laid the foundations for the different type of industrial policies, the extent and scope of state control, and the dominance of different property rights arrangements across sectors.
And these have huge consequential implications for China’s technological development, and also, importantly, with how foreign direct investment have been strategically used by the Chinese government for technology transfers, for knowledge transfers. And some of the state methods include rules on market entry, rules on business scope, rules on investment level. And unsurprisingly, in the more centralized sectors, that’s where the government may have more of an intervening role.
At the same time, that doesn’t mean that the government doesn’t involve private business, doesn’t involve foreign direct investment, doesn’t involve the role of decentralized actors, but that role of centralized governance will continue to be there. And we see these dominant patterns play out through the last several decades. And that is true across time, including beyond the political consolidation and centralization that we have seen through Xi Jinping.
And one more point I just want to make is that, that predominance, preponderance, and exponential growth of private initiative, that use of foreign direct investment, that use of private initiatives and private investment and private development of technology, and the role that state plays will have implications at the business level.
And we see that the telecommunications infrastructure, the backbone infrastructure, remains state-owned. But the value-added services, so it goes to the subsector, the value-added services that operate on top of telecommunications backbone infrastructure– the Alibaba’s, the fintech, the Alipay, even TikTok, and the social media– they operate on top of state-owned infrastructure. So they are, many of them, which are private initiatives and foreign direct investment.
At the same time, we do see if political consolidation for the Chinese Communist Party becomes an important imperative, there will be state intervention. And the last point I’m going to make is that we saw that in the 2020s in the internet blitz that intervention into fintech. And we continue to see that in different sectors of high tech.
And so the distinct global integration, the sectoral variation in dominant patterns of market governance, therefore, also industrial policy, the role of multi-level governance, different actors and the wide range of distribution of property rights, including foreign investors, but with that strategic value logic, really have shaped technological development in China today. And so that’s the institutional foundations that I want to leave us with.
ANNALEE SAXENIAN: Great.
ROSELYN HSUEH: Thank you.
ANNALEE SAXENIAN: Thank you so much for providing a framework for us. I’m going to turn this over now to Rachel Stern, who’s going to speak to us about legal technology.
RACHEL STERN: I am. Hi, everybody. Good afternoon. This is such a wonderful first for me. I’m a little bit surprised to find myself– this is my first time on a panel on political economy and on a panel on tech. Those are both firsts.
I study state society relations, and I study social legal studies. But I found myself I’ve been working on it for a couple of years on a comparative project, where the two key case studies are China and France, and I’ve done field work in both places on the politics of access to legal information.
And through this project, once legal information, specifically, I’m looking at court decisions, once both countries made a decision to make millions of court decisions available, hundreds of millions of court decisions, a legal tech industry sprung up to capitalize on this new data. So I found myself knowing something about legal tech.
So that’s what I’m going to talk about today. I want to talk about it a little bit first, just as a sector. What I mean by legal tech is it’s the use of software and technology to help legal professionals do their job more efficiently. So that’s a lot of stuff. That’s everything from automated contract drafting to e-discovery software, to billing and accounting software for lawyers.
What I’m interested in is, I’ve been interested in data analytics, legal analytics. So software that takes those millions of cases that are now publicly available, that are now in the public sphere and puts it into a dashboard. And what it delivers is information that can help predict case outcomes, analyze judges tendency to rule one way or another way, or court’s tendency, or even to look at opposing counsel.
And I’ve become, especially, interested in how courts themselves, at least in China, are using legal analytics. And I sort of backed my way into an interest in sectors, partly, by knowing Rosie and reading her really excellent piece that’s forthcoming, I think, in perspectives on politics. But part of it was it gave me a way to think about legal tech as a sector.
And what has resonated with me about a sectoral approach is that legal tech is both embedded in the global arena and in the national arena. And that’s been one of the things that’s hard for me to write about in untangle, that Rosie’s work has helped me to see.
And let me just tell you a little bit about what I mean. It’s embedded in the global arena because it’s adjacent to AI. All of these legal tech companies will tell you they’re doing AI. I mean, it’s contested what is actually AI and what’s not. But this is all like taking AI into the legal space.
And, of course, AI is, this is about global geopolitics. We’re in a global race for AI. It’s a high value strategic sector for everywhere I can think of, but, certainly, in my case studies, certainly in AI, certainly in China, certainly in France. And so thus, you have to talk about geopolitics.
And it’s also global because if you do fieldwork as I’ve done and you go to legal tech offices in different countries, it becomes really clear that they’re all really oriented towards Silicon Valley as their model of how to operate. And they’re more similar than they are different.
Like if you go to a legal tech company and probably some people in this room have been to them, what are people going to be wearing? Hoodies, sneakers. I promise you, there’s going to be breakout rooms, there’s going to be free coffee, there’s going to be free snacks, and there’s going to be a lot of slogans about disrupting something. That’s the orientation.
So this is really interesting for those of us who study the legal profession, because it’s much more so. These legal tech companies are much more Silicon Valley than they are big law. They’re much closer to any startup in the Valley than they are to Paul Weiss, for example.
And so it’s been very, just their very existence in– entry into the legal profession has been disruptive for the legal profession itself. So that’s what’s global. But then, nationally, there’s a lot that’s going on nationally as well. And I think this is what’s interesting about legal tech, is AI adjacent.
It’s a little bit of a backwater. Like, these companies are operating in what’s sometimes called a zone of public inattention. I mean, I think it’s not an accident that nobody ever talks about legal tech on panels like this one. It’s not DeepSeek. Nobody’s ever heard of these people. And so that adds an interesting kind of layer to it.
And second, it’s mostly national markets because legal systems are mostly national. The statutes are national. The case law is national. Law is stubbornly national. So it’s impossible. Not hard, but not impossible to create products that crisscross jurisdictions.
And then sectorally as a sector, we see variation in regulation across countries. So one of the things that piqued my interest in this project was in 2019, France just decided that they were going to ban judicial analytics.
They just banned the whole thing. They said, you can’t run any data on judges. You cannot run and publish data on the case records of individual judges, we’re just going to make that whole thing illegal. Just an example.
So let me bring it to legal tech in China because we’re talking about China today. And I have, and here, I’m going to draw very much on work that I’ve been doing in collaboration with Ben Liebman at Columbia Law School. So he’s been a big part of this as well.
We’ve been looking at legal tech in the courts in particular. And, for me, at least, having also just spent a year in France, one of the takeaways is to just ask, how much of what happens in China is authoritarian, and how much isn’t?
We were talking about this a little bit. It echoes a remark that John made yesterday. Obviously, a lot that happens in China is related to authoritarianism. But I think by the time you get down to the sectoral level and you’re looking at legal tech in particular, there’s also a lot that isn’t about authoritarianism. Let me give you some examples.
So first, in terms of what technology is doing in the courts, we’re interested in this because technology is front and center for the Chinese court system. This is not a secret. It’s in every high level policy document.
They’re really, really interested in integrating and using technology to do their jobs more effectively, and to some extent, to change what the job is, to analyze their own record, to learn about the kinds of disputes that are happening in society, and to become a smarter learning state. So that it’s front and center. And, of course, technology is being used to serve party priorities. It’s a one-party state. This is not a surprise for any of us who study China.
What’s been interesting for us is to see how technology can be used as what we call a policy accelerator, when the party’s priorities shift. So we saw this big quick example. We saw this big push for technology starting in the 2010s.
A man named Jun Zhang was in charge of the Supreme People’s Court. He had a whole series of initiatives called making Chinese courts smart. And a lot of the push for technology, at that point of time, was about bringing cases into the courts, increasing the caseload so that a lot of technology was introduced to make it easier to file cases online. And it was in line with the goal of accepting all cases that should be accepted.
So all these reforms seem spectacularly successful. The number of cases filed surged from just about 14 million in 2013 to 45 million in 2023. So it flips, the party becomes concerned about too much litigation. And Xi Jinping, in 2020, was talking about his fears that China would become a big litigation country.
Anyone know a big litigation country? I’m pretty sure we’re sitting in a big litigation country. So then the use of technology changes and the courts start rolling out what they call litigation risk assessment tools, where people who are thinking about filing cases can go see what their odds are of winning.
And for various reasons that I can talk about, including just the name of the litigation, risk assessment assumes that litigation has a risk. But the publicly acknowledged goal of these tools is to help litigants, rationally, choose methods for resolving disputes and to reduce the public litigation burden. So the deck is stacked. These are tools meant to nudge people away from litigation.
And yet not everything in this space is about authoritarianism. I think a big push for the use of technology has just been that the courts are overburdened. The Chinese, they talk about the too many cases, too few judges problem. Just making the courts work more efficiently to handle that massive caseload.
But I think here looking comparatively is helpful as well. This is not a problem that’s unique to China, nor is it a solution that’s unique to China. I’ve been doing some interviews recently in Taiwan, where they’ve also been experimenting with the use of technology to say, generate decisions because they also have really overburdened courts and that are under strain.
Let me conclude by just saying that where China stands out for me is with the court system’s unusual enthusiasm about technology. And this is, I happen to be looking at legal tech. I think the enthusiasm about technology is probably something that you can notice looking at almost any sector.
It’s been noticed by lots of people, sometimes called the fetishization of technology. But I think, for me, as someone who’s been looking at the Chinese legal system for a while now, I think it raises interesting questions about the future of authoritarian law and legality.
So even if the modest changes on the ground don’t live up to the court’s hype about what’s happening on the ground, because of course, there’s political incentives to oversell what you’re doing and how cool it is and how futuristic it is, I think it’s clear that legal tech is an area in which the courts want to be seen as pioneers.
So if artificial intelligence stays a party priority, which I think it’s likely to, if the courts continue to experiment with it, which I think they’re likely to, I think that Chinese courts are going to become more and more visibly futuristic, looking, at least to our eyes sitting here in the United States.
So I think what we’re looking at is where, I think, the Chinese courts will enter our popular imagination is I think they’ll come to challenge, in some ways, still common popular assumption that authoritarian legal systems are somewhat behind the times, or even benighted. I think that increasingly, that the way in which technology is being integrated in the court system is going to make the Chinese courts look like the future. And so I’ll leave it there. Thank you.
ANNALEE SAXENIAN: Great. Thank you so much, Rachel. It’s interesting that we’ve accepted that Chinese cities are technologically advanced, their infrastructure. But somehow, the courts, we don’t think about that. That’s a fascinating presentation. And now, I’m going to turn to Mark Dallas, who will speak to us about global value chains and US technology competition.
MARK DALLAS: Well, thank you very much for the invitation. As Ann said, I’m on leave from University. I’m working at the US Department of Education. So I have to say that everything I say today is my own private view and doesn’t reflect the view of the US government or US government agencies.
So I’m tasked with talking more about the global dimension, US-China relations. But then also just China in the world focused on global value chains. And I want to hit on three topics. One is to explain a little bit of what’s the big deal about global value chains. But really to focus on the scoping of security.
So a lot of the discussion today is security-oriented. When I started to study about global value chains, it was all about development issues– Chinese firms upgrading, encouraging development.
Well, that conversation over the past five or six years has really shifted to one of all about security– supply chain resilience, vulnerabilities, and whatnot, let alone other higher level military and other types of security issues as well. So I want to discuss a little bit about the scoping of security and how that concept can just keep growing and growing until everything becomes security.
And finally, which almost contradicts that second point to a certain extent, is a tension of technologies today that have become infrastructures. So internet as infrastructure, cloud computing is infrastructure, even social media as infrastructure, AI servers as infrastructure.
And I know that’s all digital-oriented, but that’s one of the concerns, is technology is not only infrastructure but it’s global scale. And it crosses all these countries. And how do we deal with that situation.
There are sovereignty issues involved there. I mean, if you think of security issues, cybersecurity is very different because there, you have nation states literally inside of each other’s networks. Most security issues don’t deal with that. It’s border issues, security on borders. Military being, obviously, the most typical example of that.
So let me start with global value chains. So just very briefly, previously, you can think of firms as used to be vertically integrated, meaning they made everything from the raw material all the way through to the final product. That’s a vertically integrated firm.
Over many decades, the firm started to break up. And they broke up more and more and more. They outsourced, and they also offshored. So they internationalized. And then they had other companies do other things. And that value chain can get fine, sliced very, very minutely. One company does one process, one task, but it has a global market to sell to.
And what that created is it created a situation of functional integration. That’s different from just globalization. We had trade, global trade, we had global investments, and things like that. Those are resources moving across the world. That’s not new.
What’s new is this functional integration. And what we’re left with is an intense degree of interdependency globally, but also global scale monopolies. In a lot of sectors, 1, 2, maybe 3 firms, sometimes dominate globally in that little minute task that they’re doing. And sometimes it’s not minute. Sometimes it’s a bigger task.
Well, China is kind of, partly, the poster child of a lot of this. They’re deeply integrated in lots of these tasks all over the place. Some are massive scale and whatnot. And so the idea of functional integration, how it differs is you’ve probably seen studies, there’s a study about trade, US-China trade or China trading with the rest of the world, studies about investments, Chinese investments inward or outward whatnot.
But a global value chains, there’s services. There’s IP licensing. There’s software. There’s also PhD students, like, where are they in the world. How much research papers are being– so this is basic research or University research. How many research papers are being published by who and who’s collaborating and everything– patents.
So those can each be a study. And they are people study that. Each of those is a topic. Let’s study it. Let’s look at the relationships, and whatnot. And they’re all great research. I’m not trying to say anything about that.
The global value chains integrates all that together. If you start with a product, let’s say the cell phone, it has all of those things all intermixed together. And that’s where the functional integration happens. So all of those. It’s not separate topics. It’s all these topics integrated into one thing. And that’s really where you start thinking about functional integration.
Now, in a world in which countries are friendly and it’s about development and win, win, no problem. That’s not an issue. And that’s the world we lived in for a very long time between US, China, China and other countries, of course, Europe, around the world. When that starts to– when the lens shifts to security, now you start looking at those value chains very differently. And that’s what has really changed.
I want to just give you a sense of how interdependence can be. What it is, and then how it can be perceived in very different ways. And so now we’re in a security world. And so now people are starting to scratch their head. And that’s where all these terms come from– nearshoring, friend-shoring, decoupling from China, de-risking from China.
There’s endless number of new words out there to deal with this. How do you deal with vulnerabilities from supply chain? Of course, COVID added to this too. We couldn’t get access to vaccines and masks and everything else. So people were like, wow, supply chains.
And, by the way, I’ve been studying that for a long time. Never was it the headline of newspapers. And then, suddenly, especially with COVID, it was headlines. I was, like, oh, wow, what I studied, actually people care about now. They actually know what the word supply chain is. So that was I guess a nice shift.
And now, the second issue about security and what are the boundaries here. I think this is a real quandary because you can already probably imagine about this interdependency and how that could lead to types of security. But the security dialogue is also shifting towards military. So advanced semiconductors.
When I said that, when I said advanced semiconductors five years ago, people were like, oh, that’s cool. We all benefit from advanced semiconductors. Today, the conversation is, well, this is going to go into supercomputers in China that are going to develop nuclear weapons or other types of things like that or AI.
AI should not be considered a military thing. There’s tons of scientific and other wonderful things, legal and everything. But the conversation is shifting. It’s also a military thing. It is a military thing. But the problem is with these emerging technologies, how do you parse that?
Some technologies like nuclear, we could parse the civilian and the military side of things relatively more easily. With a lot of these other technologies, I don’t know. It’s hard, where we’re trying to figure out what that parsing is. But right now, the security side seems to be winning.
And that’s a tremendous problem. Because we don’t know the boundaries of that what should and shouldn’t be regulated. What is a concern and what’s not a concern. So I don’t have an answer to that, where the boundaries are for security. But it’s one of the main issues we’re dealing with.
The final thing I’ll say is about digital technologies, in particular, as infrastructure. I think it’s, particularly, the case with digital. I think it can be applied elsewhere. But this is another type of what is, when we built the internet, no one was really thinking about security issues of famously, it was just, it was the open internet and everything’s good.
And once that infrastructure gets built, then, suddenly, like, oh, wait a minute, bad actors can do stuff. And I’m not just saying about state actors, like mostly non-state actors, can do bad things. Well, oh, geez, the infrastructure is already built. It’s very hard now to securitize infrastructure that’s already been built.
And so, that’s the classic case from the ’90s when infrastructure internet became sort of commercial, was passed off from government military to commercial purposes. But you build on top of this. You’ve got the internet, but then how does the internet work? There’s all this hardware, but then you build on top of the internet, obviously, social media and other things. So there’s a layered stack here of different integrated infrastructures that are largely global.
Now, there’s a lot out there on how countries are trying to regain sovereignty or trying AI sovereignty, data sovereignty, and whatnot. So we’re struggling to figure out how do we strike this balance. I think, in some cases, it’s going to be extremely difficult.
There are, just again, going back to that idea of intense monopolies. And I also just want to give you a sense of the– we hear so much in the United States about US security concerns. Let me just give you just two quick empirical examples from China’s perspective. And again, I’ll use mobile phones as an example.
There’s two operating systems in the whole world right now– it’s Apple and Android, Google and Apple right over here, right next door to each other. They control the operating systems of every single phone on Earth, including all the ones in China. A little bit Android, you can customize and stuff, but the underlying code is that.
Huawei, recently, is trying to develop their own operating system and break from that. So imagine every single phone in China is run by– the operating system is run by two companies here in San Francisco. That’s a security concern that we have to be sensitive to.
Another company– well, a company you’ve probably never heard of but is one of the most important countries in the world– ARM. They’re based in the UK, near Cambridge. Every single major, important chip in every single phone, the underlying architecture of it is designed by them in the UK– every phone in the world.
Not just phones, actually. ARM is not just in phones. It’s actually in every device that requires energy conservation or efficiency. So that means all of that– yeah, I’m just add one more comment.
ANNALEE SAXENIAN: Yeah, your fine.
MARK DALLAS: So I just want to give you a sense of the security concern of a country like China to think about that. I’ll make one quick comment and then I’m done. So the only thing I cringe, and this is just a generic thing because you hear it in the media a lot, is sort of tech catch up and technology is a horse race– who’s ahead, who’s behind, things like that.
I just want to, just general warning against, to people to just not tune too closely into that. It makes it sound like technology is one linear race, like a horse race, and you’re either behind or you’re in front and it’s just not. It’s so multilinear.
There’s so many different pathways to it. So just as a general, I don’t know, public announcement like just avoid that. Avoid that, if you read about it, which you will all the time. And I will make the same mistake. I’ll be like, who’s behind? Who’s ahead? You almost immediately fall into it, so.
ROSELYN HSUEH: I just want to really quickly make a comment about the economic security nexus. I think Mark does a good job in laying it out as that the technological or the structural attributes of sectors and the functionality of sectors and evolution of that, how that affects the economic security nexus.
But then, I want to bring politics back into that as well. And we can take this from the Chinese state’s perspective. And I think the SMIC, which is China’s, quote-unquote, “state-owned foundry” and the evolution of the corporate governance structure of that company is a really good example of that evolution of the economic security nexus.
So SMIC, in 2000, was foreign direct investment. And it was part of that kind of strategic use of FDI. But over the years, because of China’s response to internal and external pressures– I mean, in addition to technological attributes and advances, but in response to those internal and external pressures on the economic side, but also the security side, globally, but also perceived security issues internally, we saw, over time, that SMIC, through corporate governance interventions, actually is now from foreign direct investor that was foreign invested with the Shanghai state government and Peking University professor by 2010, before Xi Jinping even came into power, was, and became a state-owned enterprise, and dominated with state-owned, as well as state-controlled private entities as investors and on its board, the Chinese Communist Party. And so I just want to bring the politics back into the economic and security nexus.
ANNALEE SAXENIAN: Well, actually, just to pile on the politics, I mean, I think that a lot of the ways that US AI companies sell themselves to the state or this is security. I think we see lobbyists from OpenAI, from all of these companies in Silicon Valley, Google, going to Washington regularly and talking about the threats of falling behind China. And, therefore, we must, the US must invest in AI. So it’s politically a very–
MARK DALLAS: I think that’s the boundary conditions, is what is not security now. And security just cells–
ANNALEE SAXENIAN: Absolutely.
MARK DALLAS: –in the media and elsewhere.
ANNALEE SAXENIAN: So I want to just add a few comments about the research that I’ve done on Taiwan because US-China, we always frame it US-China. But Taiwan’s in the middle, and it’s going to stay in the middle. And it’s an interesting position.
In the 1990s and 2000s, I was studying communities of engineers who were educated in the US, but working in Silicon Valley. The story that I tell is about the way that they built very powerful institutional, economic, and technological ties between Silicon Valley, and then in Taiwan, and then after that, China.
And they built these ties because they were going back. After coming to the US, studying almost anywhere, Texas, Arkansas, they then took their knowledge, worked in Silicon Valley, marinated, learned the Silicon Valley model of doing business and learned the technology really well. And then they went to their homes, their home countries, whether it’s first Taiwan or I’ll tell the story of Taiwan. And they started companies there.
They started investing in startups. This was the early semiconductor industry era. Or even earlier, Taiwan was doing low-skill chip assembly, and then games assembly for Atari. And then they started manufacturing for IBM, the IBM PC and the Apple PCs.
So you see that the story I want to tell is that these people going back and forth between Silicon Valley and Taiwan were really the best of the global era, because they were transferring technology, they were building social and business ties that allowed both Silicon Valley to flourish and Taiwan to learn and move up the value chain. And Taiwan became, very, very quickly became very good at manufacturing and at manufacturing semiconductors and electronic systems. So good that they were really better than the US.
So starting as a low-skill assembly center, Acer, if you remember Acer technology, that was a Taiwanese company. You don’t hear much from them anymore. But quickly, through this back and forth and learning by doing and a lot of other state investments, a lot of processes, they became the leading in the world, really, in terms of the quality of and the speed at which they could design and manufacture motherboards, electronic systems, and also chips.
You hear now about TSMC. But I want to suggest that it goes deeper than that. So by the end of the ’90s, Taiwan, through this brain circulation, I call it, had have become a central node in the global value chain for manufacturing electronic systems.
In the 2000s, this value chain, this piece was moved lock, stock, and barrel to China, to Southern China, to Shanghai. And it was the same people. I mean, and so then you saw this linkage between Silicon Valley and Taiwan and China, and you saw the same people investing and starting companies, with the know-how of the, then, emerging semiconductor industry.
So the transfer of technology from Silicon Valley, Taiwan, and then to China means that now China takes over a piece of the value chain, and Taiwan needs to move up the value chain, and so does Silicon Valley.
So Silicon Valley becomes, for all of these advanced technology products, the design center, maybe the architecture, center of architecture. Taiwan moves up, becomes better at manufacturing mobile devices and things like that, more advanced semiconductors.
And then China enters first as a low-cost assembly and then builds up its own clusters of suppliers, which are, by the way, mostly managed by Taiwanese managers. So the transfer just happens across the Taiwan Straits. This process was– and actually, SMIC was created in that period. And a lot of investors from Silicon Valley and Taiwan involved in this forming of SMIC.
ROSELYN HSUEH: The founder, Weijia Jiang, is Taiwanese.
ANNALEE SAXENIAN: Yeah, exactly.
ROSELYN HSUEH: Former Texas Instrument executive, but Taiwanese. Yeah, Taiwanese-American.
ANNALEE SAXENIAN: Well, and TSMC has also started by a TI.
ROSELYN HSUEH: Yes.
ANNALEE SAXENIAN: So we have this small world that, at the same time, that you’re seeing what Mark has described as products functionally disperse across. You see these communities of people doing it. And they all know each other. They all knew each other.
So that period, however, of the global value chain sort of spreading out via people has ended. It started to end after 9/11. The US started to clamp down on migration and immigration. Xi Jinping emerges in China. Over time, there’s been anxiety. The security anxieties have led both the US and China, I think, to shut the doors more and more and more, and today they’re sort of slamming shut.
I just want to say that we all know because it’s in the news all the time. We hear about TSMC, Taiwan Semiconductor, because that’s the leading chip foundry in the world. And it really started in the ’80s and has become, again, the– well, interestingly, just to bring this to the present, people heard of NVIDIA.
Those are the most advanced chips that go into all of the data centers that power AI today. NVIDIA CEO Jensen Huang announced that he was going to onshore his advanced chip making to the US, again, to Texas and Arizona. If he does it, that’s a big deal. We’ll see if it happens.
But four of the five partners that it announced would move there were Taiwanese. So it’s TSMC, Foxconn, Wistron, and a company called SPIL, S-P-I-L. The other one is Korea of the five. So that’s not really that different than the process that happened with Apple computer.
When Apple decided to diversify its manufacturing, iPhone manufacturing out of China to India and Vietnam, it relied on Foxconn, Wistron, and a company called Pegatron, all Taiwanese companies, to do the heavy-lifting of rebuilding or building or rebuilding manufacturing capabilities. It’s tremendous know-how. There’s tacit knowledge. There’s also deep knowledge. There’s institutional knowledge. Doing this kind of very, these are very sensitive manufacturing processes.
So they’re being, it’s the same Taiwanese that set up in China and India, moved the iPhone. And now they’re being tasked to set up in the US, which helps explain the recent decision. When China announced their reciprocal tariffs on the US, they exempted chips.
And why did they exempt chips? Because they wanted to keep getting these chips even though they were designed in the US and they don’t want US products, but they were made in Taiwan and they will not be subject to the tariffs.
It’s an acknowledgment that we all, China depends on Taiwan. We depend on Taiwan. And that, in spite of politics, this interdependence continues. This functional interdependence that Mark described.
So with that note, I guess we are at now 1 o’clock. We have half an hour left. Do we want to discuss contemporary issues or open up what people have thoughts about contemporary issues like about the tariffs, or would people like to raise some questions? Looks like we have an audience that would like to get engaged. Why don’t we start here?
AUDIENCE MEMBER: So I appreciate the talk. It was an expatriate living in China during the good times when we had reform. And it was interdependence was good, not this functional integration. And I think about the higher theme of how do we avoid confrontation and war, and go back to deterrence, cooperation, competition. What is needed for that? That’s on a whoever can figure that out. I’m happy to hear– number one.
And then number two is on the domestic side, private sector. I was just in China and Taiwan two weeks ago, and privatization is still going on very well. Good thing will Xi Jinping be able to pull it off from an economy, high domestic internal consumption, keeping high export, and then also trying to become a technologically advanced country? So will he be able to pull it off?
ANNALEE SAXENIAN: We have two questions. How to avoid war? Minor question. And then the second question is about will Xi Jinping pull off this self-sufficiency in technology? Both really big really interesting questions. Who wants to start?
ROSELYN HSUEH: I guess I’ll dive in and try to see if I can answer these questions. So I mean, if we in terms of the first question, I think, we could think about the US-China trade war, when it first got launched in 2017.
Some of the very firms and sectors that were against the tariffs, the three rounds of tariffs, they did keep piling up. Not in the same way we’re seeing today. But those three rounds, some of the very businesses and sectors that were opposed are companies that needed the components from China because they were being produced in China.
And some of the components did have indigenous technology that China was producing. And so I think one very just good way politically for cooperation would be for some of these sectors and businesses to come together and step up and say, hey, we’re not going to stand for this. It’s this interdependent world and we do depend on China.
Now, I do think that it is true. There are security concerns and there some of the diversification of supply chains make it so that, maybe there’s less need or maybe perhaps there’s less opposition from sector and businesses in this round of trade war. But the domestic private interest groups can play a very big role in global cooperation in terms of, at least, this very next round.
And then on the domestic side of issues I think I mentioned. And so in my most recent book, Micro-Institutional Foundations of Capitalism, where I do compare China to India, Russia, across industrial sectors. But on the China side, part of what Xi Jinping in the party is kind of grappling with are internal pressures.
And some of those internal pressures are not just political, they are economic pressures. And so I will, I do believe, and maybe I’ll go out on a limb to predict it, is that there will be, certain tinkering and recalibration of how the Chinese government is dealing with industrial policy, even security concerns and economic policy as well.
MARK DALLAS: I’ll just add. So you were referring really to kinetic war, like real war, not just trade war or?
AUDIENCE MEMBER: Yeah, or a confrontation with Taiwan. Just saying the distance between confrontation and war is a lot shorter than you think it is.
MARK DALLAS: Sure. Yeah, I think one of the key things that can be done and should be done more of is just communication between the countries. I mean, you’re not going to solve, I mean, in an ideal world, we solve the security dilemma and we got peace. It’s been several thousand years. We haven’t solved the security dilemma. So I don’t think that’s going to happen overnight.
But what’s been declining is communication, and particularly between the militaries. That was at a real low point during the spy balloon incident and whatnot. But that’s a major concern. I think, just general communication, I mean, military-to-military communication and whatnot, to avoid anything like that sort of mistakes.
But then communication at different levels of government also would be better. And then also among civil society, I mean, I think business has a role. But national security in both countries is very much hived off intentionally from lobbying and things like that.
And so I think there is a role, but I think that won’t be like a silver bullet to solve that. Now, in terms of, I think your other question was about Xi Jinping achieving self-sufficiency. Is that a fair way to frame it?
AUDIENCE MEMBER: You, obviously, maintain the Communist Party is critical. At the same time, he’s got to continue economic growth.
MARK DALLAS: Sure.
AUDIENCE MEMBER: I think if those two don’t happen, he’s got a big problem.
ANNALEE SAXENIAN: You’re right.
MARK DALLAS: Yeah.
AUDIENCE MEMBER: –put it off.
MARK DALLAS: Yeah, I mean, I don’t know if he can pull it off. I think the pressures on, the general pressures on China’s economy, I mean, their well-known demographic, their big structural stuff, creating a social safety net and all sorts of things to encourage consumption and whatnot, the property sector and whatnot.
In terms of, specifically, global integration issues and the drive for self-sufficiency, and whatnot, I think that’s a fool’s errand. I think the answer is going to be no. Or I think actually the better answer is going to be, China’s going to have to choose.
Do they want a second tier in terms of technological sophistication? Do they want second tier technology then they can probably do a degree of self-sufficiency? Or if they want to remain at first tier, they’re going to have to integrate.
And that goes back to my point before that, the supply chains are so fragmented and there’s expertise all over the world, and those expertise are extremely concentrated. The memory, for instance, I mean, in Korea, it’s just the global market share of two companies is just enormous.
You just can’t replicate all of that in a single even one. With as many smart people as China, with as much resources, which is much will, I just don’t think that’s possible. So I it is a bit of a fool’s errand. And I think some of these, what Rosie was saying earlier, are examples of, oh, geez. Or sorry, what you were saying about keeping semiconductors open is an example– like, oh, geez, we can’t close the door.
ANNALEE SAXENIAN: Yeah. Good point. And there’s that company in Netherlands, AML.
MARK DALLAS: ASML.
ANNALEE SAXENIAN: ASML, I mean, it is really, literally, very sophisticated knowledge all over the world.
MARK DALLAS: And China is trying to replicate that in a different way. They’re taking the EUV technology in a slightly different way. But what they’re doing with that EUV technology, which is the most advanced chip making, was done back about 20 years ago by Japanese and other companies, and they felt like that was a pathway that wasn’t going to lead anywhere. So that is a good example of second tier technology, potentially.
ANNALEE SAXENIAN: OK, Rachel has a comment on these questions.
RACHEL STERN: I want to open it up for more questions. I’ll just say one line, which is just to say that your first question about how to avoid war is a really live conversation in the China studies community. Just because the Washington Consensus has so hardened in both parties around a consensus that China is a threat. And I really think that has to do with our own domestic politics, in a way, and political polarization, and China being the thing that allows us to agree and get things done.
So anyway, just to say that people are thinking about this and thinking about other sources of ballast in the US-China relationship besides the business people that Rosie mentioned, whether it’s scholars, whether it’s students, diversifying opinion so that not all the voices on China come out of Washington, because that’s so much one voice at the moment, very hawkish. So I think just to flag that this is a conversation that’s happening in the field.
ANNALEE SAXENIAN: That’s very helpful. We have a question up here. And then the next one over.
RACHEL STERN: Yeah.
MARK DALLAS: The question is, is the technology of fintech in, legal tech, excuse me, is it different than it is in France? In other words, is it being used for authoritarian for monitoring purposes the way, say, social credit was? So how is legal tech different in China– the actual structure of it and what it tells you about users, et cetera?
RACHEL STERN: So I was talking about legal tech as it’s used in the courts. And the French courts have, in keeping with Europe being kind of tech skeptical in general, have just not rolled out the same type of technology as China has. I think that what they’ve done in China is that they can use it to monitor judges. And they want to make sure that judges are doing their work.
It makes sense in a highly centralized court system, you want to make sure that everybody, that has concerns about case overload, you want to make sure that everybody’s doing their work. So on the one hand, you could say that’s about surveillance. And it is. And I’m sure it’s resented by the judges.
But on the other hand, they’re also doing that in Israel and Taiwan. So it’s not exactly about regime type in a simplistic way. I think it’s more about having a highly centralized court system, where the high court wants to know what’s happening in the lower courts.
AUDIENCE MEMBER: Your response had to do with the volume of cases. [INAUDIBLE] What about the nature of the decisions?
RACHEL STERN: I don’t think they have the capacity to monitor that right now. But one of the things that’s been the most interesting application of this, and if I had slides I would have shown you, is you can analyze, one of the ways that the software can be used is you can analyze all those past cases, and the judge can pull down some drop-down factors from a menu.
Like imagine you have a DUI case, Driving Under the Influence case, a drunk driving case. You pull down some factors. What was the blood alcohol level? How many damages were caused? What was the damages caused? And the software will analyze past similar cases based on the factors that you chose and give you a recommended sentence. Give you, the judge, a recommended sentence.
And it doesn’t tie your hands. It’s not telling you what the sentence should be, but anyone, anything. All the research at the intersection of law and psychology suggests that it has a really strong anchoring effect, such that you wouldn’t go outside that sentence.
And then once that technology exists, it becomes possible to identify maverick judges who are making outlier decisions. Now, I have no evidence that any legal system in the world is targeting maverick judges in this way. But what I know is that it’s possible.
ROSELYN HSUEH: One point I have about that is, it struck me as the information collection of the use of technology there. And that is Janus-faced. It could go either way.
ANNALEE SAXENIAN: Totally.
ROSELYN HSUEH: Yeah.
ANNALEE SAXENIAN: [? Asema ?] had a question.
AUDIENCE MEMBER: So I have a question for Rachel and one for Mark. My question for Rachel is that the initiative for this legal tech and courts, is it from the party state or is it court-generated? And then a second– I have a quick comment, which is that a professor Pei, who teaches at Claremont McKenna College, actually has a recent book on the sentinel state.
And he actually talks– the book is really about the low tech kind of aspects of the surveillance of common people, like neighborhood committees, but they also use tech like cameras and streets, and things. But how it is organized, he has a very interesting work. But I was wondering, the source of court, is it the party state or is it the court initiative on its own?
And one question for Mark. Great, very interesting presentation. My question really has to do with is there any product or any global chain where there is a disconnect or dissonance between these control policies to prevent the global value chains to existing and actual functional integration.
So has actual functional integration, is it proceeding the same way, or is there any product or any supply chain where all the protectionist and other control and self-reliant measures have actually affected the material basis of functional integration at the product level or at the supply chain level?
I was wondering because I find a dissonance. So all countries are doing all these great things. But as you said, there are a material limits to whether it can actually happen and whether there is any one supply chain where it is beginning to happen or has happened.
ANNALEE SAXENIAN: OK, why don’t we start with the courts, Rachel?
RACHEL STERN: Sure. That’s a complicated question to unravel. It comes, the initiative to a technology comes from the courts, but it also, is responsive to central priorities, if that makes sense. Like I mentioned, just in passing, that one of the key figures was the head of the Supreme People’s Court. And this is in the 2010s, a man named Zhiqiang.
But he very much framed technology as dovetailing with this national push for AI. So in a sense, it’s the courts fitting in to the broader priorities of the party state. And Xi Jinping is, around that time was, he had a huge anti-corruption campaign. And so the technology is also being framed as a way of combating corruption in the courts.
Maybe this is responsive to your question too. Letting judges know that they’re being monitored. And that it’s not clear what the mechanism is exactly, except it’s sort of a Foucauldian panopticon, is that, if people feel like they’re being watched, they’re less likely to stray, would be the logic.
So it’s kind of a back and forth. And I think in China as in France as in other places, these kinds of technology projects, they need champions. It’s hard work to do this within any state. So there’s champions within the court system, but very much framing their work as responsive to national priorities.
ANNALEE SAXENIAN: Great.
MARK DALLAS: Great. Yeah, good question. So I didn’t want to give the impression that functional integration of global value chains is equivalent to free markets. Those are very different things. And so I would, one answer to your question is, every single global value chain is shaped by various policies that you may have been referring to, whether it’s tariffs or everything else. They mold and sculpt themselves to all of those policies, of course.
And, of course, there’s a large literature about which countries are more favorable or less favorable. So a lead firm is going to decide where to go based on that. But no, there’s no value chain that is not affected by some of the things that you’re referring to. But you were also suggesting– so in a way, it’s all of them.
But you’re also, I think, suggesting is there any product that is completely enmeshed within these regulations. I mean, the classic case is military. So military goods and military companies are very much tied to the national economy. There’s lots of laws and other, depending on the country you’re in, reasons why military production is often purely national and they’re not as integrated in global value chains, not as integrated.
So Raytheon and Boeing, they’ve got military and commercial dimensions to what they do. Airplanes are very complex. They’ve got tens and hundreds, maybe even hundreds of thousands of parts. So you’re not completely walled off.
And so one of the concerns is, what is the defense industrial base at least in the United States? What is the quality of that defense industrial base in terms of producing some of these more less sensitive but critically important, but more conventional products that are more commercially oriented, let’s say?
And then all these companies like Raytheon, RTX is like their commercial arm. So even in the military space, there’s a degree to which. But if you wanted to choose one industry in which it’s much more hived off, let’s say, from functional integration, it would be that one.
ANNALEE SAXENIAN: Question here and then you’re next.
AUDIENCE MEMBER: Venezuela, I’m a policy fellow at the law school. So I have a question related to your views about the role of industrial sectors, particularly more in maintaining the international regimes in amid the more intensified geopolitics.
So in the world of climate policy, so we see technology can be a factor for international collaboration, but they can also become a factor for intensified geopolitics, even though for the globalization of green technology productions, it used to be a factor to get more countries together. Now, it becomes more of a major factor US want to decouple with China.
So then, I’m thinking about OK, state actors seems cannot be helpful, in a way, to figure out a way to keep different parts together. How about the AI sector and also the industrial sectors, are they, in some way, can actually maintain the international like systems, which used to be very helpful, particularly after the World War II because of the UN systems and the international trade? So I was kind of, yeah, see you on that.
MARK DALLAS: Do you want to, who wants to take that?
ROSELYN HSUEH: I mean, just thinking about, you’re mentioning that you work in the climate sector. And to wet that with the question about industrial sector, I think Greentech would be a good example of where, perhaps, we could see international cooperation and collaboration. And if we could get countries to work together around the idea of the Paris Agreement, for example.
And I think Greentech achieves both climate change goals as well as economic goals and growth and development for countries and also sustainability. So, in many ways, it has these multi-dimensionality goals that can mean many different both country level goals, as well as economic business incentives. And so I think that’s how I approach your question.
MARK DALLAS: Yeah, I think, I don’t know if you were referring to particular products or whatever, but new electric vehicles as an example or something like that. I think there’s potential. I mean, clearly, China is at the leading edge of that sector. And so I’m not sure in the United States, but let’s say in Europe, I could see a lot tie-ups happening between Chinese firms, both in battery as well as in the vehicle assembly, given China’s advances in that.
And the China, truly, is exceptional, has done a wonderful job in developing that. It would be great if the whole world could benefit from that for exactly those reasons that you talked about. But whether or not, I think, the United States would be resistant to that, but I think Europe could be open to that, and they would benefit greatly from that. So then, you get a degree of that reintegration over common issues. I mean, obviously, there’s money making there, but the climate is a great offshoot of that development.
ANNALEE SAXENIAN: That’s a great point. We had a question right here. And then you’re next.
AUDIENCE MEMBER: My name is Sergey Donskoy, Professional Competence Institute. That’s my research organization. And I got two questions which tackle my own research I conducted 10 years ago. So the first question is mostly to Roselyn. And the other question is to Rachel and AnnaLee.
First question, it’s been said that China is liberalizing the sectoral part in country But, at the same time, infrastructure is still under control. So what do you think, will government of China liberalize the infrastructure? That’s the first question.
And the second question is about the knowledge transfer. So it’s been said about the chain, United States, Taiwan, and China. And there was a lot of discussions about it, about stealing technologies. So yeah, we’re open to speaking about it.
So what do you think– well, is there any legal basis for regulating this knowledge transfer? That’s my first question. And maybe it would be reasonable to create this treaty between the United States and China, for example, maybe the other country too, in order to regulate this knowledge transfer instead of implying balancing tariffs.
ANNALEE SAXENIAN: Yeah. Do you want to start?
ROSELYN HSUEH: Very interesting. Thank you for those questions. So the way I conceptualize thinking about infrastructure is that you can actually take it to the sector. And so think about infrastructure as a subsector of telecommunications, for example. And then you have services. Even the value-added services, as I mentioned before, that operate on top of that infrastructure.
And so, in that sense, then, that strategic use of foreign direct investment, that reregulation that we see is, there’s even some sectoral variation within one industry like telecommunications. But to your question, is there a possibility that China would ever liberalize that critical infrastructure? Telecoms is very critical. It has security implications and so forth.
I think that simply, and I think, perhaps, this is where authoritarianism might play a role more so than other sectoral questions, is that if party state legitimacy or political stability is important political goals, then letting go of infrastructure may counter to that, to those goals. And so particularly in critical infrastructure, where even countries like European countries and US have regulations around, then I think for China, I think the question would be probably less likely.
And in terms of knowledge transfers versus technology transfers, I think technology transfers could be seen and conceptualized as knowledge transfers. But I do think that software, for example, as knowledge is sometimes less tangible than maybe knowledge in chips. And in micro institutional foundations of capitalism, where I compare China to India and Russia, and for the Russia cases, I look at post-Soviet Russia Federation and patterns of market governance across sectors.
And when I look at software, that was where because it was more intangible. And as a result, even though much of telecoms and ICT was perceived as a military industrial complex sector and was actually never privatized in the way that we the conventional wisdom tells us software was, and it was because it was less tangible in terms of knowledge transfers.
And so there, I mean, I don’t know if this is the answer to your question about legal barriers, but when it’s less tangible, maybe there are opportunities for countries to come together and collaborate on how we could think about frameworks of regulating that.
ANNALEE SAXENIAN: Rachel, do you want to take a whack at that question or?
RACHEL STERN: Yeah, just briefly. So knowledge transfer and legal bases for dealing with knowledge transfer. That was a key point of, like that’s always a key point in Joint Venture contracts. So it is legally, like, there is a legal way to negotiate it between companies that are contemplating tie-ups.
I mean, doing it as a treaty on a national level, I can’t think of any examples. I mean, it’s possible to imagine but it’s also, you’d have to be at a really different world. You kind of connects us to the first question that was asked about, we’d have to be in a world with far more exchange and far more possibilities for cooperation than the one that we currently find ourselves in. So not impossible, but.
ANNALEE SAXENIAN: Yeah, I think this is a challenging conversation because there certainly has been IP theft. I think we know that has happened. On the other hand, the knowledge transfers that I’m talking about, there’s often tacit knowledge that happens. So you cannot start a chip plant without having people go who know how to do it. And we don’t know how to write it all down.
There’s a lot of transfer of knowledge that is just moves with the people. So much to know-how exists. You can’t write down how to run a venture capital fund or how do you– there’s a lot of things that are happening.
You can explicitly license technology, or you can steal it. But I do think that the people flows made this allowed for a lot different kinds of less explicit knowledge get transferred, which are very important as well.
AUDIENCE MEMBER: So we’re talking about stealing people.
ANNALEE SAXENIAN: Well, no, these are–
AUDIENCE MEMBER: [INAUDIBLE]
ANNALEE SAXENIAN: No. I think the problem is these are Taiwanese people who worked in the US and went home. So they’re running a company that, yeah, I mean, so if you want to call it stealing, you can. I don’t think anybody, at the time, would have seen it as stealing. I think that in the industry, it was mutually beneficial.
I think with China we have, there are cases. And I think we need to follow those and try to clamp down on them. But I would not do that at this. I think it’s terrible that we no longer have students going back and forth as much as we did. The student exchanges have been incredibly productive in terms of building common understanding and avoiding war. And having those cut off is very dangerous.
MARK DALLAS: Can I take a crack at that also?
ANNALEE SAXENIAN: Yeah, please do.
MARK DALLAS: Your question is very broad. I think breaking down knowledge is important. On the basic knowledge side, basic research, it’s completely open. And there’s no controls and whatnot. So you’re talking about knowledge that is in some way proprietary, and sensitive, and whatnot. So you’re at one end of that knowledge spectrum, let’s say.
And I think Anna is absolutely that there are, quote-unquote, “controls, regulations in place,” in companies, even for people– non-disclosure, if you leave this company, blah, blah, blah. So there’s plenty that companies are regulating it. Trade secrets are an important part, not just in semiconductors, but even aircraft engines. I mean, something like that. Lots of trade secrets. It’s all based on that. And so, yeah, companies do a lot of that regulating.
I mean governments I think probably would be doing less. But I’ll give you an example for people to people. In export controls, there’s actually a thing called deemed exports. And what a deemed export is, is basically two people. One is a foreign national, one is domestic in a particular country, and they talk to each other. That’s, quote-unquote, “a controlled item.” So government I mean, the US government does intervene at that level to regulate, quote-unquote, “knowledge between people.”
And just so you know that it’s possible. But I think most of it’s done by companies in terms of regulating knowledge. And I think also that the idea it’s not right, it’s not stealing people. It’s really about, you hire them away. There are certain things that they can divulge and they cannot divulge. And there’s a time limit over when that is.
But knowledge is abstract and then specific. It’s the specific stuff that’s associated with a particular company. And their particular processes that should, cannot and should not be transferred, at least in the United States. And you can be sued for doing that. So there are knowledge controls in place.
ANNALEE SAXENIAN: I mean, your argument would lead you to think that all these people have moved between companies in Silicon Valley. They work for one, they take with them. They can’t take the trade secrets. That’s illegal. They will be sued, and they should be. But there’s so much knowledge that they transfer just from their base of experience that you don’t want the government stepping in there. I think we have time for one more question back there.
AUDIENCE MEMBER: Hi, thank you so much for this panel. It’s very fascinating. I think my question is actually very related to the last question, and it’s on digital infrastructure and the materiality of software and how that changes the way we’re thinking about it. So this also relates to a cybersecurity question that you brought up, Mark.
So I guess my question is if we think about you, several people mentioned, like the operating systems and how American corporations are dominating this sphere. If we look at the production of these operating systems, open source software is a huge component of that. One example that I’m thinking of is in 2024, the Microsoft engineer caught this backdoor that was going to be released, and that would have huge implications.
But the larger point here is when we look at the materiality of software and the different network effects of software, that has sort of moved the industry toward this open source model that Silicon Valley is very reliant on, how do we think about that in terms of cybersecurity, and how we should think about digital infrastructure as something materially different from other sources of infrastructure, and like the telecommunications industry, and how are we thinking about this? Yeah.
MARK DALLAS: When you say materiality of it, are you referring to people or?
AUDIENCE MEMBER: I guess I’m referring to the, especially the network effects of software and how the libraries are always dependent on each other and the way it’s produced, yeah.
ANNALEE SAXENIAN: When you say the open source model, you’re just referring to the development process. I mean, Silicon Valley is very split. I mean, there’s a lot of people developing software using the process, the open process that is open source, but they’re also commercially, they’re private companies that make profits. So you have that, I’m not sure what you’re referring to. Is it the process or?
AUDIENCE MEMBER: I guess I’m asking you to hold both of, holding both of these together. Like, yes, there is the proprietary knowledge, but also a lot of the software industry is reliant on open source and have funny way there’s a software joke that goes around about it.
ANNALEE SAXENIAN: Yeah.
AUDIENCE MEMBER: I wish I could pull up the meme.
ANNALEE SAXENIAN: Yes.
AUDIENCE MEMBER: But like Linux–
ANNALEE SAXENIAN: Yeah.
AUDIENCE MEMBER: –which is like–
ANNALEE SAXENIAN: I totally understand. Some people will tell you that 95% of software is run on open source. I don’t know how they figured that out, but, yeah. Go on. So you want the whole software sector. Do you want to?
ROSELYN HSUEH: I mean, I think one way I would approach this is how I answered the previous question, which is, if we think about traditional critical infrastructure as being physical. When it’s physical, in some ways, we have their existing processes and existing frameworks of thinking about how to regulate physical things.
But I think what you’re trying to get at is that when something that is open source or less physical, intangible, less tangible, as I mentioned, in software like that, then can we regulate it? And is it even desirable to regulate it?
And that kind of reminds me of one of the papers that was in one of our panels in the bringing the sector back in workshop that we just had, was that Jeff Ding at George Washington University, he presented a paper about how there’s still interdependence between US and China in the open source software space.
Because as, indeed, as you were saying, that network effect. The people to, and as Armando had mentioned, the people to people flows that’s part of that network effect, makes it. Maybe we don’t have the models yet anyway to try to regulate it. And is it desirable?
ANNALEE SAXENIAN: Well, you can’t patent it. I mean, it can be patented. So software can be.
MARK DALLAS: Do you want me to add something?
ANNALEE SAXENIAN: Yeah, please respond.
MARK DALLAS: So, I mean, it’s a fascinating topic that you hit on. It’s extremely complicated as well. Open source software in particular, I do think it’s a type of infrastructure, depending on how you define what that is. And I think, I mean, it’s a classic example of what I started off with, which is a transfer from a world in which we’re all friendly and there’s development issues to a security world.
And so, yeah, open source software was like this global, everyone contributes, it’s free, you download it, it’s free, you can customize it and then you give back to the community. It was like this global community, wonderful, almost for some anti-capitalist way of approaching things that was going to revolutionize everything.
And it still has that spirit. I think a lot of it. That’s some of that spirit is still there, although a lot of it has been commercialized, privatized, or services have been built on top of it or other ways of monetizing it.
You shift to a security world. And so I used Android. Android is open source. The source code underlying it, AOSP is open source. Everyone contributes to it. Google guides, it’s hard to say, use the word guide. They control it. What are they doing with it? I mean, technically, Google owns Android, but Android is open source.
So it’s this weird combination of private company, money making on the advertising. But with the operating system all over the world. So it feeds into their advertising model, but it’s open source as well. So it’s a strange world, which, I think, is what you were getting at.
But on the one hand, I already mentioned about China’s security dilemma, let’s say, just with something like that. On the other hand, open source is protected by First Amendment rights in the United States. So all the security folks out there are saying, hey, let’s leverage open source and really get the Chinese now, because so much digital infrastructure is built on open source. It’s a First Amendment right.
So it’s a higher bar. I’m not saying that when there’s a security issue, that First Amendment can start to get thrown out the window, too. I’m not saying that. First Amendment is a protection of, foolproof protection against security, but it’s a higher bar. It’s a stronger protection around it.
And so, I think, what you’re hitting on really it’s a topic that’s going to become more and more important. DeepSeek, for instance, was open source. That was the most recent case of a Chinese company that actually is using open source and has open sourced their product. So–
ROSELYN HSUEH: I do think this were regime type could come in, is that, if you are an authoritarian regime and you have the discretion and there’s no First Amendment protections, then maybe there’s not been any intervention, but could that potentially be an issue.
MARK DALLAS: That’s right. And I think one thing I’ll mention– I know we’re already over– is that a lot of this is based on trust. We don’t really talk about trust in the economy. I mean, there is, the sociologists do, sorry. But a lot of this is based on trust.
So when you talk about, will China develop its own open source models, they have to convince the rest of the world to build their products on top of an open source foundation that’s housed in China. And so that might be a hard sell for China. They may be able to create an internal to China open source system, but then you lose a lot of the magic of it.
ANNALEE SAXENIAN: Listen, I want to thank our panelists first, and the audience for a really fascinating conversation. Thank you so much for coming.
MARK DALLAS: Thank you.
RACHEL STERN: Thank you.
[APPLAUSE]
[MUSIC PLAYING]
[WOMAN’S VOICE] Thank you for listening. To learn more about Social Science Matrix, please visit matrix.berkeley.edu.
