Full Transcript

Nathan Labenz: (00:00) Hello, and welcome back to the Cognitive Revolution. Today, we're continuing to build foundational context for understanding China as we enter the AI era. My guest is Angela Zhang, professor at the Gould School of Law at USC and author of Highwire, How China Regulates Big Tech and Governs Its Economy. Having grown up in China, lived in Europe, and become a law school professor in The United States, professor Zhang combines an insider's view of China with a global outlook to create a unique perspective on Chinese technology governance and international technology competition. After first discussing the difficulty of studying China and how much better China collectively understands The United States as compared to how well we in The United States understand China, we unpack the dynamic pyramid model that she developed in her book to explain China's regulatory approach, as well as the prevailing characteristics of hierarchy, volatility, and fragility that it exhibits. We then examine the stark power imbalance between government and business in China, where the state clearly has the upper hand, and compare and contrast the Chinese approach with its European and American counterparts. We also discuss several notable moments in history, including the Chinese government's decision to exclude American tech companies, thus creating the market opportunity in which their own tech sector could develop. The famous 2020 crackdown on big tech that involved canceled IPOs, major destruction of the market cap of Chinese tech firms, and the sidelining of Jack Ma, who had been one of China's leading technologists and global ambassadors up until that time, as well as the more recent apparent thaw between the CCP and leading Chinese tech companies, which some see as evidence of recognition by the government that they need the most vibrant tech sector possible in order to effectively compete with The United States going forward. Along the way, as you'll hear, I ask a bunch of basic questions about the practicalities of Chinese society, including how concentrated power really is and how Internet censorship works at the operational level. On the topic of AI specifically, I was also very interested to get professor Zhang's takes on the light touch that the Chinese government has taken on AI so far, how China understands The US export controls on chips and semiconductor manufacturing equipment, how to understand the DeepSeek moment and whether we should expect another crackdown coming soon, and even whether or not Xi himself might be, as some in the AI safety community have speculated, a doomer. Please note as you listen, I am well aware that a number of the opinions professor Zhang expresses in this conversation are highly controversial. To highlight just a couple, while I'm not sold on their strategic value, I am also not convinced that the DeepSeek moment proves that export controls won't matter going forward. On the contrary, I think there's a very good chance that held in place as they are, they will. And more specifically to the Chinese tech market, I know that Jeff Ding's work on technology diffusion suggests that this is not in fact a great strength, but rather a weakness for China. That said, because my time with professor Zhang was limited and because I'm interested in hearing different perspectives on such important questions, I chose not to push back on each of these topics in real time along the way. I think it's worth taking her self description as a neutral analyst seriously and genuinely trying to consider her analysis from a neutral point of view. But, of course, this won't be the final word, and you can expect multiple strongly contrasting points of view on this feed coming up. For now, if you're finding value in the show, please send it to a friend who's interested in US China relations or the global AI landscape. Reviews on Apple Podcasts or Spotify are always appreciated too, and we invite your feedback via our website, cognitiverevolution.ai, or by DMing me on your favorite social network. Now for a look at how China regulates technology and what that means for the global AI race, I hope you enjoy this conversation with professor Angela Zhang. Professor Angela Zhang from the Gould School of Law at USC and author of High Wire. Welcome to the cognitive revolution.

Angela Zhang: (03:48) My pleasure. Okay. I think it's okay to pronounce it as "Gould" because it's a good school.

Nathan Labenz: (03:54) Yeah. Okay. Good. I like it. So quick preface for this conversation. I don't know a lot about China and I'm trying to catch up quickly because I think that the assumptions that people are making about China and its technology development and what's it gonna do in AI and what's its attitudes towards The United States or the West or whatever are, you know, these assumptions are really fundamental to a lot of the AI discourse that's happening in The United States right now. And I have a sneaking suspicion that a lot of it is not super well founded. So I'm trying to get a better understanding myself and hopefully be able to have better conversations and contribute to a healthier discourse. So you've written this book, High Wire, about how China manages and oversees and regulates and exactly to what degree, you know, those different words properly apply to its technology sector, and I would love to go through it. But maybe for one question I have to just get started, because I don't think people have a good intuition for this, what are your methods for doing this kind of research? Like, what sources do you go to? How do you understand them? You know, what tools do you use? Is AI helpful? I'm interested in that because I think we probably need a lot more people looking closely at China, and I think a lot of people might be interested. Like, I have probably no idea where to start.

Angela Zhang: (05:12) Yeah, that's a good question. It's increasingly more and more difficult for researchers to do field studies in China. But fortunately I was born and raised in China. I grew up in China. I did my undergrad in China and I also practiced a bit in China. And when I was working overseas, I also worked on China related projects. So over the years, I built up all my connections with the Chinese industries. And I also worked for 7 years at the University of Hong Kong. Hong Kong has very close proximity to China. So all these things add up and it's incredibly helpful for me to accumulate firsthand experience understanding what's going on. And I have to say, unfortunately, it is impossible to really truly understand China without being there. And also, you know, to really truly understand things, you don't get it from an interview. It's more like you have to be really absorbed in the environment and have that most experience to really understand what's going on. And also the book rely on a lot of firsthand understanding and my knowledge and conversation with numerous contacts that I have. But I also did a lot of empirical studies, looking at court cases. And so I try everything. I mean, also secondary sources, right? I mean, so I'm just kind of like an AI, you know, in the sense that I do the best I can and gather all the sources and trying to answer the question that I want to understand.

Nathan Labenz: (06:46) Yeah. I'm actually thinking of trying to arrange a trip to China because I've never been and, yeah, I feel like it would just help both my intuition a little bit at least and set hopefully kind of a positive example for like general engagement. Plus, I'm sure it'd be fun and it'd be a lot of great food.

Angela Zhang: (07:01) I will definitely welcome you and then all your podcast listeners to come and see China. Now I'm living in The United States, I think there's a lot of misconceptions about China and a lot of people worry about traveling to China. But I do think that being there in the first place is incredibly important. And, you know, the vast majority of people have fun, really incredible experience after they traveled there. And it's not as scary as the media has portrayed it is. And then just like another country. And in fact, they will be surprised by the technology when they arrive in China.

Nathan Labenz: (07:43) Okay. Several directions I wanna go and follow-up there. How helpful, I mean, you don't need them, but do you have a sense for sort of how helpful translation tools are in today's world in various contexts? Like, if I wanna go read content off the Chinese Internet, you know, and I translate it with DeepSeek or whatever, like, how much does that help me actually be able to read in a way that would foster real understanding versus sort of being too superficial? And I guess maybe same question for, like, street conversation. You know? Can I take a just a real time translator at this point and have meaningful conversations? What would you expect in that regard?

Angela Zhang: (08:22) I am not a big expert in that, but I would just say that China is a leader in voice recognition technology, and that kind of technology has been built in a lot of the softwares and is freely available. You know, when I was living in Hong Kong, my domestic helper only speaks English and my mother-in-law speaks Chinese and they communicate flawlessly with each other just using their cell phone. So I don't think that's a problem at all. Yeah.

Nathan Labenz: (08:52) Yeah. Cool. What would you say are some of the technology surprises that the average American would be kind of, you know, most taken aback by when they visit China?

Angela Zhang: (09:04) I will say first is high speed train. And it's appalling that United States, correct me if I'm wrong, still doesn't have high speed train. I think California has been trying to build its high speed train for a long time and now has finished one third of the project and who knows when it will complete.

Nathan Labenz: (09:26) Resold you one third as I understood it.

Angela Zhang: (09:29) Look. I mean, yeah, when you think about that, I mean, wow, I don't think China is 10 feet tall, but in terms of the execution capability, you just look at the capacity to build high speed train. I mean, this kind of project between, you know, high speed route between San Francisco and Los Angeles, it will take China one or two years to complete. And it's taking the state ages. And we still don't know when it will be completed. And there were numerous obstacles along the way, right? I mean, so I was talking to my student the other day, look, I mean, when we talk about US China tech rivalry, I mean, this is just one example that shows you where China is good at, that China speed, the China execution and efficiency in doing things, right? I mean, so I was just appalled at the US inefficiencies when it comes to a lot of the high-tech technology.

Nathan Labenz: (10:26) You mentioned that there's a lot of misconceptions. I wonder maybe the super highest level, one that it sounds like you're already alluding to is that, you know, China is the sort of, you know, invincible monolithic, you know, super powerful entity when you say they're not 10 feet tall. What other sort of misconceptions do you think are most prevalent or problematic in, say, The US with respect to China? And then I'm also interested in kind of the reverse of that. Like, does China understand The United States better than The United States understands China? And, you know, if so, is that just, like, because of our sort of general cultural exporting history? And, you know, what big misconceptions do you think, you know, China collectively has about The United States?

Angela Zhang: (11:09) Right. I mean, this is an excellent question. Look, I don't want to paint a rosy picture of China. Okay? The Chinese economy is at its rock bottom at the moment. Young people have very pessimistic outlook of the future. Probably this is kind of like the lowest sentiments for decades. And my book for the most part is very critical of the China's governance model. But that said, right, I mean, let me be fair, because I was born and grew up there. There are certain advantages of an authoritarian regime being very efficient, and they can get things done very quickly. And they put more emphasis on the interests of the majority. And so even for a lot of controversies that you have seen in Western jurisdictions about China, the vast majority of Chinese people actually likes it. I mean, give an example of the surveillance technology, right? I mean, in the West, when we talk about Chinese surveillance using all this camera and everywhere, facial recognition, it's all incredibly negative. But if you go to China, I mean, you ask normal Chinese grandma, grandpa, do you like that? They don't care. And in fact, some people will tell you that they feel safer because they have camera everywhere and they don't need to worry about being robbed like when the crime rate is extremely low. So you know, I think the Chinese Communist Party knows how to get endorsement from the majority of the people. I mean, that's why the popularity rating was pretty high, you know, until the mishandling of the COVID-19, okay. Now, so I think China has its clear strengths and weaknesses. We can delve more into its weaknesses as we talk about the book, but what about The United States? The other side around, I mean, does China have some misconceptions about The US? I do think they do, I guess one of the biggest misconceptions is to think The United States is a very democratic government, which this new administration clearly show it is not, okay? But for the most part, I think China understands The US much better than The US understands China. And that's understandable because the Chinese government is very opaque. The bureaucratic regime is extremely, you know, efficient, but at the same time, they deliberately keep things very ambiguous and very opaque. And also because of the internet censorship and information control, it's much more difficult to understand what's really going on inside that China bubble. So I would say that, you know, every day when I read the news in China, I mean, I got to know almost everything going on in The US, but mostly you can tell the government always tells the bad news about other countries. So, you know, depending where I live, the information I get is always biased one way or the other.

Nathan Labenz: (14:15) Yeah. I was gonna ask just briefly before turning to the framework of the book. Yeah. When you said, like, the approval was high up until COVID-19 mishandling...

Angela Zhang: (14:24) Yeah.

Nathan Labenz: (14:25) Is that, like, a felt sense or are there, like, data sources that you trust and are there data sources that have actually, like, shown a drop in satisfaction? I'm not aware of that, but it's really interesting that's out there.

Angela Zhang: (14:35) Yeah. Yeah. There is. I mean, I think, however, there's a bunch of scholars who track the public opinions rating of the Chinese government, and they show, you know, the communist party, they gained the legitimacy through a lot of times through performance, I mean, they were not democratically elected. So the government needs to ensure economic prosperity, social stability for the vast majority of their people, right? Otherwise, there will be a lot of discontent. And that's distinguished China from a lot of other authoritarian countries that tend to be extremely repressive and things like that. But by the way, you know, during COVID, when Xi Jinping was able to, you know, get things under control, particularly between 2020 to late 2020, during that period, I think his approval rating was extremely high. If you talk to the ordinary Chinese people, they feel great about that country because when they look around the world, everybody was living in chaos and China was able to achieve COVID zero and was the first one to emerge from the COVID chaos and the first one to resume production. And there was this talk about, you know, the rise of the East and the decline of the West at that time, right? But, you know, things quickly turned sour after Omicron emerged when the rest of the world started to open up when China is still under severe lockdown. Yeah.

Nathan Labenz: (16:06) Hey. We'll continue our interview in a moment after our words from our sponsors. In business, they say you can have better, cheaper, or faster, but you only get to pick 2. But what if you could have all 3 at the same time? That's exactly what Cohere, Thomson Reuters, and Specialized Bikes have since they upgraded to the next generation of the cloud, Oracle Cloud Infrastructure. OCI is the blazing fast platform for your infrastructure, database, application development, and AI needs, where you can run any workload in a high availability, consistently high performance environment, and spend less than you would with other clouds. How is it faster? OCI's block storage gives you more operations per second. Cheaper? OCI costs up to 50% less for compute, 70% less for storage, and 80% less for networking. And better? In test after test, OCI customers report lower latency and higher bandwidth versus other clouds. This is the cloud built for AI and all of your biggest workloads. Right now, with zero commitment, try OCI for free. Head to oracle.com/cognitive. That's oracle.com/cognitive.

Nathan Labenz: (17:21) Yeah. That was a really interesting moment. Maybe we'll come back to that in a little bit. Let's go to the book. Yeah. So the core framework of the book and, the subject is how does the Chinese government oversee, manage, regulate its technology industry? So you introduced this framework, the dynamic pyramid model, which you describe it, you know, a couple different dimensions. One is just sort of the traits that it has, hierarchy, volatility, and fragility. And then there's also the different actors that play in that system. Let's start with the overview of the hierarchy. We've got the top leaders, the regulators, the firms, and the platform participants, which I take to be, like, individuals and small businesses and anybody that's sort of doing stuff on these Internet platforms. It strikes me that we have basically the same model here in some important sense, but maybe, you know, that provides a good jump off point to compare and contrast, like, how the different actors are sort of positioned differently or have different levels of power or maybe have different interests in China versus what we're familiar with.

Angela Zhang: (18:23) Yeah. Great question. Right. I mean, the purpose of having this model is to help our readers to understand how China governs more generally. And in fact, this model is not only helpful for us to predict, you know, how to regulate consumer tech business, but almost every major policy that you see coming out from the Chinese government. So I also applied this framework to analyze how China had, you know, regulated COVID-19 and property crisis and etcetera. But the book focused almost mostly on the tech sector because it gave us a very good case study, you know, a massive case study to analyze this model in-depth. And so when we talk about models, the first thing we need to look at are the players, right? I mean, so the hierarchy describes the structure of the decision making in China's regulatory governance system. And so there are 4 major actors, there's top leadership, who are mostly referring to President Xi and the other members of the Politburo. And then we have the bureaucrats and regulators who are in charge of the day to day implementation of the top's initiatives. And these are the people who really did the hard work in turning out regulation and conducting enforcement. And then we have the firms, right? I mean, and that's the main subject of the regulation. And at the bottom are the platform participants that interact with the firms. Now, we obviously have the same 4 major players in other democratic settings like The United States and the EU. However, the way that they're organized are entirely different, right? I mean, the major distinction that I will see is the power imbalance. Now in the China, if we see it on the spectrum, China will be at one extreme where the government clearly has upper hand over businesses, right? I mean, so when we talk about the pyramid, at the apex of the pyramid in China are the top leadership. And then it comes to regulators because regulators are subordinate to top leadership. They're not independent, right? I mean, they're held accountable to the top, not necessarily to the bottoms. They're not so held accountable to the public because in China, that's the first and foremost important for them. And then...

Nathan Labenz: (20:46) That's the interjection. The top leadership, I just had to search, for how many members are on the Politburo, and I got 24 back as my answer from Google. So are we literally talking at top leadership, like, not like a general elite class or even sort of a, like, 500 member, you know, like the US congress sort of scale? We're literally talking, like, just a couple dozen...

Angela Zhang: (21:08) Or you can just think about it just one person. There's President Xi. Yeah. Because he basically controls the top leadership.

Nathan Labenz: (21:15) And then the regulators basically report directly to that high level and sort of you feel like their direction is coming, like, through very few links in a chain of command from the very top?

Angela Zhang: (21:26) Yes. Very, very loose chains of commands. That's a good way to put it. Right? I mean, the regulators are the governments. Right? I mean, the ministries, the departments, and the local governments. I mean, there's a huge bureaucracy spanning across 31 provincial entities and also numerous central ministries and department agencies. So all these bureaucrats, they have the expertise and the knowledge to implement the law. However, when they do so, they need to closely align their work with the top's initiatives. So they always look upstairs to see, you know, what are the shifting policy winds or what are the policy signals, right? So I mean, because in China, you know, that signal is incredibly important. In some ways, you know, that signal is policy, you know, and the policy is signal. Right? I mean, so in China, when the bureaucrats carry out the work, I think the first and foremost is to understand the policy signal because the policy wind can shift all the time. And that's incredibly important.

Nathan Labenz: (22:40) Interjection. When things happen in The United States, like, it's often the case that, this is not what the president wanted. Right? We've got, you know, sort of this is up now for contest, of course, but we've got, like, independent agencies, you know, that sort of are historically at least understood and meant to be insulated from, you know, the executive's, like, current mood. We've also got, like, state level regulators. In general, it seems like you just cannot assume that there's, like, a presidential or anybody. You know, you can only go so high before you sort of find, like, this is the executive, you know, decision maker in this particular domain. It sounds like the Chinese model is a lot of, like, pull in the sense that there are, like, local executives who have a very clear understanding that, you know, they don't have an independent power base. They're not, like, elected or appointed by some other, you know, group that they're more beholden to. It's kind of all at the discretion or the pleasure of the top leadership, and their goals are to understand what that top leadership would want them to do. It sounds like they're not necessarily always told super explicitly, but they're, like, trying to map these sort of high level notions onto their domains in a way that they, you know, hope will then be ultimately pleasing to the central leadership. Is that kinda how I should understand the...

Angela Zhang: (24:03) Yeah.

Nathan Labenz: (24:03) Trying to please who? What signal?

Angela Zhang: (24:06) Yes. Yes. So, basically, everybody answers to one boss upstairs. Right? I mean, well, depending on what level they are, they may have other bosses, but they know who is the big boss. Right? I mean, so and that's why in China, Xi Jinping is also called, it means, you know, he's big. I mean, he's a real big boss in the country. But even at his level, he may not have the time, expertise, and capacity to handle everything, right? I mean, so he has to delegate it to the government officials, the bureaucrats to implement them. So the bureaucrats themselves, they do have discretion, but they do at the same time need to closely align what they do with the top's initiatives or frame it in a way that is aligned with the top's initiatives, it's also work for their own interests, right? Because we also need to take into account agencies' own bureaucratic incentives when they carry out enforcement mandates. Now, so go back to what we were talking about, we are talking about power imbalance, right? I mean, so the key actor here are the other businesses, right? I mean, the businesses in China, they are, I put them at the lower level of the hierarchy than the agencies, because unlike other jurisdictions where you see businesses go to war with the agencies, in China, they don't do that very, very often. You see a company whether it's a public company or private company or foreign multinational company or Chinese domestic company, they almost never ever directly confront the government. I mean, think about Elon Musk, you know, when he went to China, he behaved like a sheep, right? I mean, he's trying to please the Chinese government, right? I mean, unlike in The US, he's sending a poop emoji to the government agency. Can you imagine he ever dare to do that in China? I mean, he understands. In China, you play by the Chinese rules and show respect to the government agency. So businesses understand that the government clearly has upper hand, is the most powerful actor here. And then at the bottom are the people, right? Are the public. These are the, think about the online merchants, delivery workers, the consumers, right, that interact day to day with the platform. And these people, you know, just like those in the West, they're very weak in bargaining with the platform, a very powerful platform. And oftentimes there's a lot of complaints, but the voices tend to be muted in the sense that in China, because of the censorship information control, that makes it very difficult for them to voice, have their voices out. That said, you know, the Chinese government is very strategic. If they want to have the complaints heard, they could amplify those complaints. So at that point, you do hear a lot of public opinions, because those public opinion will help the agencies to get more legitimacy of the actions. So Chinese government is extremely media savvy. They do know how to games and gain more public approval and support. Now, if you, I will play Europe at the center in a sense that in Europe, government is still powerful, but businesses do have the opportunity to challenge and appeal the cases. Just think about in the EU where I also used to live. I used to work in Brussels and also teach in London. And so in that jurisdiction, US tech businesses, investigated, almost all of them have been investigated by the European Commission. They almost always challenged the Commission in Luxembourg with the general court or the court of justice. Courts have been very deferential in Europe and partly also because of their legal culture. But occasionally businesses do win, right? I mean, and they continue to challenge and continue to fight their cases. And from the standpoint of European officials, they also don't think this is abnormal. They won't hold grudges because they fight, right? I mean, because they respect this is a rule of law system. Now moving to The United States, I will see it as another extreme, polar opposite of China, where businesses seem to have the upper hand. Just think about in the past 5 years, despite, you know, a cascade of congressional hearings and all these interviews and testimony with all these big tech CEOs like Mark Zuckerberg or, you know, all of them. Has anything happened? Nothing. The Congress hasn't passed any single law to discipline the big tech firms. The only law that it adopted is the law against TikTok to ban and divest TikTok. The United States so far doesn't have any data privacy law at the federal level. The only law that restricts data flows targeted at TikTok, the Biden administration signed executive order to restrict about transfer of sensitive data to adversarial country, including China. So the only tech regulation that we have in The United States are almost exclusively targeting at companies with the Chinese identity. With respect to its own big tech giants, The US has done almost nothing, right? I mean, and you can say, is that capture? Is it a scandal, right? I mean, when you see all this big tech CEO surrounding the president on the inauguration day, rubbing shoulders with the president right now. I mean, Trump is surrounded by all this tech moguls these days and even assisted by Elon Musk, right? I mean, you worry about another opposite to, like, the crony capitalism in The United States.

Nathan Labenz: (30:11) Hey. We'll continue our interview in a moment after a word from our sponsors. Being an entrepreneur, I can say from personal experience, can be an intimidating and at times lonely experience. There are so many jobs to be done and often nobody to turn to when things go wrong. That's just one of many reasons that founders absolutely must choose their technology platforms carefully. Pick the right one, and the technology can play important roles for you. Pick the wrong one, and you might find yourself fighting fires alone. In the ecommerce space, of course, there's never been a better platform than Shopify. Shopify is the commerce platform behind millions of businesses around the world and 10% of all ecommerce in The United States, from household names like Mattel and Gymshark to brands just getting started. With hundreds of ready to use templates, Shopify helps you build a beautiful online store to match your brand's style, just as if you had your own design studio. With helpful AI tools that write product descriptions, page headlines, and even enhance your product photography, it's like you have your own content team. And with the ability to easily create email and social media campaigns, you can reach your customers wherever they're scrolling or strolling, just as if you had a full marketing department behind you. Best yet, Shopify is your commerce expert with world class expertise in everything from managing inventory, international shipping, processing returns, and beyond. If you're ready to sell, you're ready for Shopify. Turn your big business idea into cha-ching with Shopify on your side. Sign up for your $1 per month trial and start selling today at shopify.com/cognitive. Visit shopify.com/cognitive. Once more, that's shopify.com/cognitive.

Nathan Labenz: (32:13) It is an interesting time for business. Tariff and trade policies are dynamic, supply chains squeezed, and cash flow tighter than ever. If your business can't adapt in real time, you are in a world of hurt. You need total visibility from global shipments to tariff impacts to real time cash flow, and that's NetSuite by Oracle, your AI powered business management suite trusted by over 42,000 businesses. NetSuite is the number 1 cloud ERP for many reasons. It brings accounting, financial management, inventory, and HR all together into one suite. That gives you one source of truth, giving you visibility and the control you need to make quick decisions. And with real time forecasting, you're peering into the future with actionable data. Plus with AI embedded throughout, you can automate a lot of those everyday tasks, letting your teams stay strategic. NetSuite helps you know what's stuck, what it's costing you, and how to pivot fast. Because in the AI era, there is nothing more important than speed of execution. It's one system, giving you full control and the ability to tame the chaos. That is NetSuite by Oracle. If your revenues are at least in the 7 figures, download the free ebook, Navigating Global Trade, 3 Insights for Leaders at netsuite.com/cognitive. That's netsuite.com/cognitive.

Nathan Labenz: (33:38) Going back for one second to the, just you mentioned that, like, the government knows how to sort of turn up complaints, you know, selectively when it wants to. Can you describe a little bit more how the sort of staffing or, like, actual management of the censorship apparatus works? Because I would imagine, like...

Angela Zhang: (34:04) Mhmm.

Nathan Labenz: (34:05) And that when you were first saying, like, you know, the complaints about the tech platforms themselves are, like, by default suppressed. That made sense to me because I was like, well, there's already this, like, censorship regime and they're implementing, like, whatever censorship guidelines they're getting. So maybe they'll just, you know, salt in a little of their own brand protection as they do it. But then when you said, well, they could actually tinker with that in a more granular way, now I was, you know, kind of wondering what the shape or, like, you know, who has access, who staffs what? I don't know if that is even known or clear, but...

Angela Zhang: (34:35) Well, this is a very opaque system, right? I mean, but obviously both the big tech firms and the Chinese government invested hugely on content moderation, right? Although, you know, how that, you know, has a machine operates in practice remains a mystery to almost everyone except a few insiders, right? I mean, but you know, think about it in the West, you know, companies like Facebook, Meta, they also employ an army of people to do content moderation, although they're getting rid of them these days. But in China, this is a very important part of the compliance of all the big tech platform who do social media businesses, because one of the first and foremost priority of the government's regulation is to ensure that your content is politically aligned with the government's or you can have the socialist values of the Chinese Communist Party, right? So the platforms, they will have to receive instruction from the Cyberspace Administration of China about what kind of content is allowed or not allowed. And they will receive very timely updates given the changing circumstances. But I can give you a very concrete example of how it works in practice. So in the book, I talk about China's largest ride hailing platform called Didi, which is the Uber in China. And when they wanted to go public in The US, they actually met with some oppositions from within the government and one of the agencies, the Cyberspace Administration of China, we just talked about, they were asking Didi to postpone its IPO filing to ensure that they comply with the data security regulations, because it's a huge ride hailing firm and gathered a lot of sensitive geographical data and sensitive personal data from a vast Chinese population, right? I mean, so from the standpoint of CAC, that the online listing process, including the submission of information to The US regulators can potentially pose national security concern. However, Didi pressed ahead with its IPO filing in The US disregarding the CAC's warning. And obviously, the CAC was very upset. And 2 days after its listing in The United States, the CAC conducted a cybersecurity check review over the firm. And that caused the firm's tumble more than 10%. I mean, and wiped out a lot of money from its capital market capitalization. That was really, you know, a point that really made international investors really upset that China is uninvestable. This kind of regulation is totally unpredictable and arbitrary. Now, when the CAC conducted that investigation, it was doing that on the pretext of they wanted to conduct a cybersecurity check on the firm. But that kind of check is often done according to the law is about procurement practices, whether the firm is buying some equipment that make it insecure and that proposed national security threat. It's never been used to a setting where the firm is going online listing, right? I mean, they're just submitting documents to the SEC. And so it's a very dubious act from the government. And however, in order to enhance the legitimacy, you started to hear a lot of public opinions in the Chinese media saying that Didi is a traitor of turning over its sensitive data to the US government. And Didi's founder, the president of Didi is called Liu Qing, which is an ex-Goldman banker, very successful businesswoman, is also labeled as a trader who divulged Chinese secrets and sold it to The US. So you see the Chinese government drum up this kind of nationalistic sentiment in the country. And there's also posts circulating showing that Didi actually collected a lot of sensitive government data because it can show, it can tell at certain time what the traffic is like going to different Chinese ministries and which ministry work at what time and which ministry is busy at what time. So, I mean, and when this media posts were circulated, it helped form the impression among the Chinese public. Look, I mean, this company does, you know, have the ability to reveal information that can pose national security threat. And so the government is a good thing to intervene and to conduct cybersecurity check. I mean, of course, general public have no idea what cybersecurity review means and whether there's a clear legitimate legal basis, right? You can tell this is one example. Another example is when Jack Ma's IPO was suspended, and then there's a flood of media posts on Chinese internet accusing him of, you know, extremely greedy businessman who are trying to start up this huge fintech company that rips off the Chinese people. And then there's a lot of reports about the dishonest practices of Ant Group, which is the fintech company that was supposed to go on IPO and tried to dupe consumers, right? I mean, the whole Chinese media machinery is also part of the Chinese bureaucracy, right? I mean, so these different departments, propaganda department, the media and the regulatory authority, they all work together in tandem to jointly increase the legitimacy of the actions.

Nathan Labenz: (40:38) Okay. That is quite interesting. And again, a lot of different directions we could go. Could we maybe do just a minute on sort of the rise of China's tech giants? Because I don't know a lot about this, and one question I'm particularly interested in is to what degree were these, like, the winners from a sort of, you know, relatively free and open market competition process versus at what stage were they sort of, you know, favorites of the party or somehow otherwise, like, designated winners?

Angela Zhang: (41:12) Well, in the book I did talk about how the Chinese government censorship and the Great Firewall kind of work as trade barriers for foreign, particularly US social media businesses to come and operate in China. I mean, so companies like Baidu, which operate China's largest search engines face way less competition these days because the only other US operated search engine is Bing, which now has only 1% of the market share, right? I mean, Google exited years earlier, right? Now let's consider the counterfactual that you posed that had China not imposed this kind of restraints on the Chinese internet sector. Do you think that China still have these domestic national champions? I think it is true. You know? I still think that the domestic firms will still win. Just think about what happened nowadays. Right? I mean, why TikTok is conquering all over the world? And why the majority of the top 10 most downloaded apps in The US app stores are Chinese firms, right? And the AI firms, you know, also leading and doing well. It's, I think for the most part is China's capability of adoption and diffusion is much better than The US peers. So China is not good from 0 to 1. That's why, you know, you see the major breakthroughs in technology, it almost never occur in China. Okay. And I, by the way, I don't see DeepSeek as a technology breakthrough in a way, because discovering the transformer model or, you know, like Geoffrey Hinton, figure out how to use the neural network and model the AI architecture of a neural network. That's what I see as a true scientific breakthrough, okay? When we talk about the Joseph Schumpeter's innovation cycle, we have the innovation and then we have adoption and diffusion. China is more at the second and the third stage, in adoption, and particularly good, extremely good at diffusion. And DeepSeek is just very good at diffusion because when we talk about diffusion, it's just about how we make things more efficient, how we lower the cost, how we make it more accessible to consumers, right? I mean, in terms of that, so that's why I'm saying, so from 1 to 10, China can do it really well, right? I mean, so all the consumer tech businesses that I'm talking about, I have no doubt that, you know, from a few years from now, we will see, unless The US bans all the China's apps, which is difficult, then you will see more and more Chinese firms flooding into The US market. You already see Temu, like I mean, just order from Temu the other day. I mean, prices are 30% cheaper than from Amazon if you can wait. And Temu can deliver very quickly these days too. They also have local warehouses, right? And Shein, right? The fast fashion. And it's coming into US market, not to mention DeepSeek and things like that. Yeah.

Nathan Labenz: (44:24) So that's really helpful context. It sounds like these companies, the tech giant names that we know, your Tencents and Alibabas and so on. These companies started as, like, organic and like, they succeeded on their market merits. Yes. There was this sort of exclusion of US companies, which, you know, created sort of a less competitive environment than they would have otherwise faced. But in terms of, like, why it was Alibaba or why it was Tencent versus a different Chinese company that may have been competing for that niche, these leaders like they won these markets kind of fair and square, so to speak. Is that...

Angela Zhang: (45:00) Yes. Yes. Yes. I mean, it's like extremely competitive market. Right? I mean, so before everybody knows DeepSeek, okay? DeepSeek has already been known in the country for more than a year and I was just so surprised that Silicon Valley just have no interest in China. They totally underestimated it. So DeepSeek around mid 2024 had engaged in a wage price war in the AI community, they were the one who slashed significantly the prices. And that started this predatory pricing in a way scheme, and then everybody had to follow, including the leading AI leaders, right? I mean, so that's why in China, DeepSeek was called the Pinduoduo of AI. Pinduoduo is the parent company of Temu, which is very good at, you know, huge discounts, deep discount sales. And so I was just so surprised that the Western community doesn't know this firm's existence until like 8 or 9 months later, right? All of a sudden they just woke up. I mean, this firm is doing this. Just to show that in some ways, misconception that we were talking about, right? I mean, the misconception in The United States is going like kind of also polar opposite. One school of thought thinks China is 10 feet tall, China is ahead of us, China is going to beat us and we have to control them. And the other is I think China is like 3 feet tall, totally underestimate China's innovation capability, right? I mean, so I'm just trying to be neutral. Like, I mean, present what China is.

Nathan Labenz: (46:41) So can you tell, you know, we haven't officially touched on volatility and fragility, but maybe you can sort of illustrate those with a little bit of a history from this sort of 2020 moment where you alluded to the Jack Ma sort of, you know, putting him in his place and then kind of through to the present. Because my, I know that story in the sort of roughest terms of, like, you know, he said a couple critical things. All of a sudden, he wasn't heard from very much for a while. IPO was canceled. You know, vast amounts of, like, wealth seemingly taken off the books. I mean, my sense is that the company still kind of, you know, carried on without, like, so much disruption to their operations, but it was just kind of clear that, like, there's a limit to your upside here so that, you know, can affect the stock price even if, like, today's operations are still running pretty smoothly. That's kind of my sense for sort of what happened, I guess. But why did that happen? Like, what were they really trying to accomplish? Did they accomplish their goals, or did it, like, you know, turn out differently than the decision makers had hoped? And now I have a sense, again, a little bit pretty vague from just sort of reading the news that, like, you know, even, like, very recently, like, since maybe the DeepSeek moment, there's sort of a sense that, like, okay. We maybe need to, like, bring our technology companies back a little bit, but, you know, exactly what does that mean? Like, what are we bringing them back from and, you know, sort of, again, like, what are they trying to accomplish? You know, if it's volatility, like, what's the change in direction that all this sort of amounts to?

Angela Zhang: (48:11) Right. I mean, so by volatility, I mean, is that the regulatory pendulum swing is very, very dramatic in the sense that the Chinese government either is very lax, whether it's tax regulation, or becomes extremely intense and very strict with the tech regulation. And for the most part, from early 2010 until late 2020 until Ant's IPO debacle, I will say that the Chinese consumer tech businesses enjoyed like a decade's honeymoon period with the Chinese government, where the government is like taking a very light touch approach with these firms. And so there was a breather like until, you know, from that IPO until early 2022, there was like an 18 months, very intense crackdown. And then after that, we revert back to the normal. I mean, so if you think about how China regulates AI these days, I also think it's really, really lax at the moment. So exactly what happened is very, very complicated. And obviously for the first decade's honeymoon period, because the government is not doing much, it's actually doing very little. And that also created a lot of the problems because had these problems been addressed earlier, they would not have become what it is today. It actually led to a very concentrated tech sector, extremely concentrated tech sector. We are mostly talking about two major players, right? Alibaba, Tencent, we have duopolies in China. Whereas in The US, you already have the 7 major tech firms. At least is still a bit more less concentrated in a sense, right? I mean, so that concentration problem led to a lot of issues, right? And a lot of regulatory problems. And from that respect, it's not so different from what you witnessed and experienced in a democratic settings. But I think what is really the last straw over the camel's back was the IPO of Ant Group, which is China's largest fintech company and actually also the world's largest fintech company. And it was going on the largest IPO in history in late 2020, and that IPO was suspended just at the 11th hour. And that's why it really stunned investors. That's the kind of volatility I was talking about. Now, why did they call off the IPO at the last minute was because, obviously, Jack Ma's speech was very helpful, but fundamental, helpful to contribute to the debacle in a way. But I think fundamentally, it's the regulatory tensions between China's financial regulator and Ant Group. And I think that, you know, despite the fact that you think about Alibaba's business empire mostly started from e-commerce business, but when he spun off his Alipay, which is China's payment app to start up Ant Group, the fintech business, he starts to cross the line, he starts to venture into a financial industry. And that touches sensitive nerve of the Chinese Communist Party, right? I mean, who is very obsessed with control, right? I mean, when you're just selling clothes, selling cabbages, selling toys, you know, it doesn't pose any social stability or political threat to political stability, right? But as soon as you start to venture into finance, and as soon as you grow to such an enormous scale, okay, that will put Chinese top leadership on high alert because the last thing they want is one single private entrepreneur. Just imagine Jack Ma, which used to control over 50% of the shares in Ant Group, one single entrepreneur amassed so much fortune in the largest financial institution in China, right? And so not to mention, like, I mean, think about it even from a legitimate regulatory standpoint, what if this firm collapsed? Right? I mean, and its ownership structure is also very opaque because you have a dominant controlling shareholder there. So we probably won't have time to go into all these details. I do have a lot of sympathy with the Chinese government in why they want to intervene in the IPO of these firms because to be fair, Ant Group, they do all sorts of financial services. However, they argue that they're not a financial institution and they changed the name from Ant Financial to Ant Group to avoid being identified as a financial firm. They don't even want to be seen as a fintech business. In their IPO prospectus, they call themselves a tech-fin business. So they're a tech firm. And eventually they got a very high valuation of the IPO, $320 billion US dollars, which is higher than JPMorgan at that time. It was like, that was something that really stunned the Chinese regulators. Like what? This firm was like spun off from Alibaba less than 6 years ago, and now grew up to be the largest financial institution in the world? That looks like a huge bubble in the making, right? I mean, we need to intervene before it is too late because this firm now poses a systemic risk to the Chinese financial industry. And so I think this is the major reason why, you know, the government realized these firms, they are not just doing ride hailing or selling clothes or video games. They are venturing into the most important part of the Chinese economy. And so we have to reduce the influence. We have to make sure that they listen to us. Right? I mean, we have to remind them who's the big boss. I think I mean, to just to...

Nathan Labenz: (54:04) put it very, should we understand that as Xi personally? Like, when something like this happens, should I be thinking, you know, Xi gathered the cabinet and, like, got reports from different directions and ultimately made, like, a personal decision, like, this is not going forward. I say so. Yeah. Ma goes to, you know, home arrest. That's coming from what?

Angela Zhang: (54:27) He has a home arrest. No. I don't think it's a home arrest. And by the way, Xi met with Jack Ma the other day. I mean, which is, we, it all went round circle now. Like, I mean...

Nathan Labenz: (54:36) So what's the new, what's the next turn then? Like, they, you know, put him in his place for a while. I hear the argument that, like, okay. Hey. This is now a national public good, and, you know, it's just not our style to let random entrepreneurs control such things. So, you know, you're no exception. You know, are we swinging back because for, like, competitive reasons, or what do you, how do you sort of, you know, account for the most recent apparent vibe shift?

Angela Zhang: (55:09) I don't think it's swing back in the sense that government is encouraging this kind of financial services. No. Right? I mean, I don't, I think it remains clear from Xi's standpoint, what are the priority industries that can really help China become strong and prosper. And you can see from the entrepreneurs he invited to that symposium held a couple of weeks ago, are on Xi's table, right? I mean, so these are the entrepreneurs who work on AI. I mean, even DeepSeek's founder, who nobody knows over a year ago, now is on the table. And there are a couple of consumer tech businesses, but for the most part, it's firms in AI, in EV, clean energy, robotics, and one of the most important guys is the boss of Huawei, right? I mean, so these are the people that the Chinese government relies on. These are the most important prioritized strategic industries. The government, yes, they understand. We should let you continue to sell video games and do all this consumer type businesses as usual because we still need employment. We desperately need employment. Our young people need jobs. Right? I mean, the unemployment rate is too high. But that's not going to have the transformative impact on the Chinese economy. Right? I mean, the Chinese went through a very painful structural transformation in the past few years under Xi's leadership. And I think Xi's very ambitious in trying to transform China from the old and dirty economy into a new economy. And you have to say, when you look back at what China is trying to do, they achieved mostly what they want to do, right? Even the highly controversial China 2025 project. And I think Bloomberg ran a report the other day and see China achieved most of its goals and China now becomes a leader in most of these tech, all these technologies China tried to set the goal earlier, right? I mean, China went through really painful property reforms because China doesn't want its economy to build on the property bubble. And now, you know, you can see that the economy has switched to a high manufacturing and clean energy exports, right? I mean, so we'll have to wait and see what the trajectory of the Chinese economy would look like in a few years. But you do see that government's emphasis on tech, and particularly hard tech, when talking about hardcore technology that can give China competitive edge, is clearly the main focus of the Chinese government. And also thanks to, you know, the pressure from The US, right? I mean, that China views, the urge that I have to be independent, I have to be self sufficient.

Nathan Labenz: (58:15) So how about AI? We're in the, you know, maybe the early innings of the great AI competition, which is not something I'm super enthused about, to be honest, because I feel like the technology is unwieldy enough without making it, you know, putting it right at the center of a geopolitical struggle. But, nevertheless, you know, that's the way the winds are certainly blowing for the moment whether I like it or not. I guess, you know, one of the things that has sort of surprised me, you mentioned that the sort of AI is still in its honeymoon phase. I've heard a little bit about this before where it's like, you know, there are sort of some standards, some expectations, but it's not like the government is, like, doing prelaunch reviews of all these products. My sense is they're more kind of letting people do stuff, go to market, see what happens. And then, hey. If there's a problem, you know that you're probably gonna get smacked for it, but, like, you're kind of given freedom to operate until something goes wrong. Still, though, I was surprised by just seemingly how kind of free and, like, almost whimsical the DeepSeek, you know, releases have been where I'm not sure if, and maybe nobody's sure, but, like, it's unclear to me, like, did they run that by the, you know, the top leadership? Was that something that, you know, would rise to the sort of senior decision makers? Like, hey. This company is planning to launch or, you know, wants approval to launch an open source, basically, frontier model that's on the level of the best stuff that's out there. Like, I'm not sure, I could imagine a world where you'd say, yeah, Xi would definitely be personally looking at that and signing off on it. And I could also kind of imagine a world where you might say, actually, that may have taken them totally by surprise and, you know, they heard about it on Twitter with the rest of us. So what do you think is kind of the alertness to the AI moment right now in the Chinese government?

Angela Zhang: (1:00:03) Well, I definitely think that DeepSeek is quite independent in turning out their own products. And I, you know, obviously any business, whether you are foreign or domestic in China, you operate under the authority of the Chinese government, right? I mean, you need to play by the Chinese rules. However, the true creative destruction coming out from China, it's all from private businesses, owned, independent innovations, right? I mean, I almost have to be too harsh to say that anything the Chinese government touches, it's not doing that well, to be honest. Because whenever a government tried to allocate resources, it leads to corruption, a lot of inefficiencies and loss, right? I mean, and that's what we've seen and observed in the semiconductor industry, there's a lot of corruption scandals. And so far, China is not achieving that much of its semiconductor capability. And so they need to rely on Huawei. I mean, Huawei is not a state firm, it's a private business, right? I mean, so the true innovative capabilities still come from the private entrepreneurs, right? This kind of animal spirits of the business leaders in China. And DeepSeek is a maverick in the sense that it's not even a big tech firm. It's just kind of like a random hedge fund, quantitative hedge fund, who actually kind of like, because of regulation reasons, they were restricted from trading some businesses, a financial part of businesses. And so they devoted more resources to train AI models and became like an overnight success. I mean, so I think it's a very idiosyncratic success story of DeepSeek. Although at the same time, you see all the major big tech companies like Alibaba, Tencent, ByteDance, and including a lot of AI startups, they have also made tremendous progress with AI efficiencies. It is just that DeepSeek is the best, but everybody else is not so far behind. And I don't think any Chinese government has anything to do with that. But I do think the Chinese government is very supportive of the AI development. But if you ask me what is the first and foremost priority, I will still think it is the hardware. It is...

Nathan Labenz: (1:02:23) the...

Angela Zhang: (1:02:23) the semiconductor, because this is really China's bottleneck at the moment.

Nathan Labenz: (1:02:31) So is that how we should understand the recent announcement of this, like, mega funding, a hundred plus billion dollars? You know, I've seen it called a quarter trillion dollars in PPP. Is that a hardware build out more than anything else, or is there anything else that we should infer from that? I mean, you know, we're hearing here in The United States, like, not, I would say, like, you know, super credible calls just yet, but, you know, the Overton window now extends to Manhattan Project for superintelligence. And, you know, part of the reason, you know, people are willing to entertain that is they think that maybe China's gonna do its version of a Manhattan Project for superintelligence. Does this sort of, you know, large investment tell us anything about sort of how centralized the effort might be? Or maybe another way that we could, you know, frame that is, like, we hear a lot about civil military fusion in China. And, certainly, it sounds like we have, like, sort of maybe, like, media military or media government fusion, you know, from the sort of earlier generation of more, like, social media platform companies. But, yeah, like, what, where is this money going? You know, what does it imply? And to what degree do you expect even if, like, all these, okay. These things came up independently, but should we expect that, like, there's government monitors, like, at the DeepSeek offices, like, looking at all the developments now on an ongoing basis? Or, like, you know, what's... I think people actually think that's what's happening. So if you think that's laughable, like, you know, please disabuse me of that notion. But I really am not sure what I'm... I'm ignorant about what's going on for real.

Angela Zhang: (1:04:04) Well, I do think that DeepSeek will need to closely communicate with the Chinese government given that it's such a major player in the game right now. But for the most part, I think its intentions are benign because it's open source and it's kind of providing a public good for the rest of the world, right? I mean, so, and that's awesome, you know, from that standpoint, it also makes sense why they want to make it open source. I mean, also makes them more credible and also makes China become like an industry standard setter in a way. Right? I mean, China also gained a reputation being, you know, now recognized as the second most important AI superpower. And when DeepSeek is widely accessible everywhere, in a way China can potentially also shape the standards of the AI technology with its open source AI models, right? I mean, so I think the Chinese government is playing a long game. And on the one hand, they understand they are not the best positioned to create the technology or to run the technology. But I think that they also want to guide the technology in a way that maximizes the national interest. Now, why do I think that The US semiconductor export controls, and particularly all these export controls might fail and then China may, we might experience another DeepSeek moment with China's semiconductor breakthrough is because The US export controls have what inadvertently have created is that alignment of the private sector and the government's interests, right? I mean, so in Washington DC nowadays, because of the emergence of DeepSeek, there are a lot of industry leaders, policy analysts still advocating for very strict export controls of AI chips to China because national security reasons. Like Dario Amodei was one of the most vocal proponents of that idea. He wrote an essay that we should double down on this export control, right? I mean, we need smarter export control strategy and things like that. I think this school of thoughts, it just that they totally disregard the unintended consequences, not just the DeepSeek's innovation, but also the alignment of interest from the private sector. Because at that point, the private sector knowing they can't access advanced foreign technology, they have to innovate and they have to work with the government. They have to think about solutions. Right? I mean, so you're incentivizing that strategic alignment between the private and the public sector here. And that is the unintended consequences resulting from all this US restrictions. And I actually think that it's not a remote possibility, although a lot of industry folks were saying, you know, China is like a decade behind in semiconductor technology and things like that. I am more optimistic about China's capability. And it's, look. I mean, this is a country with 1.4 billion people, tons of really capable engineers, by the way, a lot of them came back from The United States because of The US anti-espionage investigation. So US lost out a lot of talents to China in the past few years, thanks to Donald Trump's China Initiative project. And so when you have people, that is exactly what Ren Zhengfei, the Huawei boss was saying that during that symposium with Xi Jinping, the most important thing is that we have people. Even if we lose out everything, as long as we still have people that we can innovate. Right? I mean, so I think if my projection is that a few years from now, we will experience another DeepSeek moment, and they're like, holy cow. How did the Chinese crack out this really difficult problem? And at that point, I think The US-China tech war will end because there's no other important obstacle along the way.

Nathan Labenz: (1:08:27) When you say that the tech war will end, what does that mean? Does that mean, like, we enter into a new period of, like, friendship? I mean, one question I often ask that I think is, like, too, you know, too quickly glossed over in many of these conversations is, like, why are the 2 countries at odds right now? I mean, I know that there's, like, a long list of grievances that both countries will cite. But in your view, is there, like, a fundamental interests-oriented analysis that leads to, like, unavoidable tension or conflict? Or would you just describe this more as, like, a sort of malfunction of leadership potentially on both sides that has sort of created conflict for maybe, like, domestic political reasons that doesn't have a more fundamental reason to exist long term.

Angela Zhang: (1:09:17) So, Nathan, you should invite me back for another episode. This is exactly what I'm writing right now. It's called America's Legal Gambit Against China. So now to just very briefly give a preview of and to answer your question, right, I mean, just think about the deterioration of The US-China relationship. When did it start? It started from 2016 after Trump was elected, particularly after 2018. The approval rating among The US people against China dropped dramatically since 2018 and reached its lowest point, just keeps decreasing over time. So we now enter the worst time of this relationship. Now, how does that happen, right? I mean, if you think about this outcome, it's driven by both demand and supply. So from a demand standpoint, there are 2 things happening. First is the trade imbalance as the economists were talking about, because China, after US liberalizes trade relationship with China, I mean, just the influx of Chinese imports in The United States. And that took away a lot of US jobs and caused manufacturing decline and generated a lot of discontent in US society. And that is a part of a story we have all heard about, right? I mean, that's why Trump's major, major campaign promise is that we need to bring back jobs to The US, it was China that was taking, stealing our jobs, right? I mean, and the other thing is China's rise, right? I mean, China's geopolitical rise has disrupted the preexisting great power order, right? I mean, challenging US hegemony, right? So these are all from demand side, why, you know, the relationship turns sour. But what I am writing about is from the supply side. We not only need to think about why The US needs to get tougher on China, and we also need to think about who are the, you know, what are the institutions that are shaping the policy response to address these demands. And I think Trump played a pivotal role in contributing to the deterioration of The US-China relationship. Because when we talk about foreign relations, the president, executive branch has tremendous discretion, despite the fact that US is a democratic country. But when it comes to foreign policy, things are very different, right? I mean, the power is very concentrated in a way that makes US look like China. Power is very concentrated in executive branch. Now, when Trump came into power, his major, one of his major cards is China, is against China. That was his brand, right? I mean, we need to be tough with China. And Trump broke a lot of presidential norms in taking a lot of actions against China. You think about he was the first to invoke Section 301 to impose tariffs on China, accusing China for technology transfer, right? I mean, that is something that broke the norms because that can, it's a violation of the WTO commitment. Rarely has been invoked, right? And highly unpopular. No president before Trump would ever rarely use tariffs to deal with this kind of issues. Right? I mean, he was the one who started this game. Trump was the first to, you know, started all this restriction of capital flow into China. First one's really toughen up sanctions and restrictions on Chinese firms like Huawei, ZTE, the investment bans, all these stories that we heard. Trump was the first one to propose banning TikTok, restrict information flow. Trump was also the first one to start this China Initiative program to start a massive anti-espionage investigation over thousands of ethnic Chinese scientists working in The United States that eventually these people just left and go back to China. Right? So he was the one who broke with all these norms. And when he was doing that, just like the Chinese government, right, when he was supplying this, you also drum up the demand because he also talks very tough on China. So you can see the public opinions was very, very negative. Like he would say, every Chinese student who comes to live here is a spy, and he called the virus China virus. Right? I mean, putting all the blame on China. So in supplying to this demand, he also increased the demand. He inflated the demand. So even after he left, you will see that Joe Biden inherited almost all his policies and in some areas, doubled down and intensified, particularly export control. Right? I mean, so this is a part that most people don't really talk about. So I think it's important to see that institutional comparison with China. And in fact, when I finished this article, I was thinking in a way, in its efforts to contain China, The US is becoming more like China ironically in so many ways. And you see concentration of power, the structure. You also see volatility with its practices. TikTok is a very good example. Right? Now Trump is the biggest savior of TikTok, by the way. And you're going to see fragility of the regulatory outcome. You see the DeepSeek emergence is a clear example of how The US export control generates huge backlash and the unintended consequences. Wipe out over $1 trillion market cap from The US tech firms. Thanks to this export control. And what has it achieved? Right? I mean, and by the way, the people in Washington DC still these days, they are still arguing for tough export control. And Trump, there's rumors around just today, saw on Twitter, that Trump is going to appoint somebody to the BIS and then it's going to get tougher with export control. Right? I mean, it's, you see once these institutions has been set up, they are there. They will have long lasting impact. Yeah.

Nathan Labenz: (1:15:18) So how do you think the Chinese government right now understands The United States' goals with respect to it? Like, do they see this in terms of, like, increasing, you know, volatility and maybe think like, jeez, you know, Trump's probably gonna be gone in a couple years, and maybe we can just, you know, we're gonna, like, obviously develop our own domestic chip industry, but, like, other than that, you know, hopefully, we can wait this guy out, and we'll kind of get back to senior leadership. Or do you think that it's starting to take on its own sort of, you know, hard to control momentum where they may be, like, really sort of seeing The United States as an enemy long term and in a way that will be, like, hard to undo or hard to get over?

Angela Zhang: (1:16:02) I think the biggest enemy for China is never The United States, it's that they mess up themselves. I mean, this is my book, the title of the book is called High Wire, How China Regulates Big Tech and Governs Its Economy. You can also say how China has messed up themselves. That tech crackdown wiped out trillions of market cap from China's big tech firms. How do we inject any competition into the tech sector? And no entries because all the private investors have gone and fled the Chinese market. They believe China is uninvestable. They, you know, they totally sabotaged private entrepreneurial spirits. Right? I mean, it was not The United States. Yes, there's tariffs, there's export control and all these restrictions, creating obstacles for the Chinese entrepreneurs. But that's not the biggest worry of operating in China. That's not the biggest obstacle. The biggest obstacle is that how do we have really true, great vision and good rule of law system in China that gives people the confidence to invest and to prosper, right? I mean, at the end of the day, we come down to a good governance structure. Same story goes to The United States. The US biggest enemy is not China. It's our democracy. It's what people are talking about constitutional crisis at the moment. How The United States messes up in itself. It's not China. It's out competing The US. Even if you look at DeepSeek story, it is a story about geopolitics because DeepSeek created the technology because of the necessity. Right? You can't get access to US export control. But you also ask yourself, why doesn't any of The US big tech companies come up with this efficient solution to train their AI models when they have abundant resources, when they are the clear technology leader, when they know very well what China has done? Because this is an Internet age. Right? I mean, for over a year. Everybody in Silicon Valley already knows Chinese firms are very good with improving efficiency of the AI model. But none of the big tech firms show any interest in doing that or following similar strategies. People say this is a good thing from the Silicon Valley. I don't think so. I think at the end of the day, it is a deliberate business choice. I just wrote in Project Syndicate called "How DeepSeek Upended The US Tech Dominance." It's because The US tech industry is too concentrated. If you think about the AI industry, the infrastructure level are controlled by a few monopolies, right? At the very top you have Nvidia, which is a virtual monopoly. And then the 4 big cloud makers, right? I mean, these people, they also invested downstream, all of them, right? They're in a way kind of vertically integrated. The money is all upstream. That's where the profit is. That's where the profit margin is, right? Nvidia, what is the operating profit margin? Over 60%, but where they sell downstream, all the AI firms, OpenAI, do not end up in profit, these are losing businesses, right? So the money, the monopoly profits are upstream. They're not interested in competing downstream. They want to increase the demand of the AI chips, the computing resources that will continue to project a very rosy picture of the demand, the scarcity of the craziness for computing resources. That's why they have to scale things up. That's the business strategy, right? That's how Silicon Valley works, right? So if you, when we think that, oh, how DeepSeek succeeded, we also need to reflect upon ourselves why we fell. We fell not because we lack the innovative capability, which we clearly have, it's because of the concentration problem in The US that blindsided innovation and misled innovation trajectory.

Nathan Labenz: (1:20:10) How do you think about your own position in this crazy world? I mean, having grown up in China, you know, my understanding is, like, the Chinese state may still be taking an interest in what you're saying even as you live abroad. I don't know if you feel that way, but I've certainly heard people say that that is something that the Chinese state does. You're, you know, clearly willing to be critical of both countries' governments. Like, do you feel, like, you know, totally comfortable in fine doing that? Or, like, I guess or, you know, maybe what is your sort of strategy or goal in the work that you're doing in all this public communication?

Angela Zhang: (1:20:51) Well, I think that my value as a scholar, just think about, you know, if I see myself as an entrepreneur, I am also producing product, which is my scholarship. And I need to have my brand. My brand, like I mean, is I am very neutral because I grew up in China, I've worked very extensively overseas in Europe. I also know Europe pretty well. And The United States. I think I can offer a unique perspective being very neutral and honest. And when people read my work, they will see how critical I am with the Chinese government. I'm equally critical with what I'm currently writing about, about the US government. I think we need this kind of niche product at the moment, right? I mean, who can really give a fair and honest opinion on what is going on? I think that there is a demand for this. And thanks for your podcast and how my product is diffused. Right? I mean, disseminates my ideas, which is very important, which scholars don't usually do.

Nathan Labenz: (1:21:54) Niche product, but we're doing our part. What about, you know, this may be one that's, like, too far outside of, you know, what's knowable to really have an opinion on. But in AI safety circles, there is the notion that maybe Xi is a doomer. And, you know, what exactly that would mean, you know, in American AI safety circles, it's like, literally, you take seriously the possibility that AI could lead to, like, human extinction or sort of collapse of society. And as such, you're, even if you're, you know, historically somebody who's very pro technology and very libertarian in your politics, you're still inclined to have sort of a special level of caution and inclination toward regulation for this particular wave of technology that is the, you know, the frontier AI. Is there any way to read what Xi might really be thinking with respect to these extreme risk scenarios? And is there any reason to think that he, like, can or can't be engaged or reasoned with on that level?

Angela Zhang: (1:23:01) Well, I have no idea what President Xi is thinking about. But from what I observed, right, I mean, in The US and in China, both countries are pressing AI development at full steam. Right? I mean, if you're doing that, then if you're rational, then you shouldn't be an AI doomer. You should be an optimist thinking that AI is doing more good than bad things. And unfortunately, I think this is a very sad trajectory. I mean, because we already seen AI can have a lot of dangerous capability along with its great promise, right? So I wrote an article called "The Promise and Perils of China's Regulation of Artificial Intelligence." And taking this very lax stance potentially could also sow the seeds for potential disasters and catastrophes later down the road, particularly when we think about the dynamic pyramid model of China's governance, right? And things don't really get attention from the very top until it is too late, right? I mean, and that's a recurring theme in China's regulatory governance. And think about COVID-19 pandemic control, think about the tech regulation, property crisis, one after another. And so, will AI be the next disaster? We'll have to wait and see. But at the same time, The US is following China's lead in a way, you know, it's also relaxing regulation and going to another extreme. And this is really truly worrying because after all, The US is still the clear leader in this technology.

Nathan Labenz: (1:24:35) Yeah. Certainly at the very frontier, I agree with that. So last question, how much value do you see in sort of track 2 dialogues, particularly, you know, motivated by the idea that, like, jeez, this AI stuff could get really powerful. One day, we might need to step back from the brink of, you know, all out competition and have some sort of arms control or coordination or whatever.

Angela Zhang: (1:25:01) Mhmm.

Nathan Labenz: (1:25:01) How much value is there in track 2 type of work, and do you have any advice for people who want to build these sort of, you know, bridges or relationships or whatever that might one day be useful in that scenario?

Angela Zhang: (1:25:15) I think what the world needs is a disaster at the moment. Until a disaster occurs, let's hope the disaster is not so disastrous that it is going to kill us all, that is still relatively contained, that we need to show the world's leaders the dangerous capability of AI, the potential misuse of AI can turn into a huge AI accident that really threatens social political stability. Then our leaders, all the politicians all over the world, they will wake up and then we'll take proper actions. So in my opinion, the biggest obstacle to US-China AI cooperation is not geopolitics. It is the fact that we haven't seen the disaster. It's the fact that we still don't know, don't quite understand what the capability of AI is. We don't understand AI. We don't know how to regulate it. This is, I'm actually quoting Geoffrey Hinton. He's the kind of guy who came up with this idea. He was saying, we don't understand how these things work because we don't understand how the brains work. This is more or less the brain. And we just know we feed them with computing power and data and algorithm and it works. But how it works? What capability it would turn around? We don't know. So there's a lot of uncertainty and unknown when it comes to regulating AI. And that's the reason why regulators are not serious, right? You think about during the Cold War, The Soviet Union and The United States, even at the brink of a hot war, they still signed nuclear non-proliferation agreements because they know nuclear weapons can kill us. But we don't know what will happen with AI.

Nathan Labenz: (1:27:07) So that's a sort of, you know, I could imagine a more optimistic take to say the least, but I honestly don't disagree with that assessment. In the meantime, though, I think a lot of people sort of feel like, what can I do now that would be useful in a crisis later? So wanna just give you one more chance to advise me if I wanna do something to marginally improve the situation, you know, what else could I do that might build some social capital or communication bandwidth or something that might be useful when that disaster might hypothetically hit?

Angela Zhang: (1:27:42) I think there was a proposal, which is a very sensible one, of creating an independent agency modeled after the IPCC, which is like in the environmental sphere of collecting intelligence of climate and disaster. I think if we have an independent international organization, staffed with good scientists all over the world, share intelligence about AI capability, right, we already gather a lot of evidence. And then we go back to update our government about what's really going on and what are the potential consequences and disasters. At least we make sure we're on the same page and together to understand and trying to take precautions before things are too late. Right? I mean, just think about if there is a 1% or 0.01% remote chance that AI can really get out of control and the consequences could be disastrous, is it worth us investing a little bit resource in that? Clearly is, that's the whole point why we buy insurance, right? I mean, so to completely abandon AI regulation to walk away and just totally overlook this and only intervene until things too late is not a very wise regulatory approach.

Nathan Labenz: (1:29:01) Yeah. I'm with you on that for sure. Alright. I'll look more into those proposals. I think this has been a fascinating 90 minutes, so I really appreciate it. Any closing thoughts or you wanna share, like, where people can find you online or anything like that before we break?

Angela Zhang: (1:29:17) Harvard Law professor Cass Sunstein said I have the best, well-made personal website of any legal scholar in The United States. So if you search for my name, Angela Huyue Zhang, you'll find me online. He said that was the most elaborate personal website of any law professor in this country.

Nathan Labenz: (1:29:38) Alright. Well, we'll have to check it out. For now, again, thank you for doing this. Professor Angela Zhang, thank you for being part of the Cognitive Revolution.

Angela Zhang: (1:29:48) It's a great pleasure. Thanks for having me.

Nathan Labenz: (1:29:50) It is both energizing and enlightening to hear why people listen and learn what they value about the show. So please don't hesitate to reach out via email at tcr@turpentine.co, or you can DM me on the social media platform of your choice.