Dear reader,
Happy Saturday! I skipped posting yesterday since I had the honor of publishing a short summary of DeepSeek news for
. I did my best to make it as readable as possible for an I-don’t-follow-AI audience.If you liked the prior tech right analysis, I wrote up a “grand theory of the tech right” for Twitter—my provisional guess at the faction’s origins, which blend material interests, coalitions, and dispositions. Marc Andresseen himself called the thread “Very thoughtful. 🇺🇸.”1
Anyway, today’s podcast features my dear friend Tianyu Fang! Tian is a writer and researcher focused on tech policy and US-China relations. He’s currently a Tech and Democracy fellow at New America, and is a cofounder of the iconic Chinese internet newsletter
. Among other things, we discuss:Tian as “poster child of the Arab Spring”
How China journalism died
Censorship and its discontents
Has the “China model” won? (large language)
Has the “China model” won? (political)
Why everything is a “Manhattan Project” now
The Falun Gong’s alternative internet
Watch/listen by clicking above, the Substack app; or add it to Spotify, Apple Podcasts, or Pocket Casts. A full transcript and list of links is below.
Episode transcript
This transcript has been lightly edited for length & clarity.
Jasmine Sun (00:00)
This week on the podcast, we have my friend Tianyu Fang. Tian is a researcher and a writer focused on tech policy and US-China relations. He's currently at New America, and he was also a co-founder of the iconic Chinese internet newsletter, Chaoyang Trap, which is sadly on indefinite hiatus, but you should check out the archives.
Tian and I met working together on the Substack Design Team. I remember our very first call very distinctly because we realized that we are both huge fans of Stanford media historian Fred Turner, who wrote the book From Counterculture to Cyberculture. And as soon as somebody is a huge fan of Fred Turner, I'm like, “my God, we have to work together.” So we manufactured a role and we've been friends since then.
Tian Fang (00:52)
Thanks for having me, Jasmine. So good to be here.
Jasmine Sun (00:59)
You're in Beijing right now, is that right?
Tian Fang (01:05)
Yeah, I've been here since for the past month or so. I'm here for Chinese New Year; and meeting people and checking in with my sources. But I did grow up here. I lived in Beijing until was 14. And that was when I moved to the States. But I come back here pretty often, with the exception of COVID.
The end of China journalism
Jasmine Sun (01:16)
One funny thing that you told me was that you have been professionally explaining China to Westerners since you were a literal child. How does one become a professional China explainer?
Tian Fang (01:42)
I think literal child is an exaggeration. But I was in Beijing until 2014 or 2015. And that was sort of the final heyday of China journalism for foreigners. And it's right before the 2016 US election and, three years into the Xi Jinping government, So that was when a lot of longtime China watchers from the US were leaving the country, right? You had
moving back to the US in 2016. And I remember going to this public event in Beijing featuring and from the podcast. And that was when they were also moving out of Beijing to join SupChina, which has now gone out of business, unfortunately.I moved to the States for high school, and after a few years in Massachusetts suburbs, I began to write articles for English-language publications. I've always been interested in topics on Chinese politics. Back then it was the “China model debate”: whether the the Beijing Consensus was really an alternative to Western democracy. You also had a lot of interesting social movements going on on the Chinese web—social media, censorship, propaganda, that sort of stuff.
I wanted to participate in these conversations, but felt like it would be very difficult to participate in them as a high schooler. People wouldn't take me seriously. So I thought the best way to do it was to hide behind the title of a freelance writer. I didn't think a lot of the China reporting was very good, so I thought, okay, maybe I could do this as well.
Jasmine Sun (03:35)
That's so cool. I was doing the same thing, but much less noble. I wrote on medium.com in 2015. I was like, there are all these adults saying their political opinions, and they’re not very good. I'm pretty sure I can do this. On the internet, nobody knows you're a child. So I was a medium.com politics and culture writer, which is very embarrassing.
Tian Fang (03:44)
No, I think that's great. I feel like there’s a culture that both of us subscribe to, which is that you can do whatever you want as a teenager. Instead of participating in Model United Nations or writing for a newspaper, there are things you can do in the actual world that nobody's stopping you from except this mental block.
Jasmine Sun (04:19)
I must confess that I did one Model United Nations conference.
Tian Fang (04:31)
I also did Model United Nations, and high school newspaper, so there's that.
Jasmine Sun (04:35)
Okay, a bit of both. Since you've been doing writing and research about China for a long time, how has the China journalism environment changed over the last 10 years?
Tian Fang (04:52)
By now, doing journalism in China is sort of an infeasible career. Journalism in general has become a sort of unfeasible career for people around the world. Print and online media publications are going out of business.
But the environment in China is particularly bad because of tightening government restrictions in the last 10 years, and that has mostly affected Chinese media outlets. In the past, you would see very bold investigative journalism championed by publications like the Southern Weekly, which was a party publication based in Guangzhou in the early 2010s. They were very courageous about their revelations of corruption, government bad behaviors, etc. That is completely gone. The quality of journalism in China has really gone into decline. And that’s not the fault of the journalists, but rather government restrictions.
For the international press specifically, though, you basically saw the closure or downsizing of China bureaus in the last few years. Part of it was because it's become very difficult to get journalists visas for Americans. China kicked out American reporters in 2020 during COVID, in retaliation for Trump kicking out Chinese reporters the same year. That was one of the most disastrous policies under Trump's China policy in the first term, because the US lost access to some of its most important resources. So starting in 2020, a lot of the China reporting was being conducted in places like Taipei, Seoul, Hong Kong (to a lesser extent now), Singapore, and Washington DC. A lot of the China beats are following DC sources.
And in China domestically, there’s less incentive to talk to foreign reporters. You're really not rewarded in any way, right? And you face repercussions in your own organization or from the government. So people really don't want to talk to foreign reporters. It is very hard to build sources and get scoops if you're a young journalist. So we basically see the dying of an entire generation of good, high quality China journalists. Years ago, a lot of the seasoned China journalists from places like Reuters were able to figure out who was going to be on the next Politburo Standing Committee days before the actual announcement. And in the last Party Congress, we didn't see any of those stories. Nobody got it right.
The consequence of not having actual scoops from Chinese insiders is that publications instead rely on shoddy psychoanalysis of Xi Jinping. It’s like, I don't know what the internal mechanism is, so this policy must be because Xi Jinping was traumatized as a kid. Or like he said something on a particular day, but in reality Xi Jinping might have said a lot of things, and those don't all translate to policy actions. Without understanding the internal mechanisms, reporters increasingly see the country as a black box. We're always making these inferences about what is actually going on. That's not a good development.
Simultaneously, Western readers have lost their appetite for more nuanced China stories. In 2018 and 2019, there were still a lot of Silicon Valley companies trying to enter the Chinese market or build a business presence, be it reaching out to Chinese consumers or advertisers or having part of their supply chain in China. But now they're no longer doing that. If anything, they're trying to get out because because of tension between the US and China, and China’s economic decline. There was a post-2008 gold rush when a lot of foreigners went to China. China was also internationalizing very rapidly because of the Beijing Olympics. But that's also over.
A former Financial Times reporter in Beijing used to say that there are three types of China stories. There's Big China, Scary China, and Weird China. Now you don't have to tell the stories that China has a large population and it has a sizable economy. I don't think either of those two things is going to grow as fast as it used to. The country also isn't as weird anymore, given more cultural convergence. There’s Chinese students in the US. There are a lot of Americans who are very well-versed in Chinese culture. And the “Chinese people eating bats” stories, that’s not selling as well. The stories that sell the best right now are the Scary China stories. You have a lot of stories that are basically exaggeration of a small phenomenon.
One example is the social credit system that people have been talking about for years. It is almost comical that after all these years with credible journalism and academic articles written about how the social credit system isn't what people think it is, it is still widely discussed as if it was something that people have to deal with on a daily basis. I'm sure there more Americans who are aware of the “social credit system” than Chinese people who actually live here, which is befuddling but also annoying. The original reports were based on the experimental pilot program for one small city in China, in Shandong Province. And they completely ditched it a few years later because nobody was using it and there was no real repercussions.
But that story still went on. It just gets replicated. It's almost as if some center in Alabama is trying to implement a policy, and reporters write it as “This is what the US is doing.” And there's some truth to it, but it's not the whole story, right?
Jasmine Sun (11:46)
Yeah, the social credit stuff is wild because it is incredibly widespread. Everyone thinks of the Black Mirror episode, like, Did you know that China actually has that? And it's a good meme. It corresponds to the narrative of China as a foil to the US and its freedoms. It's a nice villain story, and it's very memorable. Whereas saying China has this complicated pilot policy and then they stopped doing it, but also separately there's financial credit ratings—this is not interesting to people. It doesn't stick.
Tian Fang (12:19)
Also, a lot of the financial ratings policies and systems are completely based on similar US systems. When you look at the planning documents for how to implement the so-called social credit system, when they were trying to do it, they would talk about FICO. They would talk about how Americans were dealing with it, how Europeans were dealing with it, and how they saw a similar need in China.
A lot of people fail to see that these questions are interconnected at a global scale. And Chinese bureaucrats are indeed learning a lot from American bureaucrats.
“Poster child of the Arab Spring”
Jasmine Sun (13:02)
One last personal detail: These days you write a lot about tech policy and internet dynamics, cybernetics, etc. So were you also into technology as a kid?
Tian Fang (13:18)
Yeah, I think I've always been more interested in tech than anything else. I became interested in politics mostly because when I was a kid, I was really into jailbreaking iPhones. There was a small but sizable and influential jailbreaking community in China. I started hanging out with them and realized a lot of the things you needed to download were actually on like four websites on US blogs. I started building VPNs so I can access them. This was around 2011-2012, and internet censorship wasn't that strict yet, so it was relatively easy to circumvent censorship even as a 10-11 year old kid.
I was like, OK, now that I'm on this unfettered internet, I can start reading about politics. And that's how I got interested in Chinese politics. I like to say that I was the poster child of the Arab Spring because this was around the time when the Clinton State Department under Obama was pushing for global democracy through the internet. The US must defend internet freedom so that people in Iran and China and Syria could protest their governments. So I had this techno-optimist outlook: I thought social media platform controlled by US corporations would be a vehicle for anti-autocracy movements around the world.
Of course, that didn't really happen. And now we think social media platforms are the complete opposite—the roots of all evil. But back then, I was really involved in the parallel development between information technology and global democracy. I'm still interested in that. And a lot of the work I'm working on is related to that history and that process.
Jasmine Sun (15:27)
That makes sense. If you develop your political consciousness because you were able to access a free and open internet through building VPNs, I could definitely see logically how you’d go to: oh man, if only everyone had this experience.
One thing I'm curious about: for a teenager in China, what is the perception of VPNs and the Great Firewall?
Tian Fang (15:58)
The perception has changed over time. When I was 11 or 12, I don't think people were talking about it at all.
But it's relatively common in this country, right? The reason that people use VPNs are usually not for political reasons, which is the counterintuitive thing. As Ethan Zuckerman used to say, people go to Instagram and Twitter to see cat pictures. People who most actively use VPNs in China today are not trying to look for content about the Tiananmen massacre. That's not what they're looking for. They're looking for cat pics. They're sometimes looking for pornography, because pornography is technically banned in China. And they're sometimes looking to follow South Korean idols on Instagram.
I do believe that the Chinese government has no intention to create a whitelisted internet. That was something people were worried about for a long time. Every year during sensitive dates, like anniversaries of political events and political conferences in Beijing, it gets harder to access the international internet. It is harder to use VPNs. So you know that the government has the technological capability to identify which traffic patterns are from VPNs, and they could have banned it if they wanted to. But they choose not to.
Jasmine Sun (17:21)
Why don't they want to completely cut off access?
Tian Fang (17:26)
For one, China is still an export-based economy. A lot of textile makers in Zhejiang are trying to export their stuff to the US and Europe. They need to go on WhatsApp to talk to their customers. There is a huge demand for that basic stuff.
Frankly, most people who are using these services are not trying to subvert the state. There are better ways to do it. Also, I think from a cynical perspective, it is easier for the government to monitor potential political activities. Imagine there not being Twitter—then the Chinese government would have a harder time identifying who the potential “threats” to their regime are. The last part is mostly my speculation, but there are a lot of reasons you would use a VPN that the Chinese government considers legitimate.
Jasmine Sun (18:22)
Like, it would be really bad for the economy if a lot of folks doing exports or like international business all of a sudden just couldn't talk to people.
Tian Fang (18:33)
Even the propagandists themselves. The Global Times guy, Hu Xijin, he pays for his own VPN.
A lot of Chinese people actually sign up for Twitter and pay for a VPN so that they can see these propagandists, right? Because they think it's funny. They think China is now becoming more outspoken on these social media platforms. And they want to see that in real time. So they sign up for Twitter.
Jasmine Sun (19:01)
Do you think all the rural culture stuff is propaganda? You know how everyone says Liziqi is propaganda or whatever?
Tian Fang (19:19)
I don't think so. She's just hanging out. I think she also has a team that operates foreign social media for her. Same thing for this guy, Hebei Pangzai. He chugs beer on Twitter.
I don't think they're propaganda. They just wanted to make some money. As you know, the advertising revenue on platforms like YouTube are way higher than their Chinese counterparts. They're not actually pushing for any pro-Chinese government content either. They try very hard to avoid that.
Jasmine Sun (19:55)
I do feel like a lot of people assume everything China does is a psyop, and they forget about wanting to make money. In reality, almost everything is I want to make money much more than I want to steal state secrets. Nobody really cares that much about state secrets.
Tian Fang (20:14)
Yeah, I think they really overestimate how much the average Chinese person cares about their government. It's probably the same for Americans as well.
Jasmine Sun (20:23)
My Chinese grandparents are obsessed with the idea that there are gunmen at school, and I should run and hide whenever the gunmen show up. And I was like, there are no gunmen. But it's very prominent in the Chinese imagination that every American school is full of guns.
Tian Fang (20:41)
A lot of these caricatures see countries as monolithic when they're a lot more complex than that.
Chinese vs. American AI dreams
Jasmine Sun (20:47)
One topic I really wanted to chat about was DeepSeek. My entire Twitter feed has been taken over by buzz over their latest model. DeepSeek is the current leading Chinese AI lab. They just released a model called R1, which ranks as well as OpenAI's o1 model on many metrics, but it's way, way cheaper. And it's totally open source. People have already successfully replicated the model. I tried it, and it’s a really, really good model. At certain tasks, like writing prose, I think it's way better than Claude and ChatGPT, and I use those pretty regularly.
And my Twitter feed is flipping out because everyone's like—what happened to our export controls, how could they have figured this out—because it's using a very different algorithm and different approach. I saw a tweet from Dan Hendrycks, where he was like, as soon as everyone starts talking about RL at the SF parties, then that knowledge gets to DeepSeek in China.
So I don't know, Tian. Do you think that they managed to build R1 because they just had so many Chinese spies at the SF parties?
Tian Fang (21:58)
It is my strong belief that every single Chinese person on planet Earth is a Chinese spy—or at least, that seems to be the consensus for lot of people in San Francisco.
I find that idea hilarious because in the last 20-30 years, there are two predominant ideas popular among China hawks. One is the theory of the China threat, right? China is going to be a threat to the world because the Chinese government is going to be omnipotent. It can do whatever it wants to get what it wants. On the other hand, China is incredibly weak. It cannot innovate. Despite its population of 1.4 billion people, a lot of good universities, and a lot of good companies, they can’t do anything without stealing American technologies. It's fine to believe either of these two, but sometimes people believe both at the same time.
Jasmine Sun (22:53)
Well, they created a social credit system, right? Like they created this extremely complex system where like if you jaywalk, you can't buy a train ticket anymore, and if you're mean to your neighbor, then everyone will see a floating number above your head.
But they can't create LLMs. They're too stupid to create LLMs.
Tian Fang (23:08)
They're going to create world domination facial recognition technologies, but they're never going to create LLMs. This is a ridiculous presumption.
One thing I've observed is that in SF, where I live, people like to associate technological development with world historical missions. You're not just building computers, you're trying to turn the world into like a cybernetic utopia. You're not building chatbots, you're trying to get to AGI before anybody else.
Here I'm reminded of how little these meta-narratives matter to the Chinese counterparts. The average Chinese tech entrepreneur compared to the Californian counterpart cares so much less about making utopia or dystopian technological futures and cares so much more about making raw profit in a hyper-competitive market. When 50 companies are doing exactly the same thing as you, they care so much more about making some money in a bad economy.
And the people who do care about technological futures are usually Silicon-Valley-pilled. Zhang Yiming from ByteDance is one of these people. I think people from DeepSeek are like that as well, but I'm not super familiar with their culture.
They also care much less about national interest or winning the AI war. You just don't hear that here, compared to the fearmongering and warmongering that I encounter on a weekly basis in Silicon Valley. I mean, sure, there's a lot of military technology companies and government contractors, but when was the last time you heard a founder of a Chinese consumer tech company vowing to publicly support Chinese efforts to topple US technological supremacy? I think there's sometimes lip service to the party or government. But I can't think of many founders who do that.
Jasmine Sun (25:09)
They don't have a Chinese flag and a rocket ship in their WeChat name?
Tian Fang (25:15)
No, no. That kind of stuff is way more popular among Americans. And frankly, these Chinese entrepreneurs and engineers just really want to build a company.
Jasmine Sun (25:26)
Why do you think that it is so much less ideological? Why is Silicon Valley so obsessed with meta-historical narratives? Whether it's the early cosmopolitanism, like Twitter and Arab Spring, we're going to democratize the world; or whether it’s that AGI is extremely important, and we are literally building the machine god. Is there something unique that creates these narratives?
Tian Fang (25:49)
I don't have a very well developed theory for it. But part of it is counterculture impulses. Part of it is the fact that the US is a way wealthier country, so you have the economic capital to sit in the room and think about AGI when, in a lot of other countries, people have to work their asses off.
I think there is a lot more interest in theory in the US. In contrast, I've generally found that Chinese entrepreneurs and investors are a lot more skeptical of meta-narratives, of the Hegelian view of history. After so many decades of disillusionment with Chinese socialism, it's very hard for them to be convinced that there is a particular way that history will progress. People genuinely try not to think about it. They just have more interest in succeeding in their “careers.”
Jasmine Sun (26:57)
it is interesting. I can trace the cultural roots in the US of, you know, John Perry Barlow and the Declaration of the Independence of Cyberspace. I'm like, why is that so compelling to people here?
Tian Fang (27:04)
It's also because the venture capital model in the US is different. There is more reason or incentive to do PR—to say, the next thing is American Dynamism. Or the next thing is software.
That stuff is more state-directed in China. And there is more skepticism of the state; people have a very cynical view of government initiatives. They wanna take advantage of it, but they know they can't trust it.
Jasmine Sun (27:44)
Yeah, I think
made a similar point. He was saying that part of the reason that everyone's tweeting and blogging—e.g. the crypto thing, where we’re all going to be in this decentralized network state future—is because a lot of American tech is funded by extremely early stage capital. You're in a big recruiting and talent war. The meta-narratives, like having cool blog posts and viral tweets, are how you direct venture funding. Also no one makes profit for a really long time. We all just have ideas. No one has a working product or more than one customer. So how do you differentiate yourself and show you're the founder who's gonna win? You got to have the most ambitious vision.Tian Fang (28:25)
Yeah, totally. The venture capitalists need to create a self-fulfilling prophecy. They say, look, this is going to be the future. It is often framed as a prediction when they're making it, but sometimes it wouldn't have been the future if they didn't actually push for it.
Jasmine Sun (28:50)
And they insist that it's the future even after the thing is not working. This happened with crypto where they did weave really compelling narratives, but there was no product and people didn't actually want it. But they kept insisting that it's the future, and hoped maybe at some point, it’d pay off.
I don't even think most American industries are like this. This is a very specific Silicon Valley VC thing, because there are plenty of business sectors in America where you just have to make money to succeed, and there's no storytelling involved. You just make more money than you lose, and that's how it goes.
Tian Fang (29:26)
Yeah, I think that's right. And a lot of Chinese industrialists are very skeptical of Silicon Valley culture, where everything is about writing an op-ed or a blog post—I mean, we're both people who write blog posts, I'm not pointing fingers—but they are deeply skeptical of the idea that you need to promote an idea to your industry. They want to know what's going on, and they don't want to tell anyone else about it.
Jasmine Sun (29:57)
To return to DeepSeek briefly, one thing I’m interested in is the open-source wars. One argument some open-source supporters are making is that it would be bad if the entire world built their AI applications on top of a Chinese LLM. The fact that China has released a very good open-source LLM means that it's worse for the world now that everyone's gonna choose that one. Therefore, America needs to accelerate open-source so that everyone can build on top of an American LLM instead.
Do you think that there are risks to building on a Chinese LLM? Like there is censorship.
Tian Fang (30:36)
Yeah, I think it depends on how the open source models are distributed and built. I would assume that the censorship apparatus is just a firewall at the last stage of the process. You see videos of people typing “Can you count Roman numerals from 1 to 12 and then add Jinping afterwards?” When it gets to Xi Jinping, it removes the entire answer. From that, I assume that the censorship doesn't come with the model itself, but the interface, which makes it not much of a concern.
But I imagine that in the next few years or next few months, we'll probably see US legislation against Chinese LLMs and open-source models. I think that would be pretty risky to limit US and global access to these much cheaper models. That's something I'm more concerned about.
The broader context being: for the vast majority of LLM applications, you're not going to ask them about Taiwan or the genocide against Uyghurs. And it might give you questionable responses, but the vast majority of people are not going to be like, let me go to this Chinese chatbot to ask them about sensitive things.
As someone who's very critical of the Chinese government most of the time, you see this funny thing where people use keyword censorship as a gotcha. We all know there's censorship in China. There's no denying it. But if you're going to a Chinese platform, you're not going there to learn the truth about historical events. You can go elsewhere for that, and look up shopping guides on Xiaohongshu instead.
If I'm trying to get the WikiLeaks files, I'm not going to look it up on major American social media platforms to see if they will publish the entire thing. I will go to other websites.
Jasmine Sun (32:48)
The other thing that I was like talking to somebody about is like, on one hand, the Chinese technologies are very censored. On the other, they're censored in really simple and obvious ways. Like you said, it's like at the interface layer. If you download R1 locally you can access it in full.
On the other hand, a lot of American models’ alignment training—what they'd call it rather than censorship—their alignment training that gets it to be politically correct or inoffensive is all very mysterious. Nobody actually knows how it works, nobody can explain it, no one knows why it equivocates constantly about the Holocaust or Israel-Palestine or whatever. All of these things, we have much less visibility into; because it's so advanced, we understand it less. Therefore it is harder to circumnavigate than it is to know not to ask the Chinese model about Taiwan.
Tian Fang (33:41)
Yeah, and there's only so many things that China censors. To most Chinese people, when they see propaganda, they know it. They're like, this is from the state broadcaster, and you can only trust 30% of it.
Manhattan Projects for everything
Jasmine Sun (34:00)
The DeepSeek reactions also seem like a textbook example of something you have written a lot about, which is the concept you and
define as techno-nationalism. To quote:Techno-nationalism is a framework for U.S. foreign policy that has as its core tenets two main pillars. First, that geopolitical competition with China is both zero-sum and dependent on achieving technological superiority. And second, that national security and foreign policy considerations in the competition trump domestic considerations.
You guys contrast techno-nationalism to early 2000s techno-optimism: the spirit of the internet and social media as an inherently democratizing, liberalizing force. That we should preserve the open public square and preserve free expression—that tech transcends sovereign borders, and is where we'll all get free.
This is a useful framework for understanding the shift in Silicon Valley and DC's orientation to tech policy. I'm curious for your take on where it came from. Why have we seen such a rapid shift from Clinton-era policy to the techno-nationalism that dominates today?
Tian Fang (35:10)
The elephant in the room is that the US is no longer the only technological leader, right? If you look at 2008 or 2010, almost all social media platforms were owned by US companies in this 30 mile radius around San Francisco. Now it is clearly no longer the case. We have companies like TikTok, a lot of AI surveillance companies in China. There is much more—I think legitimate—concern about what kind of values are being embedded in these technologies and who are the ones that can control it.
Second, I think there’s a decline in the globalist aspirations we used to have. At least on the center-left, the US used to care a lot about moral issues and normative issues, about human rights in other countries. These days people are more concerned about domestic issues.
Finally, a huge shift in this story is the fact that consumer technology does not make money. There's a sense that consumer internet is not real technology, that social media platforms are essentially several layers virtualized from the actual hardware underneath it, and without the hardware supply chain or advanced semiconductors, none of this would have been possible. So I think there is a weird Marxist turn—a pivot to materiality underneath the technology. Information is actually not so disembodied. Information is actually not so globalist. Rather, it is based on these hard technologies that the US can—and to a lot of people should—still dominate.
The same thing is happening in China as well. More so in many ways, since China never believed in the global public sphere. But you do see the country becoming more nationalist over time about hard tech and deep tech being the main mission the country should pursue as opposed to pushing social media platforms to go to other countries. I think TikTok is probably one of the few exceptions. And compared to semiconductor bottlenecks, I don't think the Chinese state cares as much about TikTok as the hardware breakthroughs.
Jasmine Sun (37:27)
That does make sense: US policymakers thinking the global public square made sense insofar as they could set liberal rules for how that square would operate. They assumed that the US would be good stewards of that square, but if another country was to near technological parity, they would not necessarily have the right values and rules.
The question that has been more of a mystery to me is why Silicon Valley itself has leaned so hard into techno-nationalism. I'm not saying like every single person in Silicon Valley is now techno-nationalist, but they should be the biggest champions of the open web. They should be excited about how open-source helps the “little tech agenda” or whatever. They benefit incredibly from global talent and global trade. And if they can run ads on more people around the world, they can make more money too. Like I assume that's why Facebook was so interested in the global project—because there are a lot of people in other places they can run ads on. So I'm curious why Silicon Valley right now isn't “standing up” more to techno-nationalism, for their own material interests.
Tian Fang (38:36)
That's a totally valid concern. My answer would be there is significantly less money and opportunities in building social media platforms or UberEats or another ride-share app.
There is way more interest in building things like defense tech; there's more interest in building green energy labs. These things need a lot more investment than consumer applications, and the money isn't coming from consumers. The money is coming from places like the DoD. They're coming from the NSF. They're coming from the DOE. I like to half-joke that Silicon Valley libertarians hate getting money from the government, but they love getting money from the military. There’s a lot more incentive for people to hype up the China threat or “American Dynamism” so that they can secure more funding and more government contracts.
The type of free market libertarianism that was most influential in the 90s—the crypto-libertarians trying to fight the US government because they were vehemently against export controls on both hardware and encryption algorithms—that was more or less an aberration from the Silicon Valley culture. Think about like 1970s: The largest employer in Silicon Valley was Lockheed Martin. It used to be an offshoot from the Los Angeles and Santa Monica defense-industrial complex. People now see opportunities in becoming part of the American mission to restore technological supremacy.
We were talking about how a lot of the American Dynamism portfolio companies have nothing to do with American Dynamism. All the San Francisco teams say “We’re building the Manhattan Project,” but in reality, they're just building a Chrome extension.
Jasmine Sun (40:32)
It was really funny. I was looking through the a16z American Dynamism portfolio, and you first get this really grand timeline graphic. It's like the Wright brothers take flight, and then the Manhattan Project, and it's these huge, giant, hard tech projects.
Then I'm clicking on the companies and it's like, We help make animations for science teachers in schools. I think that's awesome, but is that the same thing? Or there's one that’s literally a DEI firm. It said: We help Black, Latino, and indigenous early-career people find jobs in the trades. Like, cool. But is this the Manhattan Project? There's one about discovering your local parks. You can find where you can go kayaking near you. What is going on here?
On one hand, it's funny. And on the other hand, it's dangerous. Everyone is pretending to agree that there is this huge war we have to prepare for. And it seems like there are probably bad consequences. Even if people are just acting in their economic self-interest because they want grants, everyone's creating a false consensus, right? That we are on the brink of this Cold War.
Tian Fang (41:44)
You see this in Washington, DC as well. You have so many people who do not believe in the national security argument for things, but they have to leverage it so that they can get things done, especially with the weakening of federal agencies. Even under the Biden administration, they had to talk about national security justifications for the EV ban. They say that the real national security issue is housing.
But if we are creating this “consensus” that the most important thing is to prepare for AI war with China, then do you think building more housing, hospitals, or public transit will be the top priority? Or would the real military issues be on top? I think the Palantir and the Andurils of the world will get way more access and resources compared to the people using national security justifications to fix immigration, housing, and healthcare.
China envy
Jasmine Sun (42:44)
Yeah, totally. Here’s the thing that I feel is frustrating and hypocritical: The whole logic is that China's really bad because they have an authoritarian surveillance state. So therefore, in order to combat that, they say we should build our own authoritarian surveillance state. Larry Ellison is talking about cameras and behavioral surveillance; Instagram took down the word “Democrats.” Elon Musk on X is banning people for posting the JD Vance dossier. On TikTok, people are saying “cute winter boots” instead of “ICE” to get around censorship. And I was like, this is the Winnie the Pooh shit! Americans are doing the exact same thing.
Oh my god, I feel so crazy. I thought the authoritarian surveillance was bad. I thought that was the whole point.
Tian Fang (43:31)
Yeah, the problem is their surveillance cameras are bad. We must build better surveillance cameras so we can trump theirs.
Jasmine Sun (43:36)
Our surveillance cameras will respect liberal values. They'll have an American flag on them. We'll do a land acknowledgement, then take you into custody.
Tian Fang (43:44)
The question that people need to ask—both the ambitious Chinese nationalists and the ambitious American nationalists—is, okay, say you've won this “AI war” (which I don't think is real). Say you've won the war: What kind of country do you want to be?
More people should ask themselves that question, because if you've won this technological competition but abandoned all the liberal values that made America America, you're left with nothing. You're left with alt-CCP in the United States of America. It’s no longer the same.
At least the Trump administration is being more honest about it. You know, I think they're very envious of the CCP. They wish they had the amount of resources and power that the Chinese government has. They're very honest about, look, we don't hate China because of the values. We hate China because it's China.
Jasmine Sun (45:01)
I think “envy” is totally the frame that people miss. They think it's fear, and so much of it is envy. Even the obsession with fast trains, and tall skyscrapers, and why doesn't San Francisco look like Chongqing? And how come there's no crime, and there's no homeless people on the streets? And they make EVs, and good AI. There's all these cracked STEM grads who want to 996 at the companies.
When I see that stuff, I'm like, I'm not sure that you fear those things. I think that you're jealous. You wish that we had that. The tech right is moving in that direction. Silicon Valley is realizing it is possible to have a much closer relationship with the government. You can perfectly align capitalistic, technological and strategic goals; and the state will pick you as a winner. They will help you with your $500 billion Stargate project, and in return, you will censor whatever they want and make sure never to say anything bad.
It's very concerning because—for you and I, obviously these are not the values that we want to live under.
Tian Fang (46:02)
The current administration and a lot of people in Silicon Valley seem to have silently admitted that the “China model” has won, right? The way technology is developed in China, the way that governance is conducted in China—that has succeeded. We just have to replicate the process in the United States.
They're no longer angry about the values that the Chinese government subscribes to. They just wish they were them. It's deeply unfortunate and sad for someone who left China to move to the US to see the same patterns here.
China thinks DARPA is great, too. This obsession with strong state R&D with unlimited power is present in both countries. Chinese researchers think it's how the US won during the post-war period in the Cold War. Americans think this is how China has won. Both countries think they're copying each other, but they are actually the same.
Jasmine Sun (47:08)
There's also a broader climate of American pessimism that contributes. No longer is there a sense that there's abundance and we want to be really generous; we want to give out foreign aid and uphold these values.
Policy and tech elite aside, people in general feel like their lives are bad. Everything's expensive. They go on Xiaohongshu and look at all of these pictures of wanghong cafes in Shanghai. These Gen Z people are like, I don't care if I can't talk about political issues if I can go to that beautiful cafe and have these super cheap fast fashion clothes. Or saying, Why does privacy matter? I'm just going to download all my data and mail it to the Chinese embassy.
I understand that impulse. But it feels like American pessimism creates this yearning for the China model. It’s related to an abandonment of Western liberal values, whether that's privacy or civil rights or free speech.
Tian Fang (48:09)
Yeah, I cynically do not think that the Chinese government wants this. I think China wants—in the words of political scientist Jessica Chen Weiss—it wants a world that's safe for autocracies, but it doesn't want a world of only autocracies. The Chinese government doesn't want itself to be a liberal government, but it's also fine with other countries being liberal and democratic.
I think the destruction of American institutions, and of democratic and liberal processes in the US, has done more harm to the country than the Chinese government has ever imagined.
Jasmine Sun (48:51)
One thing I've been curious about came out of a comment from
on my last Substack post. Yes, sure, America feels like it is becoming more of a techno-authoritarian state in the vein of what China has been doing. But are there ways in which these two countries’ approaches still diverge, even as the US shifts more illiberal?Tian Fang (49:20)
The Chinese government is way more concerned about corporate power than the US. We have so many tech oligarchs sitting in the White House—that is not happening in China, for good and bad reasons. But I do think in the PRC, there is still much more emphasis in the party's institutions and the party itself. They know that they cannot give it to outsiders. This has led to a lot of economic disasters, right? A lot of bad economic policy. But tech people in China are never going to get as much political power. That's a huge difference.
I also think that China is still more globalist in its outlook than the US. Now that the US has paused aid under Marco Rubio's State Department to almost all countries except for Israel and Egypt, China is still giving out aid. China is still giving out development funds to other countries. So we’ll probably see much stronger regional and international influence when it comes to Chinese technologies going abroad. That just makes more economic sense for most countries in the Global South. The US hasn't really made a true effort to compete.
The Falun Gong internet
Jasmine Sun (50:50)
There probably are some ways that Chinese tech will develop differently as a result of cultural and political differences.
For example, one interesting and funny example that you told me is how the Falun Gong (the group behind world-renowned Shen Yun Dance), has a surprisingly advanced technical capacity, which I was very surprised by. They have their own micro-ecosystem. Like you got US internet, you got Chinese internet, and then you got Falun Gong internet.
Tell me about Falun Gong internet. How did that happen?
Tian Fang (51:19)
So the Falun Gong is a sort of religious-spiritual organization that came out of the Qigong fever in China in 1990s. They have basically been exiled from China since the late 1990s, and they're currently based in upstate New York.
This is something I'm trying to write about in the next few months. The Falun Gong pushed for a lot of early VPN software and commercialized them so that they could be widely available. Chinese people used their software to access their religious websites because they're so vehemently anti-CCP.
In recent years, the Falun Gong has become very active in the US media space—like the Epoch Times was one of the most distributed newspapers in the country, and it's also run by the organization. I've been looking at these platforms, one of them being Ganjing World, which is a carbon copy of YouTube. It seems like they're trying to create this distribution network that YouTube doesn't have to take a cut from. A lot of the practitioners who moved to the US in the late 90s and early 2000s were software engineers, so I assume that's where they got their technical capacity. I'm also trying to see what’s going to come out now that the Epoch Times has entered a Department of Justice investigation, and Shen Yun performances have seen reports of child abuse. There were quite a few former performers who have said that they were deprived of basic freedoms while they were at the Dragon Springs, which is their headquarters.
But I'm very interested in how the Falun Gong always played a role in the Clinton-era push for global democracy via the internet, because they saw information access as a core part of their pitch to the Chinese public and to access more followers and to “combat the Chinese Communist Party.”
Even though think their influence has, in China, has declined very much over the past decade or so, its influence in the US has become much more pronounced. I was in Taipei, and there were people distributing Ganjing World flyers. It's interesting. I think it's the most successful network state to date.
Jasmine Sun (53:37)
Whoa. You should you should get Balaji on this.
Tian Fang (53:56)
Yeah, Balaji should really learn from the Falun Gong.
Jasmine Sun (54:00)
They're a great idea machine, you know? Yeah, we should be copying them. Maybe instead of the tech right.
Tian Fang (54:04)
Strong organizational capacity. It's also very CCP-like, the way that it is organized. So, why not?
Jasmine Sun (54:14)
The way that you describe Ganjing World and Epoch Times, it's like a Chinese or Falun Gong version of Rumble. There's some marginalized political group that gets banned from all the major platforms, so they have to stand up Rumble or Truth Social or the Daily Wire. And then that micro-network oftentimes gains its own critical mass; it becomes a very compelling product. And that actually helps them succeed.
I talked Kevin Roose a few years ago about how the American right got good at YouTube and social media because all the mainstream cable networks kicked them off. They had to get good at technology; they had to figure out the next form of media that they weren’t censored from yet. The right got good at that, and ended up winning the next generation of the media war.
Tian Fang (55:09)
That's super interesting. Even thinking about the role that the Falun Gong organization played in building VPNs for China, I think that was pretty advanced. This is like a late 90s venture, before China had a very robust internet censorship system. I'm very fascinated by this.
A lot of libertarians and left-wing anarchists are also very sympathetic to exiled communities, network states, and places with no laws. One example is some of the largest volumes in crypto trading are coming out of cyber scam centers in parts of Myanmar controlled by warlords and Chinese crime syndicates. These outposts and these neither-here-nor-there centers of technological production are very interesting to me.
Jasmine Sun (56:12)
Normally the last question I would ask is, what's a research question you're trying to entangle? But that's a very good research question. I'm very excited to read whatever comes out of your investigation.
Thank you so much for chatting with me, Tian. This has been really interesting, fun, and kind of depressing—but depressing in an accurate way.
Tian Fang (56:17)
Thank you for talking to me, I had a great time!
Links & books
From Counterculture to Cyberculture (Fred Turner, 2006)
Out of the Sandbox (Tianyu Fang, Feb. 2024)
The Cute Cat Theory of Censorship (Ethan Zuckerman, Mar. 2008)
The Rise of Techno-Nationalism (Tianyu Fang & Tim Hwang, Nov. 2023)
A world safe for autocracy? (Jessica Chen Weiss, Jul. 2019)
A Declaration of the Independence of Cyberspace (John Perry Barlow, Feb. 1996)
Have a good weekend,
Jasmine
I did my best to be descriptive rather than normative, but it’s surprising how many people took this as an endorsement of the tech right anyway. Since it apparently needs clarifying: I am not a right-winger, lol.
Share this post