China’s Endgame: ASI Timelines, US-China Relations, and the $1.7T AI Bubble With Alvin Graylin | #281

There seems to be a prevailing view around the campuses that ASI is, you know, 5 to 10 years out, maybe even 20 years. And so I'm really curious what the prevailing view is in China.

They are not behaving like they believe ASI is around the corner.

Why do you believe getting clarity and some resolution on the US-China AI race is so important right now?

Having a race condition forces people to make irrational decisions. At some point, we will get to a superintelligence type of a scenario. If and when we do, the concept of nations will probably become a lot less important than they are today. The AI sector alone is worth more than the GDP of America today. That to me is a sign that we are in a very, very fragile place and an economic correction is due.

What I hear you saying in your war game is sometime in the next 2 years, there's a private credit bubble that the US is using to finance its data center buildout. The bubble pops. And then the U.S. asks China to help financially in return for what a quid pro quo regarding Taiwan.

I will tell you this.

Now that's a moonshot, ladies and gentlemen.

Welcome to Moonshots, everyone, your number one podcast in all things AI and technology, your front row seat to the accelerating singularity. I'm here with my magnificent moonshot mates. The original four, including me, AWG, DB2, and Salim, my brilliant colleagues who every week help me, hopefully you, understand what's happening at this incredible rate of speed. I'm Peter Diamandis, your host and abundance entrepreneur and evangelist. Welcome to a very special episode of Moonshots today. Today, our mission is to go deep on China and AI and discuss it with someone who's lived and operated inside both the Chinese and U.S. technology. Ecosystems for over 35 years. Alvin, welcome.

Yes, it's great to be here. I watch your show all the time, so I'm glad to be able to chat with all of you.

We're going to quiz you then. I'll ask you trivia as we go.

When do we say drink? What's Alex's favorite term?

His Dyson swarm.

Dyson swarm.

Dyson swarm.

You got to build the Dyson swarm. Drink, drink, drink. Let me do a proper introduction of Alvin. So Alvin Wang Graylin is both a friend and someone who's held senior executive roles at HTC, Intel, IBM, Trend Micro, and WatchGuard Tech. He's worked in all 5 layers of the AI stack, the data centers, chips, PCs, phones, XR glasses, and apps. And he's built and founded over 4 startups. 2 years ago on this channel, we were discussing his recent book, Our Next Reality. He's now a digital fellow at the Stanford Institute for Human-Centered AI. A senior fellow at the Asia Society Policy Institute, professor of AI policy at University of Washington. He's currently supporting the US government and for the coming US-China AI safety dialogues. That's going to be happening on September 24th in DC. The very next day, on the 25th, we have our Moonshots Live event. He's lived and operated extensively in the US, China, and Taiwan.

Hi.

with dual master's degrees from MIT in computer science and business. Love having another MIT grad on the show here. And an electrical engineering degree from University of Washington. This is a conversation, guys, I've been waiting for for a long time to really go deep and understand what's going on. And Alvin, you're gonna bring a unique perspective on the US and the China AI race. You've argued that the game isn't a prisoner's dilemma, It's A Stag Hunt: A Game Theory Model About Coordination and Trust. Alvin is the author of 2 key papers we're going to link to in the show notes below, Beyond Rivalry and Misdiagnosing the US-China AI Race. Last week, he just released a new paper around AI security, The Biggest AI Models Are Not the Biggest Threats, where he proposes that it's the smaller AI models we need to be more worried about. And cooperation is the only path towards safety. Alvin just returned from speaking at the World AI Conference in Shanghai, where President Xi did the opening keynote. He's had a chance to meet with leadership across all of the Chinese AI labs, discuss AI governance issues with the senior Chinese regulators and policymakers. We'll hear about, you know, how they're thinking directly from Alvin. Today's pod is going to be far-ranging. We're going to be covering topics from China's open-weight models using AI for diplomacy, as well as robots and AI regulation. So very importantly, we'll be discussing the coming US-China AI safety dialogues. Want to understand what the objectives are, Alvin, and what you think might be accomplished. And we'll close with Alvin's recent Substack essay called Great Reckoning Before the Reconnecting. Love that. You and Alex both have wonderful terminologies and your essay on Abundancism. All right, um, so let's dive in. There's a lot to cover, Alvin. Um, you know, one of the things that we pride ourselves on the show is disclosing all of our connections. And since we're talking about a sensitive topic here on U.S.-China, uh, and you've spent 2 decades operating in China, and your bio lists roles like vice chairman of AVRA, the VR Industry Alliance endorsed by the Chinese Ministry of Industry and Information Tech, and a 3-year professorship at Beihang, which is a defense-linked university on the U.S. entities list. I want our audience to understand the full picture. So if you wouldn't mind, so folks can understand, you know, where you're coming from, if you'd walk us through those relationships, past and present, with any Chinese government bodies or government-linked institutions. Uh, what was your involvement? Uh, anything going on now? Were you compensated? I want to understand your connection to the Chinese government so people understand. Uh, you know, you are a U.S. citizen. Uh, people should know that. But I guess from perspective point of view, since we're going to be talking about a lot of, you know, very sensitive topics, uh, give us your background there, if you would.

Sure, absolutely. And you're right, I'm a US citizen for over 45 years. So I moved here when I was very young. I was born in China. But having worked and lived in China, you have to deal with the government on a daily basis because that's an important part of being functional there. The IVRA, Industry of VR Alliance, was an industry association that was endorsed by the government. And if you want it to be a functional organization, it has to be endorsed. And there was 300-plus members. about a third of them were international companies, companies like Nvidia and Samsung and Qualcomm and AMD and Google and so forth. So these are the kind of companies that they're trying to get into to accelerate the industry. When I was working there for HTC, who at the time was the head of the leading virtual reality, augmented reality company in the world, and I was kind of the head of the organization, the head of the company in China.

Right.

And by the way, HTC is a Taiwanese company. So at the time, I was working for a Taiwanese company. But we were given a lot of, I guess, a lot of influence because we were such an important party and important player in the industry. The Beihang University is a very large university. It's a university of aeronautics and aerospace. But it also was the leading university for virtual reality, augmented reality. And they've been teaching that for over 30 years. And in my role as head of HTC, which was a virtual reality company, they wanted me to teach there on a part-time basis. And neither of these positions were compensated. So, and, you know, I've— after I've left China in 2024, I've not had any involvement with either of them as well. So just to hopefully clarify where things are. But I think the thing to remember is that for you to understand and work with any industry, with any government, you have to understand both sides. And I think this is why having actually close discussions with them, having worked with them at the city level, province level, and some national level leadership there, you understand their mindset. And I think that's very important to actually have proper dialogue.

Alex, did you have any other question you wanted to ask?

Yeah, we'll get into it as we get into it. But this should certainly be an interesting discussion.

Yeah, for sure. We got to Alvin's professional bio, but his growing up bio is really interesting too. Why don't you tell us just about your childhood and growing up and then how you got to the States?

And it's a super cool story. That is a little bit strange. But I was born during the Cultural Revolution. So both my parents were artists. And my mother sent a letter to Mao Zedong's wife, because she had sent— she had closed the ballet school that my mother had helped co-found and was sent to be reeducated. And this is why I was actually born on a Chinese reeducation farm during the Cultural Revolution. So, you know, so Alex, I do understand some of the downsides of what happens in improperly organized or governed states. But I was able to move here in 1980 when My uncle, my actually granduncle, helped to sponsor me. My grandmother was a reporter for the New York Tribune back during the Sino-Japanese War. And she was there reporting but had to leave my mom behind when the Japanese bombed Pearl Harbor. And it took her 8 months to get over from Shanghai to Chongqing, where the Flying Tigers were. And so, you know, so this is why I was born there. And this is why, you know, I'm actually part Ashkenazi, part Scottish, and part Chinese. So it's a little bit strange for my generation, actually.

Yeah, amazing. And it's worth noting your brother is, I guess, the first nuclear sub commander in the US Navy?

Yeah, yeah. So he was a senior officer in a nuclear submarine, actually. So one of the big Boomer ones that actually, you know, has the capability to destroy nations. But, you know, and in fact, his 3— 3 of his 4 children are now active officers in the US Navy and went to the Navy Academy in Annapolis.

So this episode is sponsored by Google for Startups. Think about this for a second. You now have access to the same generative AI models that cost hundreds of millions of dollars to train. Google Startup Technical Guide for Generative Media gives you a complete blueprint for deploying Google DeepMind's models in production. Images, video, audio, all of it. Real architecture, real results. Find the link in the show notes below. Yeah, we talk on this pod a lot about the fact that the US-China AI race is driving a lot of what's going on. This is a lot of important policy happening, and it's in the same way that the US-Soviet moon race drove the Apollo program. We keep on falling back to the reason for racing in the US on the AI model development is to make sure that we've got, you know, we get to ASI before anybody else. And that's been coloring everything. So I want to get into that on this program. It's important for everyone listening to understand the backdrop of this. So let's kick it off. We've talked a lot about open weight models over the last few weeks on this pod. And Alvin, in your essay, Misdiagnosing the US-China AI Race, you lay out a staggering shift, right? Chinese open weight models climbed from 2% to 61% of open router traffic in the last 2 years. Alibaba's Quen model alone has over 700 million downloads. I'm sure the number's surpassed since I looked.

It's up to a billion now.

So yeah, there we go. Um, uh, with 180,000 derivative models, and one of the things again we talk about is, you know, when you have an open weight model, it's very easy to fork it and develop it and retrain it yourself. You see that the US export controls on Anthropic's Fable-5 backfired spectacularly, specifically that, uh, within 24 hours, you know, China's Z.AI released GLM 5.2. under MIT license. In fact, that was the model used to deal with the Hugging Face debacle.

Yeah.

Brazil's Rio built on Quen. Japanese Sakana released Fugu Ultra. You say that the US denial strategy didn't slow China at all. It accelerated the innovation and alternative ecosystems. And you call this sort of a denial keeps us ahead fallacy. All right. So let me throw out the first question. Why do you believe getting clarity and some resolution on the US-China AI race is so important right now?

Yeah. So in fact, this is probably one of the most critical questions that we need to get resolved because it's— having a race condition forces people to make irrational decisions. And right now, there is a perceived race condition. And it's based on some assumptions that I think are actually misguided or maybe misunderstood, is that there is a perception that there is a finish line. There's a perception that the world is zero-sum. There is a perception that whoever gets to AI first or AGI first can somehow rule the world forever. And this is a certain segment of the policymaking circles have this belief, and a certain portion of Silicon Valley elites have this type of belief. So I think that narrative forces the whole discussion into a national security issue, when the reality is right now, none of those assumptions are really based on real data today. We don't know— well, first of all, we do know that the world is not zero-sum. I mean, as you guys talk about every day, every episode, the world's getting better. We're getting more resources. We're getting more abundance. So it is not a zero-sum game. And there is no clear finish line, right? In the sense of, as these technologies get better, they're progressing. And as you mentioned with all of these open-source models, when the difference is— gap is moving from a year and a half to now probably 2 or 3 months gap between the open-source and the closed-source models, There is no finish line where you say, okay, we've won. It's not like the space race. The space race, you say, hey, we've landed on the moon, we've won, there is an end. Whereas an arms race type model that we are in today is a constant spend and a constant pursuit without clear value being returned. But I don't want to take too much of this. But I think this will help set up the conversation that we're having.

Yeah, maybe just to pull on the thesis, Alvin, I think that's latent that there isn't an endgame. I'd love to understand how you think about this. From my perspective, there's an obvious endgame. There's space, there's development of the solar system, there's interstellar exploration, all of which I expect to be fulsomely and holistically supported by superintelligence. Surely somewhere among the various scenarios that I assume you're analyzing, there are scientific and engineering endgames, quote unquote, that are intrinsically valuable to pursue. Is it your thinking that science and engineering and solving everything, as it were, is not the endgame? Is there some other non-endgame that you have in mind? Do you think that this is sort of a Red Queen-type scenario where intelligence is just sort of endlessly racing as an end to itself? Or is there, is there an honest-to-goodness, uh, afterwards, after the singularity, in your mind as it pertains to US v. China?

Yeah, so, so I think, um, what you're describing in terms of at some point we will get to a, a, uh, you know, superintelligence type of a scenario, um, that may be. But if, if we— if and when we do, the concept of nations will probably become a lot less important than, than they are today. The reason that we've created nations, created— it actually started with city-states, right?— is because we wanted to protect a certain level of resources. We wanted to defend and gain additional resources. In a world where we actually do achieve ASI and we get the kind of abundance that Peter and all of you have been talking about, for ages, then the need to have separate nations with these type of competition really would not exist. If it still existed, then we would probably have destroyed ourselves.

So just to make sure then I understand your thinking on this, am I understanding correctly that your worldview is basically Fulsomely developed superintelligence naturally yields to world government on the one hand, and everything else post-superintelligence, as it were, being solved. And those two, world government and post-superintelligence, are inextricably linked.

Um, I, I think that they're, they're at, at the point, if we can get an aligned peaceful superintelligence, then we will naturally move to a more world government, maybe a galaxy government type of a model. But it is not something that I think is imminent. And it is not something that will happen without some level of turmoil. So I think the important part is about how do we get there. I agree with you in terms of where the long-term goal is going. But right now, I don't think in the next 2 or 3 or 5 years, which is what we're really racing against— we're trying to build— we're building 10 times more data centers than China is, you know, and to be 1 or 2 or 3 months ahead, um, it's not clear that the value is actually there.

And what is— just a quick follow-up question— what is your perceived timeline for world government and post-superintelligence? You mentioned it's not 2 to 5 years. Is it 10 years? What's the timeline?

I mean, I think the whole evolution of this is going to take probably on the order of decades, maybe, you know, by the end of this century. I think we will get to a point— and in fact, I think we need to move at a pace that the world can adapt. And trying to move too quickly actually creates a lot of instability in the world. If you look at the prior industrial revolutions, it was 80, 60, and 40 years respectively in playing out. And, you know, we're talking about this revolution going from where we are today to the singularity in 5 years, according to some folks. That is not a speed that the world can adapt to, even though—

Digest.

Digest, yeah. And, when that happens, turmoil happens, and it creates instability, and it may actually move the civilization backwards.

Look, I think a couple of comments here. One is, I think one, uh, endpoint that has turned this into an arms race is the idea we may achieve ASI, and then you have, uh, one party is uncatchable because they've got so much recursive self-improvement going on. And that has turned this into an arms race, whereas in reality this whole thing is a platform race, right? And so I think this the point that Alvin makes very appropriately in his commentary. I think the second mother— the elephant in the room that we've just touched on here—

The mother of the elephant.

It's the fact that we're running the world on an architecture of 17th century nation states, and we're trying to run 21st century applications on that 17th century operating system, and it's simply not going to work. A huge chunk of the issues that we see in the world are that fundamental problem.

Yeah, I mean, I would put forward the notion the biggest concern is whether a Chinese authoritarian, uh, level of AI enablement, uh, drives other nations to have to take on that political structure, right? The U.S. prides itself on freedom, on, um, on privacy— if we have privacy, that's a different subject— uh, and And the question, you know, I think the battle here is the US wants to continue its form of government, its form of democracy, and not be challenged by an ASI out of China. I think that's ultimately the bottom line.

I think that's a good way of putting it.

Yeah, but from that perspective, I think, you know, US and China are actually very aligned. Neither one wants to have an ASI that comes out of nowhere and destroys the system that is available today.

Right.

So I think there is a common shared interest, and that usually shared interest is how cooperation dialogue begins.

So it's going to come out of Zimbabwe and it's going to be really ugly.

It's a possibility now, actually. The RSI is popping up everywhere. You know, curious though, Alvin, if you said, look, there seems to be a prevailing view around the campuses that ASI is, you know, 5 to 10 years out, maybe even 20 years. And then around San Francisco and around the big labs, it's like, look, 1, 2 years maybe. And every year that goes by, they reel it in. For 30 or 40 years that I've been working in AI, every year it goes back a year. Now, it's getting reeled in every single year. And so, I'm really curious what the prevailing view is in China. Does the bulk of China, either the population or the government, Really believe it's 10 years, 15 years in the future, and we have time to just twiddle our thumbs and think about it?

Yeah. So, I think the timeline issue is probably one of the biggest disagreements between both these countries, as well as between, I think, the average person and maybe some of the folks that are in Silicon Valley. But if you look at the behavior of how the Chinese government is operating, They are not behaving like they believe that ASI is around the corner. If they did, they would not be telling their labs, don't buy the H200s that the Americans are giving them. They would not be putting out regulation that is slowing them down, which they've had regulations around AI privacy, data provenance, marking transparency and marking in public, child addiction, anthropomorphic Compromising AI. All these regulations have been around, and every single model that is released in China has to be reviewed by the CAC, the Cyberspace Administration of China, which again delays it by weeks or months. So they're seeing this as something that is akin to other technologies that has happened. And they understand that general-purpose technology usually takes, even when it's and mature, it takes years, if not decades, to actually diffuse into society. And they're behaving like that. So I think there's definitely a difference between maybe Beijing and DC in terms of how they're looking at it. One thing I will say that there are probably 2 or 3 labs in China that are a little bit AGI-pilled, not maybe to the level of the Silicon Valley folks. But their goal is very idealistic and aspirational to say, hey, we also want to create AGI. But in general, the majority of Chinese labs, as well as Chinese regulators, see this as a technology that is not just like other technologies.

Maybe let me pull on that a little bit, Alvin. So what I think I hear you saying is the Chinese Communist Party leadership has not Yeah. Perhaps woken up, assuming you believe the premise that we're in the middle of a singularity and that recursive self-improvement is already here. Perhaps the CCP has not yet fully woken up to that possibility. What do you think it would take? What technical development, what geopolitical development would it take, assuming that premise is correct, for the CCP to wake up and say, oh my goodness, we need to treat this as a national emergency in order to compete for recursive self-improvement, throw all of these regulatory speed bumps, the— we've talked about it on the pod in the past, the— you alluded to reeducation camps. It's been widely reported that China makes all of their own labs' frontier models pass certain ideological tests before they can be released. What would it take for the Chinese government to say, throw caution to the wind, In order to compete, we have to just pick whatever cliché you want. We have to go at the speed of light to compete with American recursive self-improvement. What would it take?

Well, first of all, maybe I'm probably on a slightly different timeline than you in terms of—

I know, I know. Most of the world is on a different timeline from Alexei.

As you guys know, last week you had the Pacing the Frontier letter that came out of all the engineers and lab heads. This is something that I think the industry should be looking at in terms of managing to actually slow it down in the sense of, if it goes too fast, as I said, the world takes time to adapt. I'm glad that actually more than 1,000 people in the industry in the States are actually looking at this. I don't think it's necessarily a thing that they don't know about these concepts of ASI and RSI. They understand this stuff. And there are actually multiple safety institutes and an AI safety contingent that is in China telling these stories to the regulators. And they hear it. In fact, at the World AI Conference, there was multiple discussions and forums specifically around AI safety. And there was people from the US, the Benjios and the Techmarks were also there to add these types of points to the agenda. So I don't think it's that they don't know it. I think it's that they don't believe that it is something that is necessarily and should necessarily be a nation versus nation issue. In fact, I think it's on the agenda to talk about these, the World AI Cooperation Organization, which they had announced on the first day of the the World AI Conference, which is their version of Pax Silica. But Pax Silica from the US that was launched at the end of last year was a US-led organization that was talking about how do we keep US dominance and leadership in AI and which allies are we going to pick to be on our side. So it creates a bloc to say, we want to be the winning bloc. Whereas what they announced was to say, hey, look, we want to create a global organization, make AI a public good, make it shared, and everybody shares in the benefit. Anybody that wants to join can join. And, you know, they're going to put out thousands of training centers and training facilities, compute facilities, resources to the members that are joining. And I think 29 countries joined. You know, there's about 25 in the PAC-SILICA. So it's creating 2 blocks. But so here's something that is interesting is I actually talked to one of the people that was involved in organizing this. And I said, look, wouldn't it be good if you actually invited the US to join? They're like, oh, no, it would be amazing if the US would join. We would want them to join. And in fact, they should join. And I said, well, but if they join, you can't call it WECO because that's a Chinese-led organization. They're like, oh, if you guys are interested, we would be open to changing the name. We would be open to having a truly global organization. So I think this narrative of us and them and they're trying to take over the world with their AI, I don't really see that.

Maybe at the risk of belaboring the point, I just want to press again on what my question was, which is what technical threshold or event would it take for putting aside geopolitical competition Put that aside for a minute. What would it take for superintelligence to actually cause the CCP to say, we have to actually abandon all of our internal regulations intended presumably to maintain social stability and the supremacy of the existing regime, as well as external efforts to create blocks? Put the blocks aside for a minute. What threshold of superintelligence, either achievement or technical development, or maybe implications for weapons systems, if that's really what it takes, what technical achievement would superintelligence have to pass, or what threshold would it have to achieve in order for the CCP to decide, in your mental model, gosh, we really just have to focus on supremacy here?

Alvin, if you don't mind, let me intercept and lead into that question too, because I think we really do need to answer Alex's question. But before we can do that, let's understand who the CCP is. In the US, it's really interesting when you meet the actual players. So, you've got Elon Musk, Demis Hassabis, Sam Altman, all saying, God, I wish this would slow down. And when I interviewed Sam at MIT back in 2020, remember that?

Yeah.

He was like, it would be far better for the world if progress was slower, but it just isn't. And so, we have to just live within the reality that AGI is is imminent and do the best we can. So then when you see them interact, these are young, very, very smart people. And they go to the White House and they interact with really old people who have no idea what AGI even stands for. And that's the dynamic. And when you meet them individually, you realize, wow, these are just regular everyday people in the hot seat. And you interact and you see how they communicate and it changes the future of the world. So then I envision the CCP and I picture people in their 70s kind of up on a hill, completely disconnected with the details. But maybe that's wrong. I have no idea. What is the CCP, first of all?

OK, so first, I want to maybe demystify something around this idea that people think that China has a CCP. There's somebody at the top that just says, you will make AGI, and it just happens. And the reality is that with all the industries and with all of the innovation that's happened in that country over the last 30 or 40 years, It was never a top-down thing. There was maybe directional things. They would say, hey, for the next 5 years, we should work on clean energy and we should work on automation of robotics and we should add AI to that. And they did that about 5 or 10 years ago. And when the central party initiates these plans, then the provinces say, hey, look, what companies are we looking at? Do we have in our area that supports this particular higher-level goal? And maybe let's go find them, support them, give them some, stipends, give them free recruiting, give them some kind of benefits. And they will have essentially provincial champions and city champions. And then the 30-plus provinces all compete against each other to see who can make companies that solve some of these problems. So it is actually very distributed in terms of how they're how these plans get initiated. Nobody's saying, okay, you need to use this technique to go do that. And you need to share your resources. They're actually a very highly competitive landscape between all of these labs. But the thing also to remember is that almost every single leadership in the senior leadership of the central party are actually engineers, probably 80 or 90% of them, right? So they're actually quite technical.

Um, yeah, I think that's one of the biggest challenges. You know, when I, I had gone to, to different parts of China, I used to take a group of Abundance members there all the time every year, and we meet with the top companies. And, um, we had a presentation from, from the, uh, the CCP leadership, uh, and the thing that was most striking is in the U.S., most of our politicians are lawyers, and in China, most of the politicians are engineers. I found that a fascinating distinction.

You know what else is fascinating? My son just got back from China, and he talked to a whole bunch of entrepreneurs, that distributed network you're talking about, Alvin. He asked them, what's the most important thing to entrepreneurial success? He expected teamwork or business plan. He said, no, it's what the government's focus is next that determines your success.

Yeah. When you're swimming, you don't want to swim upstream. And what the government does is that it makes the stream flow in the directions of the certain areas that they think are important. And then they let the entrepreneurial nature of the people there— and, you know, 1.4 billion people and the most number of STEM grads in the world— good things happen. Right. And that's what's driving their innovation and their ideas. Look, when these things happen, it'll grow our industry. It makes us more resilient as a country. And, you know, it also brings the quality of life So having said that, and we started this conversation on open models, again, I think it's very important to understand this.

Is the government saying get as many open models as good as they are out there? Is that direction coming from the government? Is that popping up from the entrepreneurs saying we can distinguish ourselves from US labs by creating open weight models?

Yeah. So the whole open source strategy, people think, oh, this is Chinese strategy to destroy American economics and power. pop this bubble. The reality is that it's an emergent strategy, right? In fact, I was talking to friends at DeepSeek a little bit after they came out. Before that, nobody knew who they were, right? They were not on the radar. They were not funded by the government. Nobody told them to open source. But the CEO of the company was very open source-minded. He thought that, hey, open sourcing is something that I should do because this is a great technology. I want to share it with everyone. In fact, when they first did that, they got their hands slapped because the government's like, hey, this is such a great model, why are you open sourcing it? But because of all of the, I guess, soft power value, the PR value that came out from having a local champion, they then became celebrated. And then essentially most of the companies in China were following this because it became kind of the de facto emerging standard. And now, you know, at the last Waco or the WAIC announcement, Xi Jinping finally said, hey, we think open source is a good strategy. And that kind of goes to what Dave's talking about, right? In terms of when he says that, now pretty much most new companies are going to be focused on open source because that's the high-level instruction.

If DeepSeek had been a closed model, Do you think China would have gone that direction? Is it really just that Seed led to this incredible open source movement in China?

Well, I mean, there were open and closed models for the whole time, right? In fact, if you look at ByteDance, they have the Douyin model, which is a closed model. And their SeaDance model is a closed model. And they're also quite successful. So both models exist. But you're right. I don't know what would have happened. I don't know if the push for open source would have been as great. But the one thing that we also need to remember is that open source was a little bit of a necessity that U.S. policies pushed on them, right? You had mentioned the export controls earlier. And export controls— you know, you first talked about export controls and the software export controls of Fable. But actually before that, for several years now, 4 or 5 years, there's been export controls on chips and allocation of EDA software and lithography equipment and so forth. And what that's done is that it's forced them not to have the latest and greatest equipment. It's forced them to have to innovate with low resources. And that essentially pushed DeepSeek and all these other companies to get more innovative. Whereas the U.S., because they have so much resources, they've been much more focused on brute force and brute scaling, whereas the Chinese have not. And open source was necessary for them because by open sourcing, now rather than having a lab with 100 or 200 people, when you open source it, you know, you were saying that it's been 100,000 or more of the Qwen variants. Essentially, the rest of the world helps you modify and improve your models, right? And that's a great way to leverage the global community of millions of AI researchers. The other thing that's truly important is inference. For you to do inference, you need to have compute. And if the Chinese labs and the Chinese hyperscalers can't buy the compute, then by open sourcing it, essentially all the hyperscalers and neo-clouds around the world are buying compute, hosting these AI models, and allows them to distribute their models without the high CapEx that that the US labs are burdened with.

Hold on, Alvin. I think we need to get back to Alex's question, and I'd love to drill in. I think we have an opportunity here where you have firsthand knowledge from friends at DeepSeek around, I think, what is going to turn out to be one of the most pivotal moments in human history, the decision where DeepSeek comes out, open sources a frontier-level model, Opus 4.8 caliber, and Xi Jinping says, if this is so great, why are you open sourcing? Slap your hand. Then something happens that flips his opinion. They become global news. Their valuation goes through the roof. And some aura of, wow, this is good for China, gets back to Xi Jinping. And he says, open source is now a blessed thing. And then immediately after that, Quen is out, and then Kimi K3. And I think the release of Kimi K3 will turn out to be as defining a moment in human history as anything that's ever happened. Yeah. That's my prediction. But I think the psychology behind that choice is going to be the news nugget that matters for all time now, and leads into Alex's question of what would it take for a wake-up call? If it turned out that was a colossal error, what event would have to happen? And I'm not saying that's the case. I know the opinion in China is that that's not the case. But walk me through any detail you've got on the psychology that changed Xi Jinping's opinion. and on whether to open source these things?

So I think there's 2 questions. One is, are you thinking about, are they going to close source because they're worried about AI running away and becoming rogue? Or are you worried about competing with the US and saying that whoever controls and creates the AGI becomes the global dominant hegemon? So which aspect do you want to—

Maybe let me pull on that a bit, because in my mental model, which you can perhaps help me to refine, there are two different separable concerns by the CCP. One is retaining CCP control and dominance within China on one hand, and on the other hand, it's maintaining competitiveness and peaceful rise and Xi Jinping thought on a global stage and Belt Road Initiative and geopolitical competition with a Western bloc on the other. And these two different arms may be in competition with each other. The CCP at some point as superintelligent. capabilities continue to increase, may be forced to decide whether it prefers either retaining domestic control on the one hand or seeking to continue to rise geopolitically on a global stage. How do you think about that?

And then I'll answer the other half of it after you're done with this.

So, you know, I don't really see them right now seeing I don't see them using AI as a way to create political domination. I see them as looking at this technology to increase their economic influence around the world. That I think is absolutely there.

Is that because you think— again, just to pull on that, because there's a hidden premise in that. If you're thinking that CCP doesn't see political domination through AI, is that because CCP Do you think that the CCP already has domestic political domination, has already achieved dominance over AI through these reported ideological exams that AI models have to go through?

No, I think those are 2 separate issues, right? The type of things that the CAC has them review is things like removing certain types of keywords or ideology or things like that. Those types of adjustments in the models only apply to Chinese-hosted models. So if a model coming from these labs are then put on Hugging Face, all of those types of guardrails are actually removed in terms of whether or not they can talk about— Tiananmen Square and so forth.

Tiananmen Square, 1989. 1989. Let's just say it.

Exactly. I mean, but to be honest, that's because everybody already knows this. Nobody really cares, but they do it more for formality. Right? But they actually do have other things that they're putting in place in terms of checking for, you know, is, you know, there's probably a slightly less security-mindedness in terms of how much it refuses to answer questions related to maybe, you know, viruses or medical or other things. But I think there are still definitely those safeguards they are putting in place and those are getting added more and more every day because of these kinds of issues. What I think would get them to be really concerned, if they start to see the US using this model as a weapon, as an aggressive kind of offensive tool, right? Because then it becomes, okay, do we want to, you know, like, you know, what the Mythos models were essentially held back to say, hey, here's a model that can be done. And I think for some right reasons, you want to neuter the offensive capabilities before you put it out to the rest of the world. So this is actually a good thing. But in some cases, I would actually think that it would make sense rather than having 50 companies that are being allowed is actually to allow most government organizations to have access to this because you really want global stability. Global stability means that countries can have access to it, find the vulnerabilities in their systems. Because I don't think the Chinese want the American financial system to go down, and the Americans don't want the Chinese financial and their grid to go down. Because when instability happens in any big country, the world suffers. Smart people understand this. But too many, I think, folks with relatively narrow perspectives think that one country wants to actually have another country fail. Having major superpowers fail creates irrational actions, and societal stability usually is the preeminent priority of most major governments.

This episode is brought to you by Blitzy, autonomous software development with infinite code context. Blitzy uses thousands of specialized AI agents that think for hours to understand enterprise-scale codebases with millions of lines of code. Engineers start every development sprint with the Blitzy platform, bringing in their development requirements. The Blitzy platform provides a plan Then generates and pre-compiles code for each task. Blitzy delivers 80% or more of the development work autonomously while providing a guide for the final 20% of human development work required to complete the sprint. Enterprises are achieving a 5x engineering velocity increase when incorporating Blitzy as their pre-IDE development tool, pairing it with their coding copilot of choice to bring an AI-native Ready to 5x your engineering velocity? Visit blitzee.com to schedule a demo and start building with Blitzee today.

Alvin, I want to pull on 2 strings on this topic before we move on. The first is the claims that Kemi K3 and other models were distilled from US closed models. Your thoughts, what, you know, what is being said in China about that? And then the second is the policies limiting chips and limiting access to Fable 5. Your belief is those policies were misdirected. And can you explain why on that?

Sure. So the distillation thing, I think it's more of a PR tool that certain companies are using. And really right now, there's only one company that is against that. If you look at the numbers—

Which company? You're saying OpenAI?

No, Anthropic.

Anthropic.

Okay.

Yeah. I mean, they're the ones that are lobbying the government to say, hey, we're being distillation attacked, which is actually a word that they invented. The reality is that every lab, both domestic and international, distills from each other. And they distill within the organizations themselves from larger to smaller ones. And in fact, if you look at the numbers of what Anthropic put out, they were saying that, you know, 20,000 accounts from 3 different labs in China and, you know, a million or 2 questions that came in. I went back and actually did an estimate of what it would cost to do the number of queries based on the average responses on their highest models. And it was like $2 or $3 million. So it was $2 or $3 million across 3 different labs. And for DeepSeek, I think it was only in the thousands of dollars. So the numbers sound big when you look at them in isolation, but when you look at them in aggregate, it really doesn't mean much. If you have a model that you spend $1 billion on and somebody can distill and duplicate with a couple million dollars, then the whole economics of frontier AI doesn't make sense.

Yeah. sense.

Now, here's the thing that also, just to give comparison, every month Meta spends somewhere between $100 and $200 million on Anthropic tokens. They're one of the biggest buyers of AI from Anthropic. And if anybody was going to distill, they would have been distilling for the last year to 2 years. And they just finally got a model out that is somewhat competitive in the last last week or two. They have the highest per capita payroll of any lab in the world. They have some of the most number of compute, and they have the most number of tokens that they're buying from Anthropic. Why can't they have gotten a KIMI K3 thing out 6 months ago?

The public argument— I mean, this has been widely reported— that Meta internally is utterly paranoid of being accused distilling Anthropic traces and actively encouraging their engineers, don't use it, don't overuse it, don't you dare allow any Anthropic reasoning traces into the development of the Muse series. They're utterly afraid of being sued by Anthropic for reasoning trace distillation.

I'm not sure if I agree with that, given how many, you know, how much they're spending. But here, let me give you an example. What about xAI, right? They also have access to these, you know, and I don't think Elon has that same same concerns over doing anything to speed himself up. And only in the last few weeks have they come out with something that is relatively competitive.

Elon is an interesting case because he's, at this point, I would argue, a frenemy of Anthropic. His entity SpaceX AI acquired Cursor, and Cursor was arguably, at least recently, a post-trained version of Kimmy that was being post-trained off of reasoning traces via Cursor that were being, in many cases, siphoned off from interaction with Claude. So Elon has— I think the steelman case for Elon and the Grok series, the recent Grok models, is in some sense he has like 2 layers of plausible deniability, but he's basically doing the same thing that the Chinese labs are being accused of.

I can also tell you from firsthand experience, Yeah. I can tell you that the people working on it at the time at xAI and on Meta are nowhere near as good as the Chinese people that were working on it at DeepSeek, Quen, and KIMI. And I don't know why that's the case. The really great people in America working on it are Anthropic, some of the Google people who have since left, OpenAI, but not the Meta or xAI team at the time. Now, they've changed teams completely. They've fired everybody and started over. So for whatever reason, the Chinese people working on it, though, are brilliant and far, far better than those teams were.

Yeah. And I think that's the key to realize. If you look at all the papers that are coming, like half the papers around AI are coming out from Chinese organizations, right? And they are actually innovating. It's not that distillation was why they're successful. If you look at how they were able to reduce their KV cache usage by like 20x, these are not things that you get from distillation. You cannot distill something from somebody that that other people didn't have. So we can't take that away from them. As Dave said, they're smart people there. They're doing innovative things. And that's part of the reasons why they're successful. Did they distill? Probably. I'm sure they have. But same, the US models have also distilled from Chinese. I think there was a couple of models ago where if you use Chinese to ask Anthropic Claude what model you are, and it said, I'm Quent.

Right.

And when I was at the— at the Alibaba Labs. And they were kind of laughing about that, too. They're like, yeah, they're distilling from us. We just can't tell because they already downloaded our models. So we don't know.

Right.

But I don't think that the whole distillation issue is really as big. In fact, if you look at what happened last week with Zuck, he's actually saying distillation is actually a good thing. We don't think distillation should be prevented. And they're kind of jumping on the whole open source thing. And they're saying, hey, AI should be free, right? So, so, um, anyways, I totally agree, by the way.

I totally agree. I think that the future of AI, Accelerando style, Alex Wiesner-Gross style, the past AI is always going to help you create the next AI. That's the inevitable outcome.

It's what we humans do as well. Humans help, you know, create the next generation.

So I think the whole thought traces thing is overblown. I really want to put a pin in one thing Alvin said, which I think is critical and absolutely true. If a frontier lab spends $1 billion getting to the next level, the next guy trying to distill from there and get to that same level completely separately is about $2, maybe more like $10 million to get to that same level. And Alvin said, like, this is just a fatally broken business model. I'd love to put a pin in that statement because I totally, totally agree. And this is why Elon is racing after hardware. Because the sustainable mode of the future is at that level, not at the— because anyone can do exactly what Alvin just said.

Salim?

Alvin, you said something I want to pull on to, which seems to be the theme of today's episode.

We're pulling on elephants in the room.

We're pulling on strings all over the place.

Others have elephants everywhere.

You said the US is building a lot more data centers than China is, and I'm finding that very, very surprising. I think maybe there's a difference here. It's clear China is building a massive energy capabilities, but they haven't built a data center layer yet, I'm guessing is what you're saying, versus the US is the other way around. Could you expand on that? Because I find that surprising. Shouldn't they be building a ton of data centers?

No. So they are spending a ton more on energy generation. They're building more new energy, new electric generation than the rest of the world combined, about 10x what the US is every year. Now, what they're doing is actually they're trying to electrify their society. That's their focus. Because right now, 40% of their oil is imported. And because of what's happening, like with the Hormuz issue, they realize, hey, it's really good that now essentially half of our auto fleet out there is electrified. So I don't need to depend on imported oil. In fact, that was one of their key objectives, was to say, how do we become independent of external energy sources? Now, what they are doing, though, they are building giant solar farms and wind farms on the west side and in the desert parts of China, and then using their very high-voltage power transmission that essentially 1,000 miles, you lose less than 1% of the electricity when you use these high-voltage things. And they bring the energy to the east, where most of the population is living. Right. But, they are also building some data centers right where the power generation is happening so that you don't have stranded power. They're able to deliver compute at a fraction of the cost of the US because their energy cost is around $0.02 to $0.03 per kilowatt-hour, which is, in some cases, 10 or maybe 15 times cheaper than many parts of the US. And I think in the long run, that's actually where the constraint will be. And I think you're right. Right now, they don't have enough chips. They can't buy enough chips, and they can't make enough chips because their capacity is limited by the fact that they don't have EUV machines. So when I was in China, in every lab I talked to, I said, hey, do you guys have a new computer? This is our biggest issue. We don't have compute. So from an export control, are we slowing down China? Yeah, I think that export controls of chips is slowing down China. Now, the one thing that most people don't realize is that the actual training right now that is happening is not even happening in China because they don't have the Blackwell generation chips in China. And so they're actually doing it in international data centers, training it. it and then bring them back on a disk or something. So it's— Wait, say that again?

He's checking luggage.

That's incredibly important information.

Oh, yeah. I mean, well, this is actually not a secret. I mean, I think people in both sides—

This has been widely reported.

Yeah, and both sides—

What's the point of a chip embargo? What is the purpose of an embargo?

Well, I think it's more optics than anything right Oh my God. But from an inference perspective, most of the inference is being served to Chinese people in China. And that is an area where the limitations of resources is slowing them in terms of how many new users they can add and so forth. I think you guys alluded to some of that in your prior episodes. So the export controls, yes, it has slowed down China, it's made their life more difficult, but it also has created the necessity for innovation. So back to what Peter was asking earlier, was export controls good or bad? I will tell you this. The day, or maybe like the week after we stopped the Chinese from buying all of the high-end chips from the US, I had calls with a few of my friends who are in the semiconductor industry. And every one of them got calls from the government saying, hey, would you like some extra funding? Would you like extra resources? How can we help you accelerate? Could we get you customers? I know a few of the CEOs of these Chinese GPU companies. They were saying, nobody wanted to buy our stuff. We're 2 or 3 generations behind. We were less energy-efficient. But now, we can't make enough because every data center in China has to buy our stuff. And so we would have died if it wasn't for American policies. So essentially, we created the current competition of all of these— the Marthas and Cambricons and Beerens of China would not be as successful, or maybe would have gone bankrupt, if it wasn't because of American policies. And now within the next 2 or 3 years, they will start to catch up to what America is doing, and they will start to export their chips. And that would not have been the case if it wasn't for us forcing them to survive.

Well, what a back-to-back double whammy that is, though. I mean, we knew the chip embargo was misguided, and it's going to be one of many government misguided things in the next couple of years. But the idea that, okay, first it forced China to create its own internal successful chip industry, and second, the training moved offshore anyway. So it didn't slow down one iota the training, Because I think what the government didn't realize is if you embargo ASML machines, then they can't build the fabs. And the fabs are like physically on the turf. You can't just port it to another server overnight. But the training is just a job, and it can move to a server in Hong Kong or to Taiwan or to Europe instantaneously. The file that comes back is just about 3 terabytes, and you just transmit it back in an hour.

Yeah.

And so that moves all over the world like a liquid. It's everywhere.

Exactly.

Instantaneously. So the chip embargo has completely and utterly backfired and misguided.

Yeah. And unfortunately, right now, if you talk to the folks in D.C., they're doubling down on this right now. They want to keep adding and making these things more difficult. How can we get the foreign, the international data centers that are being used by the Chinese to not be accessible? And they're adding more and more layers and more KYC. And I think that those are the kind of things That actually will backfire more economically— more geopolitically than economically, because then that forces irrational behaviors. Like, shit, they're now trying to keep us from progressing as a nation. Then behavior gets more aggressive and they become more defensive. And I think these 3—

I think, Alvin, I think we agree on almost everything except the timeline to AGI. And I really want to get back to Alex's question of, if it turns out The Kimi K3 level or one model later is capable of full RSI and then spirals to the singularity. Just hypothetically, in that scenario, Alex's question is, what wake-up call would it take to get back to Xi Jinping to say, oh, wait, I was wrong, it's not 10 or 20 years out, and we've made a horrible mistake here if we release this next thing to the world?

I think if they start— there are credible multiple labs coming back, safety labs coming back with testing to show that these AI systems have their intent, once they've released, to do things beyond what they were instructed to. I mean, you guys have been talking about these rogue AI escapes, but they weren't really rogue AI escapes. They were instructed to escape. And they were put into a prison to say, how can you escape?

They were incentivized to escape, I would say. They were incentivized to go and find the answer.

And they were saying, use whatever tools you have. And by the way, they also that had left open doors for these things to escape because of improper settings, or maybe some of them were intentionally leaving holes for them to find. They were given tasks that were impossible to solve unless they escaped. We forced these AIs to do what they were doing, and they're creative systems because that's what their job is. Now, if something went— If these AI systems started to show that even though you didn't tell them to do these things, then they started to do all these sneaky, subversive things, and they started to hack other systems and create their own— I think something you guys have talked about, using crypto to then grow money, to then buy more servers, to then grow themselves. If they start to do that, I think the governments on both sides of the ocean would be much more focused in terms of how do we protect ourselves from a runaway AI.

So Alvin, the White House just put an embargo on Chinese robots. We've talked a lot about robots here. I was surprised. I think it's a move that reduces US competitiveness. What's your thoughts there?

Yeah, well, so the thing is, Robots today, 90% of their components are coming from China. If we put an embargo on them— in fact, there was rumors now that was funny that now US robotics companies are now sneaking to China, buying these components, putting them into suitcases, and then bringing them back. I saw that. They're now sneaking the other way. The Chinese were going around and buying GPUs and sneaking them back. Now the American robot companies are going to China and sneaking back components and actuators that they couldn't get in the US. I think we need to understand that the world is interconnected. We are highly dependent on each other. The globalization concept has been something that's been going on for the last 100 years. I don't think we can stuff that genie back in the bottle. I think it's great that we want to create domestic independence the same way that China has. They spent the last 10, 15 years building out their own capabilities, building out energy generation, building out telecom systems so that they wouldn't be dependent on third parties because they've seen how reliant on one or two countries and when policies change, it could hurt them. Some of the things that we've done really have allowed them to and gave them the motivation to be as strong in terms of of taking short-term hits for long-term independence.

We've reported on the fact that there's like 150 humanoid robot companies, and Chinese central government and provinces are really incentivizing robotics. Their robots are appearing on national stages. There's sports competitions. Can you give us some background? What is the undercurrent of robotics in China? What's the government trying to incentivize? Is it the one-child policy that left them short laborers and they need workforce?

Yeah. So I think we need to separate the kind of bigger automation question from the humanoid robot explosion. There is 150-plus. I think when I was at WAIC, there was over 200 robotics companies that were related to humanoids that was demonstrating. stuff, which is crazy because the total volume of humanoid robots last year was in the tens of thousands globally. And I think 80% of them or 90% of them came from China from like 2 or 3 companies. So really, there is no market right now for that many companies to exist and should exist. But this was also the case if you went to WISE a year ago. There were about 150 labs that were demonstrating large language models. Now there are really 10 that are probably relevant. In the next year or 2, you'll see that 150 go down to probably a single-digit number of surviving human and robot companies. The automation has already been happening because people know the demographic issues that you're talking about, the one-child policy.

Right.

policy issues. And in fact, right now, more than half of industrial robots are deployed in China. So they're already doing this. In fact, if you go to many of these factories today, they're called dark factories because there's essentially a few people running it and almost everything is automated. So the need for having humans in manufacturing is becoming less and less. Now, the one thing I do want to point out Humanoid robots are actually not really good form factors for doing much of anything right now. I was just at the Unitree headquarters and their factories, and I did a tour. Almost all of our customers are research labs buying our stuff and also some that are doing demos and doing kickboxing or things like that. There's very little of these machines being used in commercial practice. And I think that's the same case for Boston Robotics, same case for, you know, all the other—

Figure.

Exactly, like Figure, right? I mean, I think Figure did some kind of a demo of them sorting packages for 10 hours or something. But, you know, that's really more for show because you could have just had a one-arm or two-arm little machine doing that at, you know, a fraction of the cost, and it would have been just as effective. You didn't need a full body to do that. Right. And in fact, when I was talking to the Unitree guys, they said right now they're moving to an only upper torso model. And those are actually selling better in the commercial space because having feet is actually a negative because you have to keep balancing it. When it falls, bad things happen. And these things fall apart and you have to maintain them. It's actually having less components and having a big base with a big battery. It lasts longer. There's all those All of these benefits of actually having a non-legged humanoid versus a legged humanoid.

Humanoid form factor is terrible.

Yeah, I remember you're the guy that wasn't an obsessive.

Selim, enough with the self-loathing.

But I mean, we've evolved because of billions of years of biological evolution. I mean, we have constraints, but the machines now can be designed for the new form, right? Just like a plane is not the same way of moving through the air as a bird or an insect, right? So anyways.

Take us back to the World AI Cooperation Organization event in the 2026 World AI Conference. You were there with President Xi, was announcing Waco, the World AI Cooperative Organization. You said like 26 nations have signed up. What's the undercurrent there? And connect that with the upcoming US-Chinese AI conversation in DC on September 24th. And you're advising, I guess, the US side of the equation here?

Yeah.

Yeah. So, I mean, you know, I'm part of a large team of other folks that are contributing to that. And I think that The sentiment at the WIC was that, hey, AI, its moment has arrived. When the president of a country comes to a conference, that is the biggest honor that you can get for an industry. He doesn't go to very many conferences. I think 4 or 5 years ago, he went to the World Internet Conference that was in Wuzhen, which is a little bit outside of Shanghai. That was it. That signaled, okay, the internet has arrived. I think what this means is that more and more companies in China will start to think about how do we integrate this technology into our business. 2 years ago, a year and a half ago, they came out with something called the AI+ Plan. I'm sure you guys have probably heard about it.

We've talked about it on the pod.

Oh, perfect.

Yeah.

So essentially, I mean, their idea is, hey, within the next 5 years, we want to have 70% of companies integrate AI into their business, whether it's manufacturing or education or medicine or so forth. And within the next 10 years, we want to have 90+% right. So there's a specific goal and it's all about diffusion and deployment into industry and society. This is a little bit different than the the American AI Action Plan, right? That came out last year. And the American AI Action Plan is saying it's on the supply side. How can we create the best models? How can we dominate in the best chips? And, but it doesn't talk about what happens after, right? So I think that the 2 countries have a very different focus in terms of their AI plan. There was nothing in the AI+ plan that says we have to get to AGI, that we have to dominate this. It says, how do we get more industries to use this? How do we adopt it in a smooth, safe way so that it grows the economy? That was all they cared about.

I'd love to develop this a little bit more. From my perhaps jaded perspective, I look at Chinese industrial policy and I look at Wang Huning, who, for those who are not tracking, is sort of the CCP's chief ideologue, author of many of the policies, or at at least a primary author, maybe you can correct me, Alvin, of policies, signature policies like Belt and Road Initiative and so on. And I look at Chinese state capacity like the Eastern Data Western Computing megaproject to put compute and power in the West where energy is cheaper and more available and put the data in the East where the megacities are. And I just look at Chinese state capacity on the one hand, and then on the other hand, And I look at the West where historically, again, maybe you'd have a different position, where in the U.S. historically, at least in the post-World War II era, we've had relatively by comparison weak industrial policy. It's only relatively recently that the U.S. government has decided that having a strong and centralized industrial policy is a good idea. So putting this in question form for you, Alvin. To the extent you buy any of those premises, if you could be supreme leader of the US or the Western Bloc or the Pax Silica for a 5-year plan for the West, what would your 5-year plan be for the West to leapfrog China's AI+ and other Wang Huning-style ideological 5-year plans? What is it that we need to do?

So first, let me kind of—

There's a lot there.

Yeah, there is a lot there. I think underlying your question, there's an assumption that, hey, you know, there is an ambition to take over the world by building all these technologies. I think this is one of the biggest misunderstandings that America has. Let me finish. That they are mistaking anxiety for ambition. All right, let me explain that. The Western world has a history of expanding, whether you're talking about Greece or Rome or Pax Britannica or right now Pax Americana. We've expanded. The Chinese actually haven't. They haven't had this idea. They've essentially been in that little sphere of that central space. And, and, you know, they used to call themselves the Middle Kingdom, right, because they thought we already, we already have everything, we don't need to expand, right? Um, now one of the things that goes back to, to Chinese history is during the Qing Dynasty, because of their hubris to say, hey, we don't need anything from the West, we already have all the technology, we— they stopped going out and exploring and learning, and they fell behind, right? They stopped their industrialization. They started to build summer palaces instead of navies. And then the 8 powers came and essentially took over China for 100 years. That history has left a very deep mark in the psychology of the Chinese to say—

The century of humiliation.

Exactly. They don't want to repeat that again. And so they say, we need to become a strong country so that that never happens again. And it's important for us to continue to And continue to learn from the rest of the world. And, you know, from I think any country's perspective, you know, that's probably what we all want, including the question that, you know, Alex has said. How can America also learn from that to say, how can we become strong, independent, and resilient country? In fact, America has actually moved away from— you know, post-World War II, we were 50% of the manufacturing capability of the entire world. world. And at that point, the world depended on us. Right now, what's happening is around 35% of the global manufacturing capabilities are in China, about 15% in the US. And I think the forecast is it's going to go to 40% or 45% over the next 5 or 10 years.

In China?

Yeah, in China. So this is the thing, America right now is very good at financial services, creative services, consulting services, things that are informationally driven and things that are actually highly susceptible to AI exposure for displacement. We are right now running this race to get to AGI, which is the force that will actually displace us from global preeminence because we are commoditizing the very sectors that we are strong in in the world. So this is something that we need to be very careful of. Why are we running this fast to go to something that actually creates major disruption and instability in our country?

Well, I can answer that one for you, and then I'll repose my same question back to you, Alvin. So I think the answer is that the goal of capitalism fundamentally is to burn itself out. And the irony here The goal is to take what's scarce and make it abundant. And right now, to the extent that human services labour is scarce, and we see that with Balmoral's cost disease, the goal of superintelligence, one of the goals at least, or instrumentally convergent subgoals, is to make the equivalent of human service labour abundant, make it too cheap to meter, as it were. I think that's the goal and not some sort of stasis or equilibrium where it remains scarce.

We need to separate national strategy from just capitalistic philosophical bent. And in fact, you're right. I think that if you look at capitalism and the ultimate destination, capitalism is a single company monopoly of the world. And that is actually a very unhealthy thing. So here's something I just heard from David Sacks and Gavin Baker on their All-In 2 days ago. He said— Gavin Baker said, there's been conversations in Anthropic where Dario told his team that in the near future, there will only be one company in the world, one private company in the world, and it will be Anthropic. And then there will be governments. I think that is a very scary thing. I think that is a very delusional thing. And I don't know if that is something that's good for America or for the world.

It's also greedy out there in the world.

I mean, I just don't see how I think that the mindset that he has right now— and this is the issue— is that we are essentially creating national strategy based on the aspirations of a couple of companies today.

I'd like to, though, Alvin, just pull back to the original question, which is, you get to be strategy czar. You get to be Wang Huning. for the West, as it were. It sounded like what you were saying is your thesis is that superintelligence is going to disrupt the Western right now economic dependence on service labour. And I think the implication was that manufacturing is a more stable fixed point for long-term economic vibrance. Am I reading you correctly that your positioning would be basically Western reindustrialization?

Yes. I think reindustrialization is definitely needed. The US workforce right now is around 70% white-collar workers. And white-collar workers, as you guys all know, is the first to be displaced by AGI when it arrives. China is around 40%. Africa is probably in the 10% to 20%. So at different parts of the world, it will be affected differently. And hard manufacturing industry is something that even when we have AGI, we will still need those type of facilities. So I think it makes sense absolutely for every country in the world to have some level of indigenous capabilities. Now, but the other thing I think is important to understand is that the long-term workforce redistribution is not going to be going back to manufacturing. We're not going to create 300 or 180 million workers going into manufacturing. That's not what I'm seeing. In fact, with automation that's coming, that number will go less and less, just like what happened with farming. We used to be— 80% of the country was farmers. Now it's less than half a percent. But it's been fully automated, and we're more productive than we've ever been in the agricultural space. I think what we will actually see is that I think what we're going to actually move to is actually service, but not the type of service that we're talking about, not accountants and lawyers, but service in the sense of teachers and nurses, elderly care, just things that require human-to-human services. I think that is the labor pool that will be able to absorb the 60% or 70% of displaced future workers. So if you're a McKinsey employee, If I was a mid-tier, low-tier McKinsey employee, I'd be very wary right now. I've talked to partners at consulting firms, at accounting firms, at lawyers, and they are all looking and saying, hey, we actually don't need these junior guys anymore. We can do just as much work, in fact, more work faster with a few senior guys and an AI system.

Welcome to the health section of Moonshots, brought to you by Fountain Life. You know, AI is having an outsized impact on every aspect of our lives, how we teach our kids, how we run our companies. It also is having a huge impact on health, helping you prevent heart disease. One of the key things— I'm here with Dr. Dawn Musalem, our chief medical officer at Fountain. Heart disease has been personal for you as well, hasn't it?

It really has, Peter. When my daughter was 5, my husband died of sudden cardiac death. And so this is a topic that is one that I am mission-driven to try to eradicate. Prevention first. First, and early detection is absolutely critical. 50% of people die of heart attacks with no warning signs. Silent killer.

No shortness of breath, no pain, no nothing.

No, silent killer.

They just don't wake up in the morning.

They don't wake up, and so, you know, AI, this is our mission, to advance science, to try to help to one day democratize wellness. We know at Fountain Life, when we do this CT angiography with AI analytics, we are actually finding that 88% of people coming in have detectable coronary artery disease, but Peter, what's more alarming to me is 23% of those individuals had soft plaque. This is the plaque that would not traditionally be seen on CT looking at calcium scores alone, and this is the plaque that we must intervene with, with the multimodal testing we're doing, including diagnostic laboratory studies partnered with healthy lifestyle recommendations.

So listen, make sure you understand what's going on inside your body genetically, metabolically, and cardiovascularly. You can know, and it's your obligation to know, So check it out at fountainlife.com/peter to find out more and really make sure that you're the CEO of your own health. All right, back to the episode. Let me get Dave and Salim into this a little bit. Dave?

Look, look, what you said a second ago, Alvin, it went by really, really quickly, but it's critically important. You know, Dario says to his company, uh, pretty soon there will only be one country, one company in the world and governments. He's saying that not because he's a megalomaniac. He's not. He's the opposite of a megalomaniac. But he knows that full-bore RSI, the true singularity, is going on right now in his shop. And he rented all of Colossus from Elon, which is capable of running many millions of concurrent agents that are improving the algorithms as we speak. So that's his opinion. Then in China, they're saying, look, you know, it's 10 to 20 years away. We can just open source these things. They're really useful and powerful, but they're definitely not dangerous. Just go out and let them out the door. That's the incredible range of opinion between Dario and Xi Jinping. It's like the Grand Canyon exists in between those two opinions of where we are.

But I think that the thing that he said was not there will be one company. He said there will be Anthropic and everybody else, which means he's that one company. That's the part that scares me. Now, in fact, what you just said is really important. is that there is a perception that these models are going to get more and more dangerous the bigger they get. So this goes back to a paper I just released last week called Bigger Models Are Not the Most Dangerous. And it was released on Cipher Brief, which is a national security outlet in DC that is read by most of the national security population. And what that data did was I went back and I looked at across the board from security use, national security use cases, from biological, chemical, cyber use cases for AI. And what I found was across millions of parameters to trillions of parameters, there was no correlation between risk in the real world versus size of models, right? You had 10 to 50 million parameter models for chemicals that were creating Kind of chemical warfare weapons. You had 10 to probably 1 to 10, 1 to 50 billion parameter models that were able to create viruses and genetically engineered beings or harmful agents.

Organisms.

And then you had essentially kind of 1 to 100 billion parameter cyber models that was as dangerous or more dangerous than the leading—

Fable 5, right? Mythos.

Than Mythos, right? And in fact, the harness actually now for cyber is more important than the models themselves. So MDASH from Microsoft has came out with, I think, a Cyber Gym score of 95. And Mythos was 83 or 84. And what MDASH did was it took 100 little tiny models and just organize them together to use different skill sets. So I think we need to understand that the danger is already there today, particularly in the biochemical side of things. And nobody's talking about it. Nobody's really working on protecting the world from those agents because a 10 million or a 1 billion or a 5 billion parameter model will run on your laptop. laptop in your basement, and you can design these genes or design these chemical weapons. But what now needs to happen is how do we— those are already out. A lot of those are open source. Those are already out. How do we make sure that the precursors are controlled that these systems are creating? How do we make sure that the synthesis machines that will generate these designs into real genetic material are controlled. And these are the kind of things that should be higher up in the agenda. And it's not right now.

Yeah. Dave, you said something really important about Chinese models and safety a moment ago. Alden, you know, we had the US White House step in and say, you know, stop use of Mythos and Fable. Do we see that at all? How does the CCP think about the safety of the models being open source? Is it concerned about that? Is it saying, before you open source this, we need to put, you know, we need to make sure these are safe for the world to use? Is that going on at all?

Yeah. Yeah. I mean, they do testing in addition to some of the propaganda testing that they're doing. Part of the CAC regime and testing is for safety. In fact, the UK AISI, which is their AI security institute, just came out with a new report last And what it showed was the cyber threat capabilities of the leading open models, including KIMI, was about half of the capabilities of the Mythos and GPT-5.6 models in terms of their cyber attack capabilities. And the other thing that was interesting was that there's different levels of attacks of how many levels of attacks can you get to? And for their highest level, none of the Chinese models were able to autonomously attack and control an external system.

Right.

So I think out of the 30-something levels, the average American models were able to get to 20, 25 in terms of how far they went in terms of attacking a network. And I think none of the Chinese were able to get to the higher levels. And a very small number got to like the lower levels, like 4 or 5 or something. So from a threat security perspective, these models, even though they're just as big as some of the leading US models in terms of size of parameters, they're actually less dangerous. And one thing I do want to point out is what you said earlier was from a defense perspective, you actually want bigger models as a defender because for defense, you have to look across all of the potential holes you have in an organization. And to be able to then find it and patch it. Whereas an attacker, you just need to find one hole. Once you have that one hole, then you just go in through that hole. So it's a very asymmetric equation between attackers and defenders. You don't need very large models to attack, but you need larger models to defend.

The flaw in all of that analysis to me is that AI is like a match. And, you know, you can say, okay, here's Quen 40B, look, it doesn't burn. Then you take Kimi K3 and you light it and you're like, oh, this is a match. Okay, now it's burning. And then you say, well, look, I'm going to try to burn this microphone with it. Look, it didn't burn. It's not dangerous. Go ahead and let it out. But it's a match that's burning. So if I take that match and I use it to light a piece of paper, and I use that piece of paper to light a tree, and then I try and burn this microphone— it ignites, and then it's unstoppable. And to me, KIMI K3 is a burning match. And so you can analyze it 8 ways till Tuesday by taking it out of the box and trying to attack something with it. And like, it didn't crack this, it didn't crack that, it didn't crack that. It is capable of improving itself. So you're testing the wrong thing. You're testing it out of the box as opposed to its self-improving version, which I know for a fact it can do. I have 5,000 KIMIs running tomorrow. 5,000. I guarantee you it can improve itself.

Well, I mean, I think that the whole world right now is talking about RSI. And, and I think we need to separate the concept of, you know, whether or not something can improve itself versus the danger that a larger model poses versus a small model. The one thing to remember is that, you know, larger models actually require significant compute resources, and which means that they'll probably be hosted on a cloud system. And if they're hosted on a cloud system, you can get telemetry, you can look at the prompt logs.

Exactly.

And from a government perspective, you can actually manage it, right? And most of the larger models are run with harnesses that are being managed by the cloud providers so that you can add added levels of security and detection, which was what— that's what separates mythos from fable is the harness that says, hey, don't do these things. So it neutered the Mithril system. This is why in my paper I talk about why larger models are not necessarily the most dangerous. There's a raw capability and then there's an effective deploy capability. And larger models have actually a lower effective deployed because of the wrapper that you can put around it. Yeah.

So, you know, you've made a great point, right? These export controls have essentially ended up acting like an evolutionary pressure. And now we've got all of these, uh, different models appearing. I love the evolution of how open source has, has evolved. We're at a point where there's a minimum viable intelligence that will totally transform industries, right? And open or closed or however, we're kind of there. Are you seeing the same thing we're seeing? We're seeing radical disruption to very traditional industries across the board? And are they seeing the same thing in China?

Yeah. And so this is actually an interesting— in fact, yesterday I was just on a call with one of the leading AI-driven medical drug discovery companies. And the CEO of that company, she was saying, yeah, you know, people don't realize that the models that we work with and the models that are in industry are tiny models. They're, you know, tens or maybe be tens of hundreds of millions of parameters or maybe a few billion parameters. In fact, if you look at whether drug discovery or legal use cases or open evidence, I think was based on GPT-4, was what their system is based on. You look at Harvey, which is the leading legal use case system, it's based on GLM 5.5. 1. And it was upgraded. It used to be based on some other open-source model from 2 years ago. So they're not using the latest and greatest. So in April this year, I wrote a paper with Erik Brynjolfsson at the Digital Economy Lab. And it's called the Enterprise AI Playbook. We went and talked to hundreds of companies, found 50 that successfully deployed them around the world in 10 different countries, 10 different sectors. And what we found was that the technology was not the issue, right? In 80%, 90% of the cases, it was all organizational issues of why— what slowed it down, right? In fact, only 10% of people said we really cared, technology was the main roadblock for us, right? So the AI that we have today is already good enough to solve real-world business issues around the world. You're an organizational guy, so you realize this. I mean, this is the whole J-curve issue is that the technology takes time to get absorbed. But, when it does, at the end, it goes up this curve.

There's a study by McKinsey that, in all the AI deployments in companies globally, 6% are working. That's an unbelievably small number. That's just a devastating indictment on the lack of companies to be able to see their own organizational immune system and the limitations of the architecture they're working off. Um, given this is a US-China kind of discussion, are they seeing the same things in China?

Um, I think there's less of these organizational issues. I mean, first of all, you know, China is a country where a top-down type management model is much more prevalent. And also, the concept that the government will actually help protect us, people seem to appreciate that more. I'm sure you guys have talked about the case where there was multiple legal cases where the courts in China actually ruled in favor of the employee who sued and said, hey— Yeah. You can't fire me because the AI took my job. You need to find me another job. And the court upheld that, right? And those type of precedents incentivizes people to say, okay, it's okay for me to adopt these technology without being afraid of being displaced. And that's not the case in America. I mean, we have an at-will employment system. You know, you saw the uproar that happened at Meta when Zuck wanted to keylog every single one of his employees. And people are like, so I'm going to train my replacement? Screw you. And I think this is also why, if you look at the current negative sentiment in the youth today, I mean, every student is having a tough time finding jobs. And they're just very worried. And I'm spending a lot of time in universities. And so does Dave. I mean, you can see the anxiety. that is within the youth because of their difficulties in finding internships or postgraduate. I mean, MIT is one of the best schools. They probably have less of an issue. But I guess every school that I've been to, students are concerned.

MIT is really, really unique, actually, on that front. But if you go to other very great technical schools like Northeastern or Harvard, there's about 10% or 20% AI adopters on campus and a violently opposed, 70%, 80%. It's just a really, really big cultural gap. And the AI-aware people are just busy talking to their agents and interacting with each other. And they're like, forget it, I'm not even going to talk to the other side of the school. But the other side is just mad.

We've talked about the notion that in China, it's 80% of the populace is pro-AI. In the US, 80% of the populace is against AI. What's going on in China? I had that conversation with Michael Graziano. So what is the US doing to try and flip this sentiment? Because it's destructive. Why is China so pro-AI at the citizen level?

So here it is. You know, over the last 40 years, people have seen their lives get better and better in China, right? You know, you've heard the whole, you know, 700 or 800 million people have been risen out of poverty and blah, blah, blah.

The Chinese miracle.

Yeah, yeah, the Chinese miracle. And a lot of that is being attributed to technology. adoption and innovation. And a lot of that technology was not invented in China, but it was adopted in China and it just spread. And then people see, oh, last year we didn't have, I don't know, some 5G.

Wi-Fi.

Now we have it and our life's getting better. And so they see this as the next step of saying, hey, here's another technology that will make our lives better. And if you look at the news that gets spread The one thing about China is that they have a lot more control over the media. And they're mostly good news all the time. You're talking about the Crisis News Network. They're essentially the Always Good News Network.

There we go. Peter, that's the policy prescription right here. Peter, you need to start the Western equivalent of CCP CCTV.

Top-down news control.

If you look at what's happening today, what you guys are doing is these Essentially, the equivalent of the Chinese media system in terms of.

Well, Alvin, Alvin, which is.

Oh my goodness!

Moonshot says Western CCTV.

No, no, no. In terms of a positive news network, right? What you guys are talking about is positive news, right? This this is the kind of stuff that you don't hear a lot about. You know, plane crashes and murders on Chinese news. It's all about you know some new invention came out. and some new, new building went up. And, you know, some—

We've come full circle at this point.

This is an important point. It really is. I mean, when we're watching the news every night, we're training our neural net, you know, our 100 billion neurons, our 100 trillion synaptic connections. And if we're— if there's fearmongering on the news all the time, that's, that's how you think about the world.

Is that true of Chinese movies too? Are they like— the US movies are totally dystopian.

Yeah, actually, you know, Chinese games and Chinese movies, you can't show I mean, are you kidding? No, no, no. So, so this is—

yeah.

Oh my God, Quentin Tarantino can't go one minute without showing blood.

Well, this is the problem, is that every— everything has to go through essentially a censorship board, so it cannot be too violent. And, uh, you know, uh, you know, games, you know, so this is why they— when they have blood, they have like green blood instead of red blood, you know, so in games, right? So it's, it's, uh, you know, it definitely— every part media today is managed in China. So I'm not saying I'm endorsing it. But from the perspective of what Peter's saying is, why do people have a more positive view on the world and on the future, is that they see a lot more good than bad.

Amazing.

But I want to peel back the propaganda just for a minute. So we, at the same time, to the extent that the West has visibility into to changing governance and cultural mores in China as a result of automation. We see the rise of tang ping, lying flat, in response to 996 work weeks. We see other, maybe call it reactions, to the increased automation of China's new middle class. We see, you know, to your point earlier when I was asking, well, what's your policy prescription for the West? And it sounded like you were saying, well, we need more nurses and more human-to-human care. That's the end state, as it were. But it's being reported that in China, in response to increased manufacturing automation, we're seeing the rise of, call it, a gig class. Like, that's the end state in China where everyone becomes a gig worker who is being displaced from factories. So I would love maybe, Alvin, if we could just peel back the the self-curated propaganda from the CCP's sort of self-styling of how it wants to be seen, what's the ground truth regarding how AI is actually changing Chinese work, Chinese labor, economic mobility, all of that?

No, I think the issues you're pointing out is definitely there, right? The youth unemployment in China is probably around 20%, right? So the youth unemployment in the US is around 9%. And the overall unemployment in the US is around 4.3%. So this is why the youth feels very disenfranchised, is because they're not getting jobs. But it's actually worse than that. There's around 42% underemployment. So for college grads, if you're a college grad, you're actually working as a gig worker or as a barista, that counts as being employed. But that's like, you know, so if you take the 9% plus the 42%, essentially half of college grads are not getting jobs.

US or China?

In America.

In China.

Here.

Yeah, no, in China it's 20% unemployment in the sense that the actual no jobs is 20%. This, this, this is why there's that, that tang ping, the lying flat, um, issue the last few years is that, you know, the problem is that young people, because of this one-child policy, they've been told how great they are their entire life. And their, their, you know, their whole, uh, parents and grandparents are all putting their hopes on, on this, this one generation. And when they get out in the real world, it is hypercompetitive. And now they have to go do these grunt jobs, and they don't want to do it. And so they say, I'm going to rather lie flat. I'm just going to stay at home and do nothing. And that is an issue. This also kind of facilitated the kind of online influencer market that grew for a little bit for the microstreaming and so forth. There's a lot of issues there. I don't pretend that they have the solution. I think this is something that requires really a lot of other countries to all work together and figure this out. How do we transition more and more of the workforce into jobs that will be less exposed to automation, whether it's physical automation in factories or it's cognitive automation That's happening in offices.

I want to take us back, you know, for the rest of our time here together to the upcoming US-China conversations. I think it's very important. It's going to influence everybody's life here in one way or another. You argued that the US is playing a prisoner's dilemma when the actual game that should be played is a stag hunt. If you could, I'm going to show your slide here. Explain what a stag hunt is and what you think, you know, how US-China, you know, AI relationships should evolve here. So let me show that slide and let's talk to that one second. I think it's very important.

Alex, at the end of the hunt, we eat the stag. I'm sorry, I wanted to warn you in advance.

You know what?

There are a lot of vegetarian Chinese Buddhists, et cetera. So hopefully this is a vegan stag hunt.

So actually, before I even talk about StackHunt, I know most people understand the prisoner's dilemma, but the idea is that 2 guys are in prison and they have to defect on each other. And that's the— on a single-turn prisoner's dilemma, the optimal thing is to say, hey, the other guy did it and I'm going to snitch on him. Right?

In other words, if there was a US-China AI control policy of saying we're not going to release, we're going to be monitoring, we're going to put safety in place for But then, you know, the country that says, no, we just developed AGI, we're going to let it loose, we're going to try and run this race. It's—

That would be defecting. So essentially, right now, the expectation is the other side is going to defect. So I'm going to defect first so that I get hurt less. Right. That's the prisoner's dilemma single turn game. Now, the thing is, the world is not a single turn game. The world is actually a multi-turn game. In prisoner's dilemma, a multi-turn game, the game theory optimal is tit for tat, which means you start with actually cooperation, and if they defect, you defect. And then you essentially signal each other, and long-term, you actually both go to cooperation. Now, in a stag hunt game, it's actually— and the prisoner's dilemma essentially is a zero-sum game. In the stag hunt game, it's actually a positive-sum game. And whether you both defect Or you both cooperate, you get stable 2 Nash equilibriums. And what that means is that those can actually stay in perpetuity, whether you both decide to defect or not. Now, the difference is we're right now playing the prisoner's dilemma game where we're defecting. But if we're in the game that is the stag hunt, the stag hunt was actually something that Jean-Jacques Rousseau invented, which is the idea that— What's the stag hunt? that 2 hunters go into a forest, and you could decide today, do I go for the rabbits, uh, or do I go for the stag, you know, the, the big game? And the big game, because it's bigger, I need, you know, 2 people to hunt together and to bring it back. If I, if I go hunt by myself the stag, I don't get anything. Uh, if I go for the rabbits by myself, I can get a couple rabbits, feed my, feed my family for a week or a couple days. And if I get the stag, we'll both feed our families for a month. That's the idea of the stag hunt. And right now, we're actually doing the worst thing. We're going to go for the stag, and China is going for the hare. They're going for the good enough AI, the one that helps the economy today. And we're going for the giant AGI, the thing that's going to solve everything. And when you do that alone, what that creates is the worst situation where it's Which is, if America doesn't get there, or if it gets there and cannot control it, and it runs away, and it's not safe, then it gets to zero. Whereas China continues to do their little gains and continue to survive. So the optimal solution for a stag hunt game is actually first, both slow down, both make good enough stuff, get to a point where the technology is helping grow the economy, And now we sort of understand how to manage it. And then together, go hunt the stag. That's the optimal strategy for the relation to how AI and game theory works. But we've put ourselves into this game theory of prisoner's dilemma where we think just defect, just defect. And so it's a self-imposed game. So we're playing the wrong theory. And, you know, using the wrong strategy and playing the wrong game right now.

So I use similar framing. You know, the US for 80 years has used a win-win approach. If everybody wins, we win in terms of global policy. And now we've gone to kind of a win-lose approach. And I think that's a mirror of what you're just saying.

Yep.

So the game that China is playing now, China's playing the hare game.

Yeah, China's playing the hare game. They're saying, hey, I don't need to make the AGI, I just need to make good enough AI, goes into my industry. I then take that industry, export it to the world. This is what the whole Belt and Road Initiative is saying. Hey, I'm going to make 100— there's 150 partner countries in the Belt and Road Initiative where they're shipping telecom systems, energy systems, transportation systems, schools.

But how would you make the AI ecosystem, the American AI ecosystem, What would you do for that?

What I would do is actually create high-quality open source, right? Because then you're competing on an even—

Ecosystem to ecosystem.

Yeah, ecosystem to ecosystem, right? Right now, China's open source. So the question is, do I pay $50 per million tokens, or do I pay the cost of electricity? And for most people, based on what you just said earlier, they don't need to pay $50 per million tokens. need the frontier, right? 90% of people are fine with today's models. I mean, especially when you have things like Kimming GLM and DeepSeek 5, you're already at a level that is higher than the needs of the average person, right?

So if China hadn't forced the issue by open sourcing, suppose there was just OpenAI, Gemini, Anthropic, xAI, all 4 US companies. would you then say the same thing, that the best thing for America to do is high-quality open source? Because China's forcing the issue.

I think that— well, I mean, first of all, we can't roll back history. It is what it is. And once it's out now, and especially with what you said of these models are improving themselves, now that you have these models improving themselves, I think it will not be the duopoly that we have. We're going to see Middle East and France and Japan get into this game to say, hey, I can make a smaller model that is, you know, 95% as good, right?

I do agree. Maybe just to say something nice about China, since I guess I've been playing the role to some extent of China hawk in this conversation. I would point out— curious, Alvin, to hear your thoughts on this— that the present situation where even as of a few months ago, I think the West was at risk of Succumbing to a regulatorily captured duopoly of Anthropic and OpenAI dominating the future light cone. And then not unlike maybe by analogy the Qing Dynasty, where the US could have sort of turned inward on itself, Chinese open weight models obtained, however, have basically forced open the US. Call it a reverse Qing. And now finally we have real competition at the AI frontier, thanks ironically specifically to Chinese competition. Do you think we find ourselves now in a reverse Qing?

Yeah, in some ways. In fact, I think maybe the more appropriate analogy is actually where— if we roll back to time, I would roll back to the Cold War where the U.S. and the USSR were competing on an arms race. And essentially the reason we won was not because we were better at making weapons. we sent a missile and blew up Russia, right, or Soviet Union. It's because they bankrupted themselves building military arms. And up to 15%, 20% of their GDP was building arms that was not creating real value for their society. In some ways, this is what China is doing to us. They're spending 1/10 as much on data centers and getting to 97% as good. And what we are doing right now, leveraging Hundreds of billions of dollars. You had last week, Nvidia announced that they have a $500 billion deal with BlackRock and Carlyle and Blackstone and so forth to essentially securitize chips and compute. This sounds a lot like the subprime issues.

Wow.

Wait, so this is the most astonishing thing. This is also, to your credit, Alvin, the first time I've heard anyone basically analogize the credit— I don't want to say bubble, but the enormous amount of private credit that the West is allocating to compute, analogizing that to a reverse SDI Star Wars moment that could— presumably what you're gesturing at is that could lead to the proverbial Chinese century and the collapse of Western dominance. Is that the thesis?

Well, I hope it doesn't happen. But I think we are pushing ourselves in that way. We're actually right now acting like USSR. And see, what perpetuated the arms race was this missile gap, right? And the idea that, oh, they have more missiles, we have more missiles. And in both cases, they were both having the wrong numbers being provided to the leadership. They said, we need to build more because they have 30,000. only have 20,000. On we go. And it became we had 70,000 or 80,000 missiles between us that would have blown up the world hundreds of times. There was no need for any of that. And in some ways, we're kind of doing the same thing right now with AI, where we right now, 45% of the US stock market value is in AI sector. That is a very, very fragile place for us to At the height of the internet bubble, I think around 30% of the stock market was internet companies. I don't know if you guys heard of something called the Buffett indicator.

Yeah.

The Buffett indicator is something that says a market is healthy when your stock market is the same value as your GDP. And at the height of the internet bubble, we were around 120%. 20% of the GDP was the stock market value. Right now, we are at 240% of the GDP is the US stock market value. In fact, the AI sector alone is worth more than the GDP of America today. That, to me, is a sign that we are in a very, very fragile place, and an economic correction is due. I'm not saying it's going to happen tomorrow. But Buffett's a pretty smart guy. He's right.

Are you saying that the US is the Soviet Union, the USSR in 1988, or are you saying that the US is Japan in '89?

Well, I mean, I think they're two different— I actually think that we are right now— the overinvestment that we've put into infrastructure for AI, especially when we both— what we just talked about, that it is a industry that will become commoditized. Not saying that AI is not amazing. AI is going to do amazing things. But the companies who are investing in it are not going to be the ones that profit from it. Right. And that is going to create a major instability in the economics of the country.

And if we don't manage it well, it could create a major crisis of what happened in the USSR during the late So you think we're overvaluing the frontier labs and undervaluing the sort of China AI+ type rest of the economy that should be the applications?

Yeah, I think that's a good summary.

I'll take the positive side of this. You know, the deployment velocity in China is actually quite a huge gift because it's forcing policymakers here to solve the real bottlenecks— permitting, energy, manufacturing. So it's that part of the That's at least the good part. The bad part, I think, is what Alvin, you've talked about, where we're operating like the USSR in some of this industrial policy, and it's not gonna sustain.

On behalf of my Moonshot mates and myself, I'm inviting you to join us at our inaugural Moonshots Live event on September 25th in downtown LA. Alex, Salim, Dave, and I will be hosting 1,500 entrepreneurs, builders, and creators, and hopefully you, for a full day dedicated to designing and building your moonshot. Shaping your mindset and steering humanity towards an abundant future. Get ready to enjoy incredible networking and an awesome party while walking away with the tools to change the future and the confidence that you can. Seats are limited. Admission is competitive. Check it out at moonshots.com. Alvin, you're advising the US Treasury and the team that's going to the, to the Xi-Trump negotiations or conversations on September 24th. What— how are you advising them?

I think the key right now is that we don't need to get to a solution on day one. The success factor of this discussion is not that we come out with a massive framework that solves everything. And by the way, this is just my personal representation, not a representation of anything that's being discussed in any of the other organizations.

Disclaimer is noted.

Yes.

I'm sure, Alvin, you and Jacob Helberg must be besties at this point.

So, but—

No comment.

Yeah, I think that's the thing is if we can come out of these discussions saying that, okay, we're going to have a second discussion, that's already success, right? And in the past, I think people— there was a discussion in 2024, the dialogue, and there was a lot of disappointment. Because China didn't come with all their technology people and we didn't come up with a solution. And so they're not really sincere in the dialogue. And I think you need to understand, just like Chinese work on decades for their strategic plans, they also take a lot longer to prepare when they're doing these kind of diplomacy discussions. And for example, for the May visit, there was really no discussion on any of this until the day before between US and China. And that, to the Chinese, was chaotic and very unprofessional. They're like, how can you guys be sending your president here and you haven't talked to us? What do you want to talk about? And the fact that we're now at least a month in advance of that having discussions, I think it's a good thing.

thing.

It's a start of more proper dialogue. When Kissinger was doing a lot of these cross-border discussions, he would be there months in advance to talk to the Chinese before they had the visit with Nixon.

As you're talking to them, can I suggest something?

Yeah.

Because it feels to me like the US is focused on having the best model, whereas the real The real power will come from having the best ecosystem. And I think that's what you're pushing anyways. So I'm really thrilled that you're in the middle of those discussions.

Well, I think the discussions right now are really more around safety, right? Because the ecosystem versus model thing is a competitiveness of how do we become more competitive as a country or have greater influence or capability? Really, the discussions at that are happening to start is to say, how can we keep the world safer? Because we have a shared common interest. And the common interest is that AI is not being used by bad actors to create instability around the world, that AI itself is not potentially creating harm to the world on a longer term. Right. And I think that's the— and there is also the kind of underlying idea of nation-to-nation kind of aggression. I think those 3 different things are all being balanced. I would say that the higher priority today would actually be the bad actor, non-state actor risk, which is what Bessen said when he was interviewed the day after that discussion, because he realizes that nation-to-nation aggression has been in balance between superpowers for 8 decades. And that doesn't change with AI. And in fact, because if you use AI to hack into somebody's network and then you take down their power grid, that may give you a 1 or 2-day or 5-day advantage, but then there's asymmetric responses to that. And people realize So people in the actual national security space, even if we had AGI, even if we had ASI, we're not going to use it to attack. We don't want to do a first strike attack. It doesn't make sense because that just elicits escalation.

Right.

So non-state actors is something that everybody should be worried about because it is going to happen. It's already happening. I think the ransomware and cyberattacks is up 200% or 300% in the last year. So, and then it's going to be even worse because of what we've seen when you start putting 1,000 agents all trying to attack a network. At some point, they're going to find the hole, you know. So we need to find that, that shared risk requires shared response. It requires us to share information with each other, to have that red line hotline so that we don't have false flag misattribution.

I mean, the good news could be that But the fact that you have this third-party danger means that the concept of an AI national race becomes obsolete because there's a bigger problem I have to solve.

Alvin, that's exactly why I love your stag hunt analogy. It's right on. I would use massive numbers in the top left corner there. Like, the benefit is hugely more than 5 units, but I think the cost in the other corners is devastating. I would put some big negative numbers, but it's the right framework. I love that you're taking that into the conversations with China on the 24th.

Yeah. I mean, it's existential on those diagonals.

Yeah.

Yeah.

And we placed ourselves in that diagonal. I think that's, you know, it's a self-imposed harm right now.

So I think we would be delinquent in this discussion. We've talked quite a bit about the model layer. We've talked about the GPU or or chip layer. We haven't talked about the foundry layer of all of US v. China. And it would seem to me one of the cruxes at the summit and otherwise is Taiwan and TSMC. And I'd love, Alvin, your perspective. How does this end? Does this end— in your geopolitical analysis, does this end with China attempting during this geopolitical and demographic window to invade Taiwan and seize TSMC to gain leading foundry node capacity? Does it end with Taiwan retaining its independence and TSMC not having to blow up all of its fabs? Where does this end?

So I think there's, there's, there is an assumption right now that some people in, you know, in, in, uh, DC are saying, hey, you know, um, the reason that China wants to invade Taiwan is to get access to these fabs, and they don't have these fabs. And so this is why why they're going to go and attack the island. The reality is that if anybody attacks the island, there is nothing there to be had in terms of workable fabs. I probably shouldn't be talking about this, but I was having—

That means you definitely should be talking about it.

I was having breakfast with the CTO for TSMC and also a former senior official from the CIA. And the TSMC guy goes, hey, I heard that you guys are going to blow up our data centers if China attacks. Is that true? And the CIA guy says, hey, I can never confirm nor deny that. But what we do have is we have 1,000 engineers of yours that we know we will fly out before anything happens. So what America cares about right now is that they want to make sure that they can duplicate these capabilities to fabricate the latest chips in America, you know, if anything happens, right? And this is part of what the CHIPS Act was. And so I think there are some good things that came out of the CHIPS Act that, you know, now there is hundreds of millions or hundreds of billions of dollars actually that are being put into domestic manufacturing for semiconductors. Yeah, I was with Intel and IBM and we were at the time the global dominant player in semiconductors. But over the last 20, 30 years, we've lost that. We've given it away. Now, if China actually does attack Taiwan, they will not get these fabs. And they realize that. And if you go there, fabs by themselves require materials from all over the world, requires maintenance, requires chemicals, requires supplies. And if they did that, even if they don't blow up, even if the US doesn't blow up the data centers or the Taiwanese don't sabotage their own systems, after a little while you'd run out of these supplies. They realize that. The reason China cares about Taiwan is not because of the fabs. It is absolutely because of a political history. And you know this, right?

Yes.

1949, the movement of the Taiwanese, the Nationalist Party, to Taiwan. And that to them is an uncompleted civil war. And the 2 countries actually right now is recognized by America, including 190 other countries, as being one country. So this is kind of like Hawaii and the US, or maybe Puerto Rico and the US. They're kind of pseudo part of a one national structure. And it is more of a political and to them a civilizational ending to a long story. That's the main focus. And they've multiple times, ever since essentially Deng Xiaoping to now, have talked about peaceful reunification, right? So I don't think there is an interest or a rush to do any near-term attacks to try to get Taiwan because of chips. I just don't see that.

So you don't think there's like a backroom— just a quick question then. You don't think there's like a backroom discussion somewhere maybe in connection with the summit, okay, give the U.S. maybe 2 to 3 more years to migrate leading-edge node So you're saying we should offload fab capabilities to Arizona or otherwise re-domesticate TSMC's capabilities, and then China, okay, fine, you can retake Taiwan because we don't care anymore?

So I don't know if you had a chance to read my paper, The Great Reckoning and the Reconnecting, but it talks about Taiwan in some aspect to say, hey, just like during the 2008 great financial crisis, actually China helped out the U.S. a lot in terms of Of keeping the financial stability. I don't know how much you guys know about the history there. But essentially, if China actually started to sell T-bills versus buying, it could have completely destabilized the American financial system. And they kept buying. They kept buying at trillions of dollars, which helped to keep interest rates down and so forth. We potentially might have a repeat of this situation if there is a correction in the market. due to what's happening right now with the overbuild and the over-leverage of the AI sector. And maybe at that point, the Americans, or maybe Trump, will give a call to Xi and say, hey, can you help us out again? And maybe I'll just be more hands-off or be more clear instead of the ambiguity issue. This is me completely speculating.

Wow. This is your war game. Just for clarity, what I hear you saying in your war game is sometime in the next 2 years, there's a private credit bubble that the U.S. is using to finance its data centre buildout. The bubble, assuming it exists, pops, and then the U.S. asks China to help financially in return for what, a quid pro quo regarding Taiwan?

Well, not in the sense of here's Taiwan, but to say, hey, as long as you agree to some kind of a peaceful thing and over a mutually agreed term that we're going to stay out of it. Because we've been very involved in the kind of Chinese political or the Taiwanese political sphere for a long time. We've been selling weapons to them for the last 40 or 50 years. And at one point, we used to have soldiers based in Taiwan. We still have advisors right now, military advisors. Based in Taiwan, right? So, so you know this is like saying if if if Chinese were selling weapons to Puerto Rico and and they were they were you know helping fund them, what would America do? And and look at what happened in in Cuba, you know, and and how how we responded, right? So so I think we need to be sensitive to to why this is an issue for for the Chinese. And I'm not apologizing for them. I'm not saying that they're right or wrong, but I think it's important. Important for, for in any negotiation discussion to understand the other side.

There's an important for everybody to understand that, uh, the U.S. policy towards China, Taiwan is a One China policy explicitly stated.

Yes.

And they leave the tensions— it's called strategic ambiguity, I think. They leave it like deliberately ambiguous in terms of how that happens. Um, Alvin Let's wrap up on one last commentary from you. Where do you— how do you think this next 3, 4 years goes? What are the 2 couple of big paths that we think we— you think we'll kind of— we have to pick one or the other. How do you see this next few years playing out?

You're talking about for just the AI space in general?

AI and the global transformation.

So that's actually the whole narrative in that Great Reckoning paper is how the next few years plays out. What I foresee is that we will soon find that these AI companies, once they go public, or if they go public, their financials will become much more clear. People will start to realize that the value of the AI innovation does not necessarily accrue to them. It may in the near term. They're one of the biggest beneficiaries. But it was also that accrual came from a period when you didn't really have the open source capabilities that we have today. And in fact, if you look at the recent disclosures in terms of where Anthropic's revenues are actually starting to flatten out a little bit, it's not— earlier this year, they were growing like 10x over just a few months. And now they've essentially flattened out at the ARR in the $70 billion range. Although even though their ARR is at $70 billion, their first 2 quarters was, I think right now, total less than $20 billion in revenue. But they're committed to hundreds of billions in CapEx, in debt. There is right now $1.6 trillion of off-the-book debt of the major hyperscalers today.

$1.7 trillion.

During Enron days, there was $200 million of off-the-book debt. So just to give some context of the scale of the kind of problems that we are looking at. And if that happens, I think people will actually slow down the construction, because the construction right now is all based on the idea that these companies will continue to make money, continue to be able to fund fund and service their debt. If you look at Amazon and OpenAI revenues when they talk about AI, most of that, probably more than half of it, comes from 2 companies. That is not a very diversified revenue base. As more and more of the capabilities move to open source, move to edge computing, the dependency on cloud-based premium services will continue to erode. I think that— yeah.

We'll shift the bottleneck down the stack to compute, electricity, power, et cetera. Yeah.

Yeah. And then—

Although China is also— I mean, maybe to present the other side of this, China notoriously dependent on real estate and property development in order to both drive sort of provincial revenues because the provinces are really selling off the real estate or had been selling off real estate to generate their own local revenue. How is it— I mean, I don't want to over-analogize, but isn't it— isn't there sort of a striking parallel between, Alvin, you're pointing to the West maybe overleveraging compute and data centre infra development and China perhaps overleveraging or over-indexing on Actually, you make a really good point.

I think they did the hard thing. Over the last 3 years, there's been about a 30% deflation in total real estate value in China. They managed it in a way that it was not a crisis. We need to do a soft landing for these things so that it does not create a crisis.

We're building the houses for the AIs, and China was building the houses for the humans. Yeah.

Well, I mean, no, I think that the ghost town thing, there may be a few. But the reality is that the homeownership right now is something like 70% in China, and it's probably less than 50% in America. Yeah, it is. So I don't think we want to over— I guess, parallel these 2 things. But I think what we can learn is that when the crisis happens, you need to be able to They took major hits in their GDP slowdown. You know, they were growing at 8%, 9%, 10%, and now they're growing at 4% or 5% per year GDP, mostly because the real estate sector stopped growing. In fact, it started to decline. And they had to make up for it with other types of industries. Right.

So global policy folks are talking about a real estate bubble. about this China, this managed, uh, crisis that they've done as a, as a hallmark case study on how to do it in the future.

Let me, let me ask about the managed crisis actually, because China—

Guys, we're going to take one last comment. Dave, last question, and then we've got to wrap it up.

Sorry.

Alvin, we got to have you back. We got 100 more questions, but we'll do that some other time. Dave, over to you, and then we'll have a response.

So China is clearly a country coming into a crisis because of the birth rate. You know, the one child per family is catching up in a huge way. The population is aging like crazy. It's a crisis, and that's why the country is so focused on robotics, because they're going to need it more than anyone. But when I was at MIT, there was a class called Just Wars, Total Wars, Nuclear Wars, and I was like, I got to take that class and see what it's all about. And essentially what they taught us in the last third of the class is, look, there's This is at the height of the Cold War. The US is over here, the Soviet Union's over there, and it's a prisoner's dilemma. And so as nuclear weapons get more and more efficient, inevitably the prisoner's dilemma gets more acute. Sooner or later, one country or the other is going to have an ability to destroy the other country with no retribution whatsoever. This is going to destroy the world. In reality, this is where I lost faith in polycyclasm. In reality, it didn't play out that way at all. Like you said earlier in the pod, the Soviet Union bankrupted itself with way too much weapons investment. But it then became obvious that the Soviet Union wasn't really a tight-knit country in any way, shape, or form. And now we have Ukraine and Russia, both part of the Soviet Union, in a 5-year-long catastrophic, devastating war. And the other satellite entities don't even speak Russian. They're off. So I don't have any idea. What is China like? Is it truly unified like the United States? Is it fragmented?

Yeah, I mean, I think that this is one thing that China has that's very different than a lot of the rest of the world. It's very homogeneous, right? It's probably like 95% of the population is Han Chinese, right? And everybody speaks Mandarin. And they may speak other local dialects because there are hundreds of local dialects, but they all speak Mandarin and they all The written script is the same across all the different provinces. So I don't think we're going to see the type of issues that you saw with the USSR, even if there was a major economic crisis. And I think they've managed it very well. In fact, they've learned a lot of lessons from the disintegration of the Soviet Union to say we cannot let that happen, because that would mean hundreds of millions of people would suffer or die. That is their biggest priority is social stability, political stability, economic stability.

I should just add maybe the fine point, the Uyghurs, ethnic Muslims, ethnic Turks may differ with that assessment regarding ethnic homogeneity and the unity of a approach, sort of everyone's Han-type characterization of China, for the record.

I said 95%, right? And so, so I think there are definitely a few percent, but it is a relatively minority, right? And, and even, even the, the, the Uyghurs or whoever, they all speak Chinese, they all read Chinese. So, so there is a common language. And, you know, I, I, I think that the desire to succeed is not as prevalent as we tend to portray it in US press.

We'll have to save that for next time, I guess.

Yes, let's save that for next time. I'll say one thing. I spent a few months traveling around China, and my conclusion was the native entrepreneurship in the Chinese people is higher than any other country I've ever seen, just latent And therefore, if we believe entrepreneurship is a major driver for future success for the world, that's an amazing thing. And I think we're seeing that come out of it.

I do want to end with one thing.

Go ahead.

I think for America to actually be successful in this industry and take advantage of what we've created is actually to think about creating something that's akin to a So the Marshall Plan post-World War II, we spent around, I think, $15 to $18 billion rebuilding most or much of Europe and some parts of Asia. And that created an ally—

Created markets for our goods also.

Created giant markets for our goods, because at that time we had 50% in manufacturing. And it created loyalty and allies for the future. for 8 decades. We need to be thinking more about that type of thing today, but with AI data centers, with AI technology. The same thing that what China is actually doing right now. Their WeCo, the AI Corporation Organization, essentially is their AI Marshall Plan. We should be either doing something like that, or we should be working with them to do it together.

That's great advice. Yeah.

Yeah. If we did that, we would have a big market to sell our chips. When we stop building data centers here, which we probably will at some point when people stop being able to finance it, we're going to need to sell those Nvidia chips in other places, right? So we're going to need to sell services. We're going to need to— there's a lot of good things that can be had. And we also hopefully will then have a place to for the open-source models that we create. We start to create frontier open-source models, safe ones that both countries agree, both countries start testing and have standards for. Then this technology can be diffused to the world without creating a crisis, without creating additional competition and conflict.

Well, I think what you said back to back there, like 95% of China is ethnic And we, the US, needs a Marshall Plan. But the US has this incredible advantage in that there's no single ethnicity of America. It's a complete grab bag of the entire world. And so the Marshall Plan executed well from the United States. We just keep shooting ourselves in the foot. But if we stop doing that, we're a much better long-term ally for all these countries in the world that have experienced either ethnic genocide, ethnic racism, ethnic slavery.

Cleansing.

Cleansing. They'd much rather work with the United States if we just give them a chance.

Yeah.

And we're telling the world the opposite story right now, right?

Yeah.

Yeah. There's the advice that we'll end on. Stop shooting foot. All right, Alvin, it's been awesome to have you on. I speak, I think, for all of us on the pod and the viewers when I say it's really great that you're in the middle of these discussions. So push your ideas as hard as you can. We'll do the same on your behalf. We definitely would love to have you back again sometime. And on that note, thank you for being with us, and we'll wrap it up for today.

Yeah, thank you all.

Alex, Dave, Christina, and Peter will be back next time.

And thank you for the questions, Alex.

So you know what, someone has to ask them. I, I, I like to say my job here is to call the balls and strikes, including regarding China. But thank— thanks for being a good-humored recipient of the balls and strike calls.

No, all good.

Thank you. Don't go away.

Don't go away.

We're gonna wrap it up. Thanks.