China's Endgame: ASI Timelines, US-China Relations, and the $1.7T AI Bubble With Alvin Graylin — Transcript

There seems to be a prevailing uh 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 USChina AI race is so important right now? >> Having a race condition forces people to make irrational [music] decisions. At some point we will get to a super intelligence type of a scenario. If and when we [music] do the concept of nations will probably become a lot less important than than they are today. The AI sector [music] alone is worth more than the GDP of America today. That to me is a sign that [music] we are in a very very fragile place and a economic correction [music] 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 [music] to finance its data center buildout. The bubble pops and then the US [music] asks China to help

[00:01:02] financially in return for what a quid proquo regarding Taiwan. >> I [music] will tell you this. Um >> now that's a moonshot ladies and gentlemen. >> Welcome to Moonshot everyone. 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 that including me, AWG, DB2, and Seem, uh my brilliant colleagues who every week help me, hopefully you understand what's happening at this incredible rate of speed. I'm Peter D. Mattis, 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 US technology ecosystems for over 35 years. Alvin,

[00:02:00] welcome. >> Yes, it's great to be here. I watch your show all the time, so I'm glad I'm glad to be able to chat with all >> We're going to quiz you then. I'll ask you as we go. [laughter] >> When do we say drink? What's Alex's favorite term? [laughter] >> Dyson's form. Dyson's form. You got to build drink. >> Let me do a let me do a proper introduction of Alvin. So, uh, Alvin Wang Grlin 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 five layers of the AI stack, the data centers, chips, PCs, phones, XR glasses, and apps, and he's built and founded over four startups. Two years ago on this channel, uh 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

[00:03:02] University of Washington. He's currently supporting the US government and for the coming USChina AI safety dialogues. That's going to be happening on September the 24th in DC. Uh the very next day on the 25th, we have our uh our Moonshots live event. Uh he's lived and operated extensively in the US, China, and Taiwan 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. Uh this is a conversation guys uh I've been waiting for for a long time to really go deep and understand what's going on. And Alvin, you're going to bring a unique perspective on the US and the China AI race. You've argued that the game isn't a prisoners dilemma. It's a stag hunt, a game theory model about coordination and trust. Alvin is the author of two key papers we're going to link to in the show notes below.

[00:04:01] Beyond rivalry and misdiagnosing the US AI, 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 toward safety. Alvin just returned from speaking at the World AI conference in Shanghai where President Xi uh did the opening keynote. Uh 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 policy makers. 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 openweight models using uh AI for diplomacy uh as well as robots and AI regulation. So very importantly we'll be discussing the coming USChina AI safety dialogues want to understand what

[00:05:00] the objectives are Alvin and you know what you think might be accomplished. So and we'll close with Alvin's recent substack essay called great reckoning before the reconnecting. Love that you and Alex both have uh wonderful uh terminologies and your essay on abundantism. 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 US China, uh, and you've spent two decades operating in China and your bio lists roles like vice chairman of Av, the VR industry alliance endorsed by the Chinese Ministry of Industry and Information Tech, and a three-year professorships at Behang, which is a defense link university uh, on the US entities list. Uh, I want our audience to understand the full picture. So, if

[00:06:00] you wouldn't mind so folks 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? Want to understand your connection to the Chinese government so people understand uh you know, you are a US 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 and you're right. I'm I'm a US citizen for over 45 years. So I moved here when I was very young. I was born in China. Um but you know having worked and lived in China you have to deal with the government on a daily basis because that's you know an important part of of being functional there. Um the IV industry of VR alliance was a uh industry association that was endorsed by the government and if you

[00:07:00] want it to be a functional organization it has to be endorsed right so um and there was 300 plus members and about a third of them were international companies companies like Nvidia and Samsung and Qualcomm and AMD and and Google and so forth. So these are the kind of companies that you know they're trying to get into to um accelerate the the industry when I was working there for HTC who you know at the time was the head the leading um virtual reality augmented reality company in the world and um you know and I was kind of the the head of the organization the head of the the company in in China right so and by the way and HTC is a Taiwanese company so at the time I was working for a Taiwanese company but we were given a lot of um uh I guess a lot of uh influence because we were such a important um party in important player in the um in the industry. Uh the Beehong University is a very large university. is a university of aeronautics and aerospace but um it also

[00:08:03] is the was the leading university for virtuality 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 to uh teach there on a part-time basis and neither of these positions uh were uh compensated um so and you know I've uh 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 um for you to understand and and work with any industry with any any government you have to understand both sides and I think this is this is why having actually close discussions with them having um you know having worked with them at the 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. >> Um Alex, do you have any uh other question you like to ask? Yeah, we we'll

[00:09:01] get into it as we get into it, but uh this should certainly be an interesting discussion. >> Yeah, you we we got Alvin's professional bio, but his growing up bio is really interesting, too. Why don't you why don't you tell us just like about your childhood and growing up and then how you got to the states and >> it's a super cool story. [laughter] >> That is it's a little little bit strange, but I I was born during the culture revolution. So my uh both my parents were artists and my uh my mother sent a letter to Xiin not to uh Ma Dong's wife uh because she had sent uh she had closed the ballet school that my mother had helped co-ound and was sent to be re-educated and this is why I was actually born on a Chinese re-education farm uh during the cultural revolution. So, so you know, so, so Alex, I do understand some of the downsides of what happens in uh in improperly organized or or governed uh states. Um, but I was able to move here in 19 uh 80 when um my

[00:10:01] uncle uh my actually grand uncle helped to uh to sponsor me. My my grandmother was a reporter for the New York Tribune back during the SinoJapanese War and uh she was there reporting um but had to leave my mom behind when the Japanese bombed Pearl Harbor. Um and uh it took her eight months to get over from Shanghai to to Tongqing where the Flying Tigers were. Um and so uh you know so this is this is why I was born there and this is why uh you know I'm actually part um part Ashkanazi, part uh Scottish and part Chinese. So it's a little little bit uh strange for my generation actually. >> Yeah. Amazing. And it it's worth uh noting uh your brother is I guess the first nuclear subcommander in the US Navy. >> Yeah. Yeah. So um he was a senior officer in a nuclear submarine actually. So u one of the big boomer ones that actually you know has a capability to

[00:11:01] destroy nations um but um you know and in fact his three three of his four children are now active officers uh in the US Navy and went to the the Navy Academy in Napapolis. 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's startup technical guide for generative media gives you 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. Uh 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. uh we keep on falling back to

[00:12:01] the reason for racing um 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 on this program. It's important for for everyone listening to understand the the backdrop of this. So let's let's kick it off. We've talked a lot about open weight models over the last few weeks uh on this pod. Um, and Alvin, uh, in your essay misdiagnosing the US China AI race, uh, you lay out a staggering shift, right? Chinese open models, uh, 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 numbers surpassed since I looked. >> It's up to a billion now. So, [laughter] >> 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

[00:13:01] yourself. You see that the US export controls on anthropics fable 5 backfired spectacularly specifically that uh within 24 hours you know China's Z.AI released GLM 5.2 under under MIT license. In fact that was the model used uh to deal with the hugging face uh debacle. uh Brazil's Rio built on Quen Japanese uh Sakana released Fugu Ultra. You say that the US denial strategy didn't slow China at all. It accelerated the uh innovation and alternative ecosystems and you call this sort of a denial keeps us ahead fallacy, right? So let me throw out the first question. um why do you believe getting clarity and some resolution on the USChina 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 um having a race

[00:14:01] condition forces people to make irrational decisions. And right now um there you know there is a perceived race condition and it's it's based on some assumptions that I think are actually misguided or mis maybe misunderstood is that they there's a perception that there is a finish line. There's a perception that the world is zero sum. There's a perception that whoever gets to to AI first or AGI first um can somehow uh you know rule the world forever. Right? And this is this is a certain a certain segment of the policym circles have this belief and a certain portion of Silicon Valley at least have this type of a belief. Right? So I I think that that that narrative um you know forces the whole discussion into a national security issue uh when the reality is right now uh none of those assumptions are really based on uh real data today right we we don't know um you

[00:15:02] know well first of all we do know that the world is not zero sum I mean as as you guys talk about every day right every every episode you know we're that the world's getting better we're getting more resources we're getting more abundance And so it it is not a zero- sum game. Um and there there is no clear finish line, right? In the sense of uh as these technologies get better, they're they're progressing and as you can you know as you mentioned with all of these open source models, uh when the the difference is gap is moving from a you know a year and a half to now probably two or three months gap between the open source and the um and the closed source models. Um there there is no no finish line where you say okay we've won right or it's not like the it's not like the space race right the space race you say hey we've landed on the moon we've won there there is an end uh whereas a a arms race type model that we are in today uh is a is a constant spend and a a constant pursuit uh without um clear clear value being

[00:16:01] returned so um but I I don't want to you know kind of take too much of this but but I think you know this will help set up the conversation that we're we're having. >> Yeah. Maybe just to pull on the the thesis, Alvin, I think that's latent that there isn't an endgame. I' I'd love to understand how you think about this. From my perspective, there's an obvious endgame there. There's space. There's development of the solar system. There's interstellar exploration, all of which I expect to be wholesomely and holistically supported by super intelligence. Surely somewhere among the the various scenarios that I I assume you're analyzing there are scientific and engineering endg games quote unquote that are intrinsically valuable to pursue. Is it is it your thinking that science and engineering and solving everything as it were is is not the endgame? Is there some other non-endgame that you have in mind? Do you think that this is sort of a a red queen type

[00:17:00] scenario where intelligence is just sort of endlessly racing as a as an end to itself or or is there is there an honest to goodness uh afterwards after the singularity in your mind as it pertains to USB China? >> Yeah. So, so I think um what you're describing in terms of at some point we will get to a a you know super intelligence type of a scenario um that may be but if if and when we do the concept of nations will probably become a lot less important than than they are today. Right? the the reason that we've you know created nations created it's actually started with city states right is because we wanted to protect a certain level of resources and um we wanted to to uh uh defend and and gain additional resources in in a in a world where we actually do achieve ASI and we get the kind of abundance that you know Peter and all of you have been talking about for you know ages uh then the the

[00:18:03] the need to to have uh separate nations with these type of competition really would uh would not exist. If if it still existed then we would probably have destroyed ourselves. >> Right. >> So [clears throat] ju just >> to to make sure then I I understand your thinking on this. Am I understanding correctly that your worldview is basically wholesomely developed super intelligence naturally yields to world government on the one hand and everything else post super intelligence as it were being solved and those two world government and posts super intelligence are inextricably linked. Um I I think that there there at at the point if we can get a aligned peaceful super intelligence then we will naturally move to a more world government maybe a a a galaxy government type of a of a model but it is not something that I think is imminent and it is it is not something that will will

[00:19:00] happen uh without some level of turmoil. Right. So so I think it the important part is about how do we get there right? Um I I I agree with you in terms of where the long-term goal is going, but um right now I don't think in the next, you know, two or three or 5 years, which is what we're really racing against. We're trying to build, you know, we're building 10 times more data centers than China is, you know, and to be one or two 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 super intelligence? you mentioned it's not 2 to 5 years is it 10 years what's the timeline >> I mean I I think the whole um evolution of this is going to take probably on the order of decades maybe you know by the end of this you know century I think we will get to a point and and in fact I think we need to move at a pace that the world can adapt and uh trying to move

[00:20:00] too quickly actually creates uh a lot of instability uh in the world if you look at the prior industrial revolutions it was uh 80, 60 and 40 years um respectively in playing out and you know we're talking about this uh revolution going from you know where we are today to the singularity in five five years according to some some folks that that um that is not a speed that the world can can uh adapt to even though >> digest >> digest yeah [laughter] because uh and and when that happens uh turmoil happens And you know it creates instability and and it may actually move the move the civilization backwards. >> Look I think a couple of comments here. one is I think one uh end point 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

[00:21:00] turned this into an arms race whereas in reality this whole thing is a platform race right uh and so I think this is the point that Alvin makes very appropriately in in his commentary I think the second mother the elephant in the room that we've just touched on here. >> This is the mother the mother of the elephant [laughter] >> a big one >> the mother is is the fact that that we're running the world on an architecture of 17th century nation states and we're trying to run 21st century uh uh 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 US prides itself on freedom on um on privacy if we

[00:22:00] have privacy that's a different subject uh and and the question you know I think then the battle here is US wants to continue its form of government, its form of um uh of democracy and not be challenged by an ASI out of China. I think that I think that's the ultimately the bottom line. >> I think that's a good way of putting it. >> Yeah. But from from that perspective, I think, you know, US and China are actually very aligned that neither one wants to have a ASI that comes out of nowhere and and destroys the system that is available today, right? I so I think there there is a a common shared uh interest and and that usually shared interest is how uh cooperation dialogue begins. So >> it's going to come out of Zimbabwe and it's going to be really ugly. [laughter] >> That's a possibility now actually the RSI is popping up everywhere. Um you know curious though Elvin if if you said look there there seems to be a prevailing uh view around the campuses

[00:23:00] that ASI is you know 5 to 10 years out maybe even 20 years and then around the San Francisco and around the big labs it's like look one two years maybe and every year that goes by they reel it in you know for 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, you know, does do does the 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 to, you know, just twiddle our thumbs and think about it. >> Yeah. So, I I think, you know, the timeline is issue is probably one of the the biggest um kind of disagreements between both these countries as well as um you know, between I think the average person and maybe some of the folks that are in Silicon Valley. Um but if if you look at the behavior of how the Chinese government is operating, they are not behaving like um they believe that ASI

[00:24:00] is around the corner, right? If if they did, they would not be um telling their labs don't buy the H200s that the Americans are giving them. they would not be uh putting out regulation that is slowing them down which you they've had regulations around AI uh you know privacy data provenence uh marking uh you know transparency and marking in in public um child addiction you know anthropomizing AI all these regulations have been around and every single model that is released in China has to be reviewed by the CAC the cyber space administration of China which again delays it by you know weeks or months Okay. So, um they're they're seeing this as something that is akin to other technologies that has happened and you know they understand that general purpose technology usually takes uh even when it's um you know invaded and mature it takes uh years if not

[00:25:00] decades to actually diffuse into society and and they're behaving like that. So, so I I think there's um there's definitely a a difference between maybe Beijing and DC in terms of how they're how they're looking at it. Um one thing I will say that there are probably two or three labs in China that are a little bit AGI pilled, not maybe to the level of the of the um Silicon Valley folks, but you know, their their their goal is very kind of idealistic and aspirational to say, "Hey, you know, we also want to create AGI." Um so but in in general the the majority of China uh Chinese labs as well as Chinese regulators see this as a technology um that is not dislike other technologies. >> Maybe let me pull on that a little bit Alvin. So what I think I hear you saying is the the Chinese Communist Party leadership has not yet perhaps woken up, assuming you believe the premise that

[00:26:00] we're uh 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 re-education 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 uh pick whatever cliche you want. We have to go at the speed of light to compete with American recursive

[00:27:01] self-improvement. What would it take? Um well first first of all maybe I'm I'm probably on a slightly different timeline as you in terms of >> I know I know >> of when most of the world is in a different timeline from Alex [laughter] >> as you guys know I mean last week you had the the pacing the the frontier letter that came out of all the lab uh you know engineers and lab heads. So, so this is something that I think um you know the industry should be looking at in terms of managing to actually slow it down, right? In the sense of if it goes too fast, as I said, the the world takes time to adapt. Um and I'm glad that actually the that you know more than a thousand uh people in in the industry and in the states are actually looking at this. Um I I don't think it's necessarily a thing that they don't know about uh you know these concept of of ASI and RSI. They they understand this stuff and and there are actually multiple safety institutes and and a a

[00:28:00] safety AI safety contingent that is in China telling these stories to the to the regulators and they hear it. In fact, you know, at the World AI conference, there was multiple um discussions and forums specifically around AI safety. So and and there was people from the the US you know the the the Benju and the tech marks were also there to to to you know add these type of points to the agenda. So so I don't think it's that they they don't know it. I think it's that um they don't believe that it is something that is necessarily um and should necessarily be a a nation versus nation uh issue. In fact, uh you know, we I think it's on the agenda to talk about these the world AI cooperation organization which they had announced um at on the first day of the um the uh World AI conference which is you know their version of PAX silica but you know PASICA in from the US that was launched at the end of last year was a US-led organization that was talking

[00:29:01] about how do we keep US dominance and and uh um leadership in AI and which allies are we going to pick to be on our side. So it creates a block to say um you know we want to be the winning block whereas what they announced was to say hey look we we want to create a global organization uh make AI a public good make it shared and uh everybody uh shares in a benefit anybody that wants to join can join and you know they're going to put out thousands of of of training centers and and and training facilities comput facilities resources to to the members that are joining and I think 29 countries joined um you know so and there's about 25 in the pack silica so it's creating two blocks but I so here's here's something that that is interesting is I actually talked to one of the people that was involved in in organizing this and I said look you know wouldn't it be good if if you actually invited the US to join they're like oh no we would it would be amazing if the US would join we would want them to join and in fact they should join and I said

[00:30:01] well but you know if if they join you can't call it you know uh Waco because that's a a Chinese-led organization they're like oh you know if you guys are interested did we would be open to changing the name, you know, we we would be open to to having a a a truly global organization. So, I I I I think this narrative of a us and them and you know, they're trying to take over their world with with their AI. Um I don't really see that. I I 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 super intelligence to actually cause the the CCP to say we have to actually uh abandon uh all of our internal regulations intended presumably to maintain social stability and the supremacy of the existing regime uh as well as uh external efforts to create

[00:31:02] blocks. Put the blocks aside for a minute. what threshold of super intelligence either achievement or technical development or um maybe implications for uh weapons systems, if if that's really what it takes, what technical achievement would super intelligence have to pass or what threshold would it have to achieve in order for the CCP to decide in in in your mental model? Gosh, we really just have to focus on uh on supremacy here. Alvin, if if you don't mind, let me let me intercept and lead into that question, too, because I I think I think we really do need to answer Alex's question. But before we can do that, like, let's understand who the CCP is. You know, in the in the US, it's really interesting when you meet the actual players. So, you've got Elon Musk, Dennis Sabis, Sam Oldman, all saying, "God, I wish this would slow down." And when I interviewed Sam at MIT back in 2020, remember that? He was like, you know, it would be far better for the world if progress was slower, but it just isn't. And uh and so we we have to

[00:32:00] just live within the reality that AGI is is imminent uh 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 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 I 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? Okay. So So first I I want to um maybe um demystify something around around this idea that people think that China has a CCP. There's somebody at the top that just says, you know, you will make AGI and and it just happens. and and and and the reality is that with all the industries and with with all of the innovation that's happened in that country over the last 30 or 40 years, it

[00:33:00] was never a top- down thing. There was maybe directional things. They would say, "Hey, you know, for the next 5 years, we should work on, you know, clean energy and we should work on, you know, automation of of of robotics and we should you add AI to that, right?" And and they did that about five or 10 years ago. And when when the the central party initiate these plans, then the pro 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, you know, give them some, you know, stipens, give them, uh, you know, free recruiting, uh, give them some some kind of benefits, right? And and they will have essentially provincial champions and city champions. And this is and then and then you know the the the 30 plus provinces all compete against each other to see who can make uh companies that that solve some of these problems right so um it is it is actually very distributed in terms of of how how they're uh how these plans get initiated. Nobody's saying okay you need

[00:34:01] to use this technique to go do that and you know you need to share your resources that they're they're actually a very highly competitive um landscape between uh all of these labs right but um the thing that also to remember is that uh almost every single uh leadership in in the senior leadership of the central party are actually engineers probably 80 or 90% of them right so they're they're actually quite technical um >> yeah I think that's one of the biggest ch you know I I had gone to to different parts of China. We 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 US most of our politicians are lawyers and in China most of the politicians are engineers. I 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

[00:35:00] entrepreneurs, you know, that distributed network you're talking about, Alvin, and he asked them, "What's the most important thing to entrepreneurial success?" And he expected teamwork or business plan. He said, "No, it's what the government's focus is next that determines your success." >> Yeah. So, so this is like a when you're swimming, you don't want to stream swim upstream. And what what the government does is that it makes the the stream flow in the directions of the certain areas that they think are are important and and then they let the entrepreneurial nature of the people there and you know 1.4 four billion people and and the most number of of STEM grads in in the world. Um good things happen, right? And that that's that's what's driving um their innovation and their idea is 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 up for the overall population and >> so having said that and we start this conversation on open models again I think it's very important to understand this is the government saying get as

[00:36:00] 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 openweight models >> yeah so so the whole open source strategy people think oh this is you know Chinese strategy to destroy American economics and and you know pop this bubble. the the the reality is that it's an emerging strategy, right? And in fact um you I was talking to friends at at DeepSeek a little bit after they um they came out and you know 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 to to open source but uh the CEO of the company is very uh open- sourceminded and he he thought that hey open sourcing is is is something that I should do because this is a great technology I want to share it with everyone and 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 reopen sourcing it. Um, but because of all of the the uh kind of I guess soft power value, the PR value that came out from,

[00:37:00] you know, having a a local champion, they then became celebrated, right? And then uh essentially most of the companies in in China were following this because it became kind of the de facto emergence standard. uh uh and um and now you know at the last Waco or the the WIC announcement uh Ci Jinping finally said hey we think open source is a good strategy and that that's kind of goes to what Dave's talking about right in terms of when he says that now pretty much you know most new companies are going to be focused on open source because that's the high level instruction right yeah >> if deepseek had been a closed model that succeeded do you think China would have gone that direction is it really just that seed led to this, you know, incredible open- source movement in China. >> Well, I mean, there were open and closed models uh for, you know, the whole time, right? In fact, if you look at bite dance, they have uh the Dobal model, which is a closed model, right? Uh and and you know, their their C dance model is is a closed model. Um and they're

[00:38:00] also quite successful. So, so both both models exist. Um but you're right, I I I I don't know what would have happened. I don't know if the the push for open source would have been as great. Um but the one thing that we also need to remember is that open source was a little bit of a necessity that US policies pushed on them right. Um you had mentioned the export controls earlier and export controls you know you first talked about export controls in in the software export controls of Fable but actually before that for several years now four or five years there's been export controls on on on chips and and allocation of of uh you know EDA software and lithography equipment and so forth. And and what that's done is that it's 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 um essentially pushed Deepseek and all these other companies to to get more innovative whereas the US because they have so much resources they've been much more focused on brute force and brute scaling uh whereas the

[00:39:02] the Chinese have not and open source was necessary for them because by open sourcing now rather than having a lab with a 100 or 200 people when you open source it you know you were saying that it's been 100,000 or more of of the Quen variants. Essentially, the rest of the world helps you uh modify and improve your your models, right? And and that's a great way to to to leverage the the global community of millions of of AI researchers. The other thing that's truly important is inference, right? Uh for you to do inference, you need to have compute. And uh if if the Chinese labs and the Chinese hyperscalers can't buy the compute then by open sourcing it [snorts] uh essentially all the hyperscalers and neoclouds around the world are buying compute hosting these uh these AI models and allows them to distribute um their their models without uh the high capex that the that the the US labs are burdened with. >> Hold on. I I I think we need to get back to Alex's question and I would love to

[00:40:00] to drill in like I 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 kind of caliber and Xi Jinping says if this is so great why are you open source slap 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 Quinn is out and then Kimmy K3 and I think the release of Kimmy K3 will turn out to be as defining a moment in human history as anything that's ever happened. That's that's my prediction. Uh but I think the psychology behind that choice is going to be the, you know, the the news nugget that matters for all time now and leads into Alex's question of what's going to take what

[00:41:00] would it take for a wakeup call. You know, if if it 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 but walk me through any detail you've got on the psychology that changed Xi Jinping's opinion on whether to open source these things. >> Um, so I I think there's there's two two questions. One is are you uh uh you know thinking about are they going to close source because they're worried about AI running away and and and becoming rogue or are you worried about uh competing with the US and saying that whoever controls and creates the AGI uh becomes the the global dominant hedgeimon right so I which which aspect are you want may >> maybe let 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

[00:42:01] Jinping thought on a global stage and belt road initiative and geopolitical competition with a western block on the other and these two different arms may be in competition with each other. The CCP at some point as super intelligence 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. [laughter] Um, so I, you know, it it I I don't really see them uh right now seeing um AI as a way to create u political um domination, right? I I see them as looking at this technology to increase their their economic influence around the world. That that I I I think is absolutely there. Right. Is that because you you you think again just to pull on that because there's a a hidden premise in that if if you're thinking that CCP

[00:43:02] doesn't see political domination through AI, is that because CCP already has domestic political domination and has already achieved dominance over AI through these reported uh ideological exams that AI models have to go through. >> No, I I I think those are those are two separate issues, right? uh the the type of things that that the CAC um has them review is, you know, things like, you know, removing certain uh types of keywords or or ideology or or things like that. And and and those uh types of adjustments in the models only apply to Chinese hosted models, right? So if if if a a model coming from these labs are then put on hugging face, all of those types of guardrails are actually removed in terms of you know whether or not they they they can talk about you know um >> Tman Squareman Square 1989 1989 let's just say >> exactly I mean but you know to be honest that that's because everybody already knows this nobody really cares but they do it more for formality right um but

[00:44:02] they they actually do have other things that they're they're they're putting in in in place in terms of checking for you know uh It is, you know, there there's probably a slightly less um security-mindedness in terms of how much it refuses to to answer questions related to maybe uh you know, viruses or medical other things. Um but I I I think there are still definitely those safeguards that are putting in place and those are getting added uh more and more every day because of these kind of issues. what what I think would would get them to be really concerned uh if they start to see the the the US using this model a as a as a weapon as an aggressive um uh kind of offensive tool right um because then then it becomes okay do we want to you know like you know what the the the mythos models were essentially held back to say hey you know here's a model that can be done and and I think for for some right reasons um you know you want to neuter the the

[00:45:01] offensive capabilities before you put it out to to the rest of the world. So um this is actually a good thing. But um in some cases I would actually think that it would make sense rather than having 50 companies that uh are being allowed is actually to allow most government organizations to have access to this because you really want global stability and global stability means that countries can have access to it. find find the the vulnerabilities in their systems cuz I don't think the Chinese want the American financial system to go down and then the Americans don't want the Chinese financial and their grid to go down because when when instability happens in any big country uh the world suffers right and and smart people understand this um but you know too many I think folks with relatively narrow perspectives think that you know one country wants to actually have another country fail having major superpowers fail uh creates uh irrational actions

[00:46:01] and and societal stability usually is the uh the the the preeminent um priority of most major governments. >> This episode is brought to you by Blitzy, [music] autonomous software development with infinite code context. Blitzy uses thousands of specialized AI agents that think for hours to understand enterprise scale code bases with millions of lines of code. Engineers start every development sprint with the Blitzy platform, [music] bringing in their development requirements. The Blitzy platform provides a plan, then generates and pre-ompiles 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 preIDE development tool, pairing it with their

[00:47:00] coding co-pilot of choice to bring an AI native SDLC [music] into their org. Ready to 5x your engineering velocity? Visit blitzy.com to schedule a demo and start building with Blitzy today. >> Alvin, I want to pull on two strings on this topic before we move on. Uh the first is the claims that Kim K3 and other models were distilled from US uh closed models. Uh your thoughts uh what you know what is being said in China about that? And then the second is the the policies you know limiting chips and limiting access to Fable 5. Uh your belief is those policies were misdirected and can you explain why on that? >> Sure. Um, so the distillation thing I I you know it's it's um I think it's more of a a PR tool that's you know certain companies are using and really right now there's only one company that is kind of against that right and uh if you look at the numbers

[00:48:02] >> which which company you're saying open AI >> no anthropic >> anthropic okay >> yeah I mean they're they're the ones that are that are lobbying the government to say hey we need to you know we're being distillation attacked which is actually a word that they kind of invented um the reality is that every lab uh both domestic and international distill from each other, right? And they and they distill within the the organizations themselves from larger to to smaller ones. Um and in fact, if you look at the numbers of what anthropic put out, uh they were saying that you know 20,000 accounts from three different labs in China and you know a million or two uh questions that were came in. I went back and actually did a a a estimate of what it would cost to do the the the the the number of queries based on the average responses on their highest uh models and it was like two or three million, right? So it's two or three million dollars of of across three different labs and and for DeepSeek I think it was only in the thousands of dollars, right? So, so the the numbers

[00:49:02] sound big when you look at them in isolation, but when you look at them in in aggregate, it it really doesn't mean much. If if you have a model that you spend a billion dollars on and somebody can can can distill and and duplicate with a couple million dollars, then the whole economics of of Frontier AI doesn't make sense. Right? Now, here's the thing that also just to give comparison, uh, every month, uh, Meta spends somewhere between$1red and $200 million on anthropic tokens, right? They're one of the biggest uh buyers of of AI from Anthropic. And if they if anybody was going to distill, they would have been distilling for the last, you know, year to two years. And they just finally got a model out that is somewhat competitive in the last, you know, uh week or two, right? So they have the highest per capita uh 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 from Anthropic. Why can't they have gotten, you know, a a Kimmy K3 thing out

[00:50:01] 6 months ago, right? >> Well, the the the public argument, I mean, this has been widely reported that Meta internally is utterly paranoid of being accused of 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. Um I I'm not sure if I agree with that given how many you know how much they're spending but let me give you an example. Uh what about XAI right they also have access to these uh you know and and I don't think Elon has that same same concerns over uh doing anything to to speed himself up right and only in the last two weeks have they come out with something that is you know relatively competitive right so Elon is an interesting case because he's at this point I would argue a frenmy of anthropic he acquired his

[00:51:00] entity SpaceX AAI acquired ired cursor and cursor was arguably at least recently a a post-trained version of Kimmy uh 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 for Elon and the the Gro series the recent Grock models is in in some sense he has like two 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 just from firstirhand experience >> I 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 Deep Seek Quinn and Kimmy. Um 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 uh OpenAI, but not the Meta or XAI team at the time. Now they've changed teams completely. They've fired everybody and started

[00:52:00] over. So for whatever reason, the Chinese people working on it though are brilliant and far far better than those teams were. >> Yeah. So so 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 you know organizations, right? and they are actually innovating. It's not that distillation was why they're successful. If you look at, you know, how they were able to reduce, you know, their their KB cash usage by like, you know, 20x and, you know, these are not things that you get from distillation. You cannot distill something from somebody that they other people didn't have. So, we we can't take that away from them. Um, you know, as Dave said, they're there smart people there. They're doing innovative things and that's part of the reasons why they're successful. you know, did they distill? Probably. I'm sure they have. But same, you know, the the the US models have also distilled from Chinese, right? I think there was a couple models ago where if you use Chinese to ask anthropic claude uh what model you are and they said, "I'm Quen, right?" So, you know, it and and when I was at the at the Alibaba labs and they were kind of laughing about that, too. They're

[00:53:00] like, "Yeah, they're distilling from us. We just can't tell because they already downloaded our models, so we we don't know, right?" Um but I I I don't think that the whole dissolation uh issue is really as big. In fact, if you look at what happened last week with um with Zuck, you know, he's actually saying this is actually a good thing. We don't think this should be uh prevented and you know the and you know, they're 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 Wizner grow style the past AI is always going to help you create the next AI that's the inevitable outcome >> it's humans do as well help you know create the next generation >> so I think the whole [clears throat] bot traces thing is overblown but 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 a billion dollars getting to the next level the next guy trying to distill from there and get to that same level

[00:54:00] completely separately is about two maybe more like $10 million to get to that same level. And Alvin said like this is a just a a fatally broken business model. I'd love to 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. Uh Selim, >> um Alvin, you said something I want to pull on to which seems to be the theme of today's >> We're pulling on elephants in the world. >> pulling strings all over the place. >> Everywhere. >> You said the US is building a lot more data centers than China is. And I'm finding that very very surprising. I I think the maybe there's a difference here. It's clear China is building up massive energy capabilities, but they haven't built the 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? Cuz I find I found that surprising. Shouldn't they be building a ton of data centers? >> No. So, so they they are spending a ton more on energy generation. They're, you

[00:55:01] know, they're building more new energy, new electric generation than the rest of the world combined, right? About 10x what the 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 and you know because of what's happening you know like with the Hermoose issue they realize hey it's it's really good that now essentially half of our of our auto fleet out there is uh is electrified so I don't need to depend on imported oil. In fact, um that was one of their their key um objectives was to say how do we uh become independent of external energy sources. right now. Uh what they are doing though they are building you know giant um you know solar farms and wind farms on the the west side and in the desert parts of China and then using their very high um voltage power transmission that you know essentially a thousand miles you lose less than 1% of the of the electricity

[00:56:01] when you use these high high voltage things and they're bringing the energy to the east where most of the population is living right and but they are also building some data centers right where these uh the power generation is happening so that you know you don't have stranded power um and and they're able to to to deliver um uh compute at a fraction of the cost of the US because their their energy cost is around two to three cents per kilowatt hour right which is in in some cases you know 10 or maybe you know 15 times cheaper than many parts of the US right so what and I think in the long run that's actually where the the the constraint will be and I think you're you're right um you right now they they don't have enough chips. They they can't buy enough chips and they can't make enough chips, right? Because their their capacity uh is is limited by the the fact that they don't have EU EUV machines, right? So So when I was in China and every lab I talked to, I said, "Hey, do you guys have have

[00:57:00] no computer?" Like, "No, this is our our biggest issue. We don't have comput." So So from a a um export control, are we slowing down China? Yeah, I I think that that export controls of chips is slowing down China. Now the one thing that that you most people don't realize is that the the actual training right now that is happening is not even happening in China because they don't have the the the blackwell generation chips in China, right? And so they're actually doing it in international data centers, training it and then bring back on a disc or something, you know. So so uh it's it's >> Wait, say [laughter] that again. >> That's that's incredibly important information. >> Oh, yeah. I mean, well, this this is actually not a secret. I mean, I think people in in in both >> 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 I I think it's it's it's more uh you know uh optics than anything right now, right? >> Oh my god. >> So So from a from an inference perspective, most of the inference is being served uh to Chinese people in

[00:58:02] China and that that is an area where the limitations of 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, um, so the export controls, yes, uh, it it has slowed down China. It's made their life more difficult, but it also has created the the necessity for innovation. So, so back to what what Peter was asking earlier, um, you know, was expert controls good or bad? I I will tell you this. Um the the the day or maybe like the week after we uh stopped the uh the Chinese from buying all of the um the high-end uh chips from uh from the from the [clears throat] US. I had, you know, calls with a few of my friends who were in the semiconductor industry and every one of them got calls saying, "Hey, you know, would you like calls from the government saying, "Hey, would you like some extra funding? Would you like extra resources? how can we help you accelerate uh you know could we get you

[00:59:00] customers and you know I I know a few of the CEOs of of these of these uh Chinese GPU companies and they were saying nobody wanted to buy our stuff you know we're two or three generations behind we were less energy efficient but now we we can't make enough because every data center in China has to buy our stuff and so we we would have died if it wasn't for American uh policies so so essentially we created the the current competition of all of these the the morres and cambercons and beerans of of of China would would not be as successful or maybe would have gone bankrupt if it wasn't because of American policies and now within the next you know two or three years they will start to catch up to to what 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 uh us forcing them to survive >> well what a backto-back double whammy that is though I mean we knew the ship embargo was was misguided and it's going to be one of any government misguided things in the next couple of years. But the idea that okay, first it forced

[01:00:01] 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 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 three terabytes and you just transmit it back in an hour. >> Y >> and so so that moves all over the world like you know like a liquid. It's everywhere instantaneously. So the chip embargo is completely and utterly backfired and misguided. >> Yeah. And unfortunately right now if you talk to the folks in DC, they're doubling down on this. Right. right now they want to keep adding and making these things more difficult. How can we get the foreign um the international data centers that are being used by these Chinese to not be you know accessible and they're they're adding

[01:01:00] more and more layers and more KYC and and I think that those are the kind of things that actually will backfire more economic more geopolitically than economically because then that forces uh irrational behavior like they're now trying to to keep us from progressing as a nation you know we're then then behavior gets more aggressive and they become more defensive and I think these >> 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 like if it turns out that Kimmy K3 level or one model later is capable of full RSI and then spirals to the singularity you know just just hypothetically in that scenario Alex's question is what wakeup 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 we've made a horrible mistake here if we release this next thing to the world. I I think if if they starts there are credible multiple labs coming

[01:02:01] back uh safety labs coming back with testing to show that these AI systems have their uh intent um you know once they've released to do things beyond what they were instructed to right I mean you guys have been talking about these rogue AI escapes but they weren't really rogue AI escape they were instructed to escape and they were put into a prison and they were incentivized they were incentivized to escape I would to go and find the answer. >> And they were they were saying use whatever tools you have. And and by the way, they also led had left open doors for these things to escape and because of of improper settings or maybe some of them were intentionally leaving uh holes for them to find. Uh and they were given tasks that were impossible to solve unless they escaped. Right? So so we we um we forced these AIs to do what they they were doing and and they're they're creative uh systems, right? because that's what their job is. Right now, if if if something went and if if these AI systems started to show that even though

[01:03:01] you didn't tell them to do these things, then they started to do all these, you know, sneaky subversive things and they started to to hack other systems and and create their, you know, their own, you know, I think some something you guys have talked about, you know, using crypto to to then uh, you know, grow money to then to then buy more servers, to then grow themselves. If they start to do that, I I I think uh I think the governments uh on both sides of the ocean would uh be much more focused in terms of how do we protect ourselves from a a runaway AI. >> So Alvin, the White House just put an embargo on Chinese robots. >> We've talked a lot about robots here. Uh I was surprised. Uh I think it's a move that, you know, reduces US competitiveness. What's your thoughts there? >> Yeah. Well, I I I so the thing is robots today u 90% of their components are coming from China, right? Uh if we put an embargo on them, in fact, there was rumors now that was funny that now ch US

[01:04:01] robotics companies are now sneaking to to China, buying these components, putting them into suitcases, and then bringing them back, right? They're they're now they're now sneaking the other way. You know, the Chinese were were going around and buying GPUs and sneaking them back. Now the the American uh robot companies are going to China and sneaking back components and actuators that they couldn't get in in in the US, right? So I I you know I think we need to understand that the world is interconnected. We are highly dependent on each other. This is this is has been a the globalization concept has been something that's been going on for the last 100 years and I don't think we can stuff that genie back in the bottle. Right. I I think it's great that we want to create domestic uh independence the same way that China has. they they spent, you know, the last 10-15 years, you know, building out their their own capabilities, building out energy generation, you know, building out uh telecom systems so that they wouldn't be dependent on on third parties because they've seen uh how reliant on on you know, one or two countries and when

[01:05:00] policies change it could hurt them, right? And so some of the things that we've done really have um allowed them to and gave them the motivation to to be as strong in terms of of taking short-term hits for for long-term independence. >> We 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 uh you know sports competitions. Can you give us some background? What's what is the undercurrent of robotics in China? Uh what's the government trying to incentivize? Is it the one child policy that left them short laborers and they need workforce? >> Yeah. So, so I think we need to separate the kind of bigger automation question from the humanoid robot explosion, right? Um there is 150 you know plus uh I think when I was at uh at WIC there was over 200 robotics companies that

[01:06:01] were related to humanoids that was demonstrating stuff which is crazy because the the total volume of human robots last year was in the tens of thousands right globally right and and and I think 80% of them or 90% of them came from China from like two or three companies so really there is no market right now for that many companies to to exist. exist and should exist. But um you know this was also the case if you went to WS a year ago there was about 150 labs that was demonstrating large language models and now there's really 10 that are probably relevant right so in the next year or two uh you'll see that 150 go down to you know probably you know a single-digit number of surviving um uh human and robot companies and the automation's already been happening right because people know the demographic issues that that you're talking about the the one child policy issues. Um and in fact right now you

[01:07:01] know more than half of of of of industrial robots are deployed in China. So they're already doing this right. In fact if you go to many of these factories today they're they are called dark factories because there's essentially you know a few people running it and almost everything is automated. Uh so the the need for having humans in manufacturing is becoming less and less. Now the the one thing I do want to point out is um humanoid robots are actually not really good form factors for doing much of anything right now, right? Uh I you know I was just I was just at the Unit headquarters and their factories and I did a tour and and there's like almost all of our customers are research labs buying our stuff and and also some that are doing demos and you know doing you know you know kickboxing or you know things like that. But 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 um all you know all the other

[01:08:00] >> figure exactly like Figure right. I mean I think Figure did some kind of a a demo of them sorting sorting uh packages for 10 hours or something. But yeah 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 and in fact, when I was talking to the unitary guys, they said uh right now they're moving to a a uh only upper torso uh model and those are actually selling better in the commercial space because having feet is actually uh a negative because you have to keep balancing it. When it falls, you know, bad things happen, right? And these things fall apart and you have to maintain them. It's actually having less components and having a big base with with a big battery, it lasts longer. It just there's all of these benefits of actually having a a non-legged um humanoid versus a legged humanoid. >> Humanoid form form factor is terrible. >> Yeah, I remember [laughter] [01:09:00] like enough enough with the self-loathing. >> So, um but I mean you know we we we've evolved because of of you know billions of years of of biological evolution. I mean with it it we uh we have constraints but the machines now can be designed for the new form like just like a plane is not the same way of of moving through the air as a as a a bird or an insect. Right. So anyways >> take us back to the world AI uh cooperation organization uh event um in the 2026 World AI conference. you were there with President Xi was announcing Waco the World AI cooperative organization. Uh you said like 26 nations have signed up. Uh what's the undercurrent there? Uh and and connect that with the upcoming US Chinese AI conversation in DC on September 24th. Um and you're advising I guess the US side

[01:10:01] of the equation here. >> Yeah. Yeah. So, I mean, you know, I I'm I'm part of a large team of other folks that are that are contributing to that and and I I think the the the sentiment at the at the WIC was that hey uh you know, AI has its moment has arrived, right? When when the president of a country comes to a conference, that is like the biggest honor that you can get for an industry. Yeah. He he doesn't go to very many conferences. I think four or five years ago he went to the world uh internet conference and that was in in in Woodsen which is a little bit outside of Shanghai. uh and that would that signaled okay the the internet has arrived right so um so I I I think this what this means is that uh you know more and more companies in in China will start to think about um how do we integrate this technology into into our business right and you know uh two two years ago a year and year and a half ago they came out with something called the AI plus plan I'm sure you guys have probably heard about it but

[01:11:00] >> talked about it on the pod >> oh perfect yeah so essentially the I mean their their idea is hey within the next 5 years we want to have 70% of companies integrate AI into their their business whether it's you know manufacturing or education or medicine or so forth and within the next 10 years we want to have 90 plus% right so there's the specific goal and it's all about the fusion and deployment into into industry and society um this is a little bit different than the the American AI action plan right that came out last year and the the American AI action plan is it's on the supply side. How can we create the best models? How can we dominate in the best chips and uh and but it doesn't talk about what happens after, right? So, so I think that the the two countries have a very different focus in terms of their AI plan. There was not there was nothing in the the AI plus plan that says we have to get to AGI that we have to dominate this. It's 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?

[01:12:00] That was all they cared about. I I'd love to develop this a little bit more from my perhaps jaded perspective. I I look at Chinese industrial policy and I look at Wong Huning who for for those who are not tracking is sort of the CCP's chief ideologue uh author of many of the policies uh or at least primary author maybe you can correct me Alvin of policies signature policies like belt and road initiative and so on and I I look at Chinese state capacity like the eastern data western computing mega project to to put compute and power in the west where energy is cheaper. and and more available and put the data in the east where the mega cities are and I just look at Chinese state capacity on the one hand and then on the other hand I look at the west where historically again maybe you'd have a different position where in the US historically at least in the postworld war II era we've had relatively by comparison weak industrial policy. It's only relatively

[01:13:00] recently that the US has the US government has decided that having a a strong and centralized industrial policy is a good idea. So question put putting this in question form for you Alvin if you buy to the extent you buy any of those premises if if you could be uh supreme leader of the US or the western block or the pax silica for a 5-year plan for the west what would your five-year plan be for the west to leapfrog China's AI plus and other Wong huning style ideological 5-year plans what is it that we need to do Um so first let me let me kind of um >> there's a lot there. >> Yeah, there there is a lot there. Um I I think underlying your question there's an assumption that hey you know there there there is a ambition to to take over the world by building all these technologies. Um I I think this is this is one of the I think biggest misunderstandings that America has that they they well let me let me finish that

[01:14:01] they are mistaking mistaking uh anxiety for ambition. >> Right. Let me explain that. Um, you know, the the the Western world has um has a history of expanding. You know, whether you're talking about, you know, Greece or or Rome or or Pax Britannica or right now Pax Americana, right? Um, we've expanded. Uh, the Chinese actually haven't. They they haven't had this idea. they they've essentially been in that little sphere of that central space and and you know they used to call themselves the 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 theqing 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 um their industrialization they started to build you know summer palaces instead of

[01:15:01] instead of navies and then the the um you know the the the the eight powers came and essentially took over China for 100 years right that that history has left a very deep mark uh in this the psychology of of the Chinese who say >> 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 innovate and continue to to to learn from from the the rest of the world. Um and you know from I think any country's perspective you know that's probably we what we all want including the question that that uh you know Alex has said how can how can America also learn from that to say how can we become strong independent and and resilient country in fact uh America has actually moved away from you know post world war ii we were 50% of the manufacturing capability of the entire world right and at that point uh the world dependent on us right now what's happening is around 35% of the global

[01:16:00] manufacturing capabilities in China about 15% in the US right so so uh and I think the forecast is it's going to go to 40 or 45% over the next 5 or 10 years so >> in China >> yeah in in in China right so um you know this is the thing is America right now is very good at financial services creative services consulting services you know uh things that areformationally driven and things that are actually highly susceptible to to AI exposure for displacement Right. So we are right now running this race to get to AGI which is the the force that will actually dispel displace us from global preeminence because we are commoditizing the very um sectors that we are strong in in the world. So so this is this is this is something that we need to be very careful of why are we running this fast to to to go to something that actually creates major disruption and instability. uh in our country.

[01:17:00] >> Well, I I can answer that one for you and then I'll I'll repose my same question back to you, Alvin. So, the I think the answer is that the the goal of capitalism fundamentally is to burn itself out. Uh and the the irony here, right, the the the goal is to take what's scarce and make it abundant. And right now to the extent that human services labor is scarce and we see that with Balm's cost disease the goal of super intelligence one of the goals at least or or instrumentally convergent sub goals is to make the equivalent of human service labor abundant make it too too cheap to meter as it were. I think that's the goal and not some sort of like stasis or equilibrium where it remains scarce. we need to separate um kind of national strategy from just um capitalistic uh philosophical bent, right? And and in fact, you're right. I I think that you if you look at capitalism and the ultimate destination capitalism is a single company monopoly of the world, right? That that is and and that is actually a very unhealthy thing. So, so here's something I just heard from David Saxs and Gavin Baker uh

[01:18:02] and on their I'll end two days ago. He said Gavin [snorts] Baker said uh there's been conversations in anthropic where Daario's told his team that in the 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. Right? I 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 is good for America, right? Or for the world. >> I I agree there. I mean, I just don't see how that gets there. [laughter] >> Yeah. No, I I realistic agree. I but I but I think that that the mindset that that he has right now and and this is this is the issue is that we are essentially creating national strategy based on the aspirations of a couple of companies today. >> I I'd like to though Alvin just pull back to the original question which is you get to be strategy sorry you you get

[01:19:00] to be Wongghooning for the west a as it were. It it sounded like what you were saying is your your thesis is that super intelligence is is is going to disrupt the the western right now economic dependence on service labor and I think the implication was that manufacturing is a more stable fixed point for long-term economic vibrance is am I reading you correctly that your positioning >> your positioning would be basically western re-industrialization. >> I I think and and re-industrialization is is definitely needed. um you know the the the US workforce right now is around 70% uh is white collar workers right and white collar workers as you guys all know is the the first to be displaced by by AGI when it when it arrives China is around 40% Africa is probably in the 10 to 20% right so so at different parts of the world it will be affected differently and um hard manufacturing industry is something that um even when

[01:20:00] we have AGI we will still need those type of facilities. So, so I I think it makes sense absolutely for for every country in the world to have uh some level of of ind indigenous capabilities. >> Um now, but the other thing I I I think is important to understand is that the long-term workforce uh redistribution is not going to be going back to manufacturing. We're not going to create, you know, 300 or 180 million uh workers going into manufacturing. That's not what I'm saying, right? >> [snorts] >> uh in fact with automation that's coming uh that number will go less and less just like what happened with farming. We used to be, you know, 80% of the country was farmers. Now it's less than half a percent. Right? That but it's it's been fully automated and we're more productive than than we've ever been, right? Uh in in the agricultural in the agricultural space. Um I I think what we will 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, um you know, elderly care, uh you know, just

[01:21:02] things that require humanto human services. I think that is the labor pool that will be able to absorb the the 60 or 70% of of displaced future workers. >> M if you're Mackenize employee, you know, start getting >> if I was a a mid-tier, low tier Mackenzie employee, I'd be very worried right now, right? I I' I've talked to to partners at at consulting firms, at at accounting firms, at at lawyers, and they are all looking and saying, "Hey, we actually don't need these junior guys anymore. We can do, you know, just as much work, in fact, more work faster with, you know, a few senior guys and 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. Don Musalem, our chief medical officer at Fountain. Heart disease has been personal for you as well, hasn't it? It really has, Peter.

[01:22:00] When my daughter was five, my husband died of sudden cardiac death. And so this is a topic that is one that I am missiondriven to try to eradicate. Prevention first and early detection is absolutely critical. 50% of people die of heart attacks with no warning signs. >> 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 angography with AI analytics we are actually finding that 88% of people coming in have detectable coronary arter 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.

[01:23:00] >> So listen, make sure you understand what's going on inside your body genetically and 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 Salem 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, Daario 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 fullbore 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 they're

[01:24:01] really useful and powerful but there's definitely not dangerous. Just go out and and let them out the door. That's the incredible range of opinion between Daario 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, right? Which means he's the that one company. That that's part that that that scares me. Um now in fact what you just said is is really important is that there's a perception that uh these models are going to get more and more dangerous the bigger they they they get. Um, so this goes back to a paper I just released last week, um, called, you know, uh, bigger models are not the most dangerous, right? And it was released on on cipher brief, which is a a national security outlet in DC that is, you know, read by most of the national security population. And what that data did was I went back and I looked at uh, across the board uh, from from uh, you know, security use, national security use cases, from biological, chemical, cyber use cases

[01:25:02] for 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 10 to you know 50 million parameter models for chemicals that were creating, you know, uh war uh kind of chemical warfare weapons, right? you had um you know 10 to uh probably one to 10 uh one to 50 billion parameter models that were able to create viruses and and and uh you know genetically engineered uh beings, right? Or or or harmful agents, right? And then and then you had essentially kind of one to 100 billion parameter cyber models that was as dangerous or more dangerous than the the the leading um >> fable, right? mythos and mythos, right? And in fact, uh the harness actually now for for for cyber is more important than the models

[01:26:00] themselves, right? So, uh so Mdash from Microsoft has came out with I think a cyber gym score of 95 and Mythos was 83 or 84, right? And and and what what Mdash did was it took a hundred little tiny models and just organized them together to to use different skill sets, right? So, so I I I think we we need to we need to understand that um the danger is already there today, particular in the the the the the bio chemical side of things, right? Um and and nobody's talking about it. Nobody's really working working on protecting the world from those agents because you know a a a 10 million or a 1 billion or a 5 billion parameter model will run on your laptop in your basement and you can you know design these genes or or design these chemical chemical um weapons. But what's what now needs to happen is how do we those are already out right there. A lot of those are open source. Those are already out. How do we make sure that

[01:27:00] the precursors are are controlled that that these systems are are creating, right? How do we make sure that the synthesis machines that will generate these designs into real uh genetic material are are controlled? You know, these are the kind of things that that should be higher up in the agenda and and and it's not right now. >> Yeah. I Dave, you said something really important about uh 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 and fable. Do we see that at all? How does the how does the CCP think about the safety of the models being open- sourced? 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 they they do testing in addition to some of the propaganda testing that they're doing. You know part of the CAC regime in testing is for safety. In fact, um the

[01:28:02] the UK AISI is their AI security institute um uh have just came out with a new report last month and what it showed was the the cyber threat uh capabilities of the leading open models including Kimmy was uh about a half of the capabilities of the Mythos and and uh GBT 5.6 so models in terms of their their cyber attack capabilities, right? Um, 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 their highest level, none of the the Chinese models were able to autonomously attack and control uh a a external system. Right? So, uh I think out of the 30s something levels, the average American uh models were able to get to 20 25 in terms of how far they they they went in terms of attacking a a network. And I think none of the Chinese were able to get to the the higher levels and

[01:29:01] a very small number got to like the lower levels like four or five or something. So, so from a from a a threat security perspective, these models even though they're just as big as as uh you know some of the the the leading US models in terms of size of parameters, um they're actually less dangerous. And one thing I do want to point out is you know what you said earlier was uh from a defense perspective you actually want bigger models as a defender because for defense you have to look across all 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 it's a it's a very asymmetric uh um equation between attackers and defenders. You you you you don't need very large models to attack, but you need larger models to defend. The the flaw in all of that analysis to me is that AI is like a match. And you know,

[01:30:01] you can say, okay, here's Quen 40B. Look, it doesn't burn. Then you take Kimmy K3 and you light it and you're like, oh, this is a match. Okay, now it's burning. And 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 it ignites, and then it's unstoppable. And to me, Kimmy K3 is a burning match. And and so, you can analyze it eight 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 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 Kimies running tomorrow. 5,000. I guarantee you it can improve itself. Well, I mean, I think that the the whole world right now is talking about RSI and and I think we need to

[01:31:00] separate the concept of, you know, whether or not something can improve itself versus the the danger that that a a larger model uh uh poses versus a small model. That 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 you can look at the prompt prompt logs and and from a from a government perspective you can actually manage it right and and most of the larger models are run with harnesses that are being managed by the cloud providers so you can add added levels of security and and and detection which was what um you know that's what separates mythos from fable is the harness that says hey you know don't do these things right so it it neutered the the mythos system this is why in in my in my paper I talk about why larger models are not necessarily the most dangerous. There's a raw capability and then there's a a effective deploy capability and larger models have actually a lower effective

[01:32:03] 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 uh 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 where we're seeing radical disruption coming to very traditional industries across the board? And are they seeing the same thing in China? >> Yeah. and and and so this is actually um an interesting in fact yesterday I I was just on a call with um one of the leading uh AIdriven medical uh drug discovery companies and uh 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

[01:33:00] that are in industry are tiny models they're they're you know tens or you know maybe tens of tens of or hundreds of millions of parameters or or maybe a few billion parameters in fact if you look at whether drug discovery or legal use cases or you know like open open evidence I think was was based on uh GPT4 was what their their their system is based on right um you look at um Harvey which is the leading uh legal use case um system it's based on GLM 5.1 right so I mean and and it was it was upgraded it used to be based on some other open source model from you know two years ago so so they're they're not um they're not uh using the latest and greatest. So, uh, in in April this year, um, I wrote a paper with Eric Benosen at the digital economy lab and it's called the, uh, enterprise AI playbook, right? We went and talked to hundreds of companies, found 50 that successfully deployed them around the world in 10 different countries, 10 different

[01:34:00] sectors, and what we found was that the technology was not the issue, right? in in in 80 90% of the cases it was all organizational issues of why uh 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 the AI that we have today is already good enough to solve real world um business issues um around the world right so um you know and you're you're an organizational guy so you you realize this I mean this is this is the whole um you know a curve uh issue is that you know the the technology takes time to get absorbed but when it does it at the end it goes up this curve right >> there's a there's a study by McKenzie's that in all the AI deployments and 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

[01:35:02] limitations of the architecture they're working Um given this is a USChina 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 is um a a country where a top- down type uh management model is much more prevalent right u and and also the the concept that you know the the government will actually help protect us. People seem to appreciate that more. I'm sure you guys have talked about the case where, you know, there was multiple legal cases where the courts in China actually ruled in favor of the employee who sued and said, "Hey, you can't fire me because the AI took my job. You need to find me another job." And and and the the court upheld that, right? And that those type of precedents um incentivizes uh people to say, "Okay, it's okay for me to to adopt these technology without being afraid of being displaced." And that's not the case in in America. I mean, we have an at will

[01:36:01] employment system. You know, you saw the uproar that happened at um at Meta when when Zuck wanted to to, you know, key lock every single one of his employees and and people were like, so I'm going to train my replacement, you know, screw you. Um and and I think this is also why if you look at the the current negative sentiment in the youth today, I mean, you know, every every uh student is having tough time finding jobs and they're they're just very worried, right? And you know I'm I'm spending a lot of time in universities and so so does Dave. I mean you can see the you know the the anxiety that is uh within the the youth because of you know their difficulties in finding internships or or postgraduate. I mean MIT is one of the best schools. They they probably have less of an issue but uh I I guess every school that I've been to students are concerned. >> Is really really unique actually on that front. But if you go to other very great technical schools like Nor Eastern or Harvard, there's about 10 or 20 percent

[01:37:00] AI adopters on campus and a violently opposed 70% 80% and this really really big cultural gap and the AI aware people are just busy talking to their agents and and interacting with each other and there like forget it. I'm not even going to talk to the other side of the school and 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% is the populace is against AI. What's going on in China? I had that conversation with Michael Katzios. So what is the US doing to try and and flip this sentiment because it's destructive? Uh why is China so pro AAI at the citiz at the citizen level? >> So so here it is. Um you know over the last 40 years uh people have seen their lives get better and better in China, right? You know, you've heard the whole, you know, 78 or 800 million people have been risen out of poverty and blah blah blah. >> The Chinese miracle. >> Yeah. Yeah. Yeah. The Chinese miracle. And and a lot of that is being attributed to to technology uh uh adoption, right? And and and innovation.

[01:38:00] And a lot of that technology was not invented in China, but it was adopted in China and and it just spread. And then people see, oh, you know, last year we didn't have uh you know, I don't know, some some 5G. Now we have it and and our lives getting better. And so they they 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 the news that gets spread, you know, the the one thing about China is that that they have a lot more control over the media, >> right? And and they they're mostly good news all the time. It's it's kind of you you're talking about the crisis news network. Yes. >> They're they're essentially the the always good news network, right? And so >> there we go, Pete. Peter, that's the policy prescription right here. Peter, you need to to start the the Western equivalent of of CCP, CCTV, >> top down news control. >> No, I mean, if you look at what's happening today, what you guys are doing is essentially the equivalent of the Chinese media system in terms of >> Alvin. Alvin [laughter] [01:39:01] Western CCTV. >> No, no, no. In terms of like in terms of 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 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. [laughter] >> 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 our 100 billion neurons, our 100 trillion synaptic connections. And if we're if there's fear-mongering 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 blood. I mean, >> you're kidding. >> No, no, no. So, so this Yeah. >> Oh my god. Quinton Tarantino can't go one minute without showing blood. Well, [laughter] this is the problem is that every everything has to go through a

[01:40:00] essentially a censorship board that said 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 [laughter] blood you know so in games right so it's it's u you know it definitely every part of media today uh is is managed in in in China so um you know I I'm not saying I mean I'm endorsing it but from from the perspective of what um what Peter's saying is, you know, why do people have a more positive view on the world and on the future is that, you know, they they see a lot more good than bad, you know, [clears throat] >> amazing. >> But I I want to peel back the propaganda just for a minute. So we at at the same time to to the extent that the west has visibility into changing governance and cultural mores in in China as as a result of automation we see the rise of Tongping uh lying flat in response to 996 work weeks. We see other maybe call it reactions to the the increased

[01:41:01] automation of Chinese 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 it sounded like you were saying well we need more nurses and more humanto human care that's the the end state as it were but it's being reported that in China in response to increased manufacturing automation we're seeing the the rise of call it a a a gig class like that's the end state in China where everyone becomes a gig worker who's being displaced from factories so I I I would love maybe Alvin If we could just peel back the the self-curated propaganda from the 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 I think you know the issues you're pointing out is that definitely there right the the youth unemployment in China is probably uh around 20%. Right. So the the youth unemployment in the US

[01:42:01] is around 9%. So, and the overall unemployment in the US is around 4.3%. So, you know, this is why the the youth feels uh very um uh disenfranchised is because they're not, you know, getting jobs, but it's actually worse than that. There's around 42% undermployment. So, for for college grads, if you're a college grad, you're actually working as 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, there's essentially half of the of of college grad are are not getting jobs here, US or China. >> In China, here. >> Yeah. No, in China is 20% unemployment in the actual no jobs is 20%. This this this is why there's that that Tping the 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 they're they're you know their whole uh parents and grandparents are all putting their hopes on on this this one generation. Um

[01:43:00] and when they get out in the real world it is hyperco competitive and now they they have to go do these you know 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 and that is that is a a an issue. Uh you know, this also kind of facilitated the the uh kind of um uh online uh kind of influencer market that that grew for a little bit for the the the the micro streaming and so forth. So, you know, there there there's a lot of issues there. Um and I I don't pretend that they have the solution. I I think this is something that requires really a lot of other countries to all work together and figure this out. And how do we transition uh you know more and more of the workforce into jobs that that will be less uh exposed to to automation whether it's it's physical automation in factories or it's cognitive automation that's happening in in offices. >> I want to take us back um you know for the rest of our uh time here together to

[01:44:02] the upcoming US China conversations. Uh I think it's very important. It's going to it's going to influence everybody's life here in one way or another. Uh [snorts] you argue that the US is playing a prisoners dilemma when the actual game that should be played is a stag hunt. >> Um if you could I'm going to show your slide here. Explain what a stag hunt is and what you think you know how USChina uh you know AI relationship should evolve here. So, let me show that slide. Let's 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 there are a lot of vegetarian Chinese uh Buddhists, etc. So, hopefully this is a vegan stag hunt. [laughter] So, so actually before I even talk about seconds, I know most people understand the prisoners dilemma, but but um the idea is that two guys are are in prison and you know they have to defect on each

[01:45:00] other and that's the on a single turn prisoner's dilemma. Uh the optimal thing is to 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 first. 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 and run this race." Um, it's it's >> that would be defecting. So, essentially, right now, the expectation is the other side's going to defect. So, I'm going to defect first so that I I get hurt less, right? That that's the the prisoner's dilemma single turn game. Now, the the thing is the world is not a single turn game. The world is actually a multi-turn game. And even in Prisoner's Dilemma, a multi-turn game, the optim the game theory optimo is tit fortat, which means uh you do you start with uh uh actually cooperation and if they defect, you defect and then you essentially signal each other and a long term you actually both go to cooperation. Right? Now in in a stack

[01:46:00] hunt game, it's it's actually and and the prisoner dilemma essentially is a zero- sum game. Right? In a stack hunt game, it's actually a positive sum game. And whether you both defect or you both cooperate, you you get a a a stable two nash equilibriums, right? And what that means is that you know those can actually stay in perpetuity whether you you both uh decide to defect or not. Now the difference is we're right now playing the prisoners dilemma game where we defecting. But if we're in a game that is that is the uh the stack hunt. The stack hand was actually something that John Jacqu Rouso uh invented which is the idea that two two 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 two people to to hunt together and to bring it back if I if I go hunt by myself the 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

[01:47:00] family for a week or a couple days and if I get the stag we'll both feed our families for a month, right? That's that's the idea of the the stag hunt. And right now, we're actually doing the worst thing. We're doing the we're going to go for the stag and China's going for the hair, right? They're going for the good enough AI, the the the the one that helps the the economy today. Um and we're going for the the giant AGI, the thing that's going to solve everything, right? Um, and when you do that alone, what what that creates is the worst situation, which is the 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 zero. Whereas, you know, China continues to to to do their their little gains and continue to survive. So the the optimal solution for for a stack game is actually first both slow down, both make good enough stuff, you know, get to a point where the the the technology is helping grow the economy and now we sort

[01:48:00] of understand how to manage it and then together go hunt the stack, right? That that's the the the optimal strategy for uh the relation to how you know AI and game theory works. But we've we've put ourselves into this game theory of prisoners dilemma where we think just defect just effect and and it's so it's a it's a self-imposed game. So we're playing we're playing the wrong theory and you know using the wrong strategy and playing the wrong game right now. >> 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. I think that's a mirror of what you're just saying. >> Yeah. >> China is playing now. China's playing the hair game. >> Yeah. China's playing the hair game. They're saying, "Hey, I I don't need to make the AGI. I just need to make good AFI goes into my industry. I I then take that industry, you know, export it to the to the world." You know, this is what the whole bell and row initiative

[01:49:00] is saying. Hey, you know, I'm going to make 100 there's 150 partner countries in the bell and roll initiative where they're they're shipping telecom systems, energy systems, you know, transportation systems, schools. >> How would you make the AI how would you make the AI ecosystem, the American AI ecosystem indispensable? What would you do for that? >> What I would do is actually uh create high quality open-source, right? because then then you're competing uh on a even >> ecosystem to ecosystem. >> Yeah. Eco e ecosystem ecosystem. Right. Right now China is open source. So so the question is do I pay $50 per million token or do I pay the cost of electricity? And for most people, you know, based on what you just earlier, they don't they don't need the frontier, right? 90% of people are fine with today's models. I mean, especially when you have things like Kimmy GM and, you know, Deep Seek 5, you're you're you're already, you know, at a level that is higher than the needs of the average person, right? >> So, if China hadn't forced the issue by

[01:50:02] open sourcing, suppose there was just OpenAI, Gemini, Anthropic, uh, XAI, all four US companies. >> Uh, would you then say the same thing that that the best thing for America to do is high quality open source because China's forcing the issue? I I I think that uh well I mean first of all we can't roll back history it is what it is and once once it's out now you know and especially with what you said of you know these models are improving themselves now that you have these models improving themselves um I think it will not be the duopoly that we have you know we we're going to see uh you know middle east and France and Japan get into this game to say hey I I can make a smaller model that is is you know 95% as good right >> I do agree maybe just to say something nice about China since I I guess I've been playing the role to some extent of China hawk in in this conversation. I I I would point out curious Alvin to to hear your thoughts on this that the the present situation where at even as of a

[01:51:00] few months ago I think the west was at risk of succumbing to uh regulatory captured duopoly of anthropic and open AI dominating the future light cone and then not unlike maybe by analogy theqing dynasty where the US could have sort of turned inward on itself Chinese openweight models obtained however uh have basically forced open the US, call it a reverse ching, uh, and now finally we have real competition at the AI frontier, thanks ironically to Chinese competition. Do you think we find ourselves now in a reverse ching? Um yeah in in in in some ways in fact I I I think it's maybe the more um appropriate analogy is actually we're we're if we roll back to time I would roll back to the cold war where the US and the USSR were competing on a arms race and essentially the reason we won was not because we we you know sent a missile and blew up the Russia right or

[01:52:00] Soviet Union is because they bankrupted themselves building um military carry arms and you know up to 20 15 20% of their GDP was building arms that was not creating real value for for their society in some ways uh this is what China is doing to us right they're spending onetenth as much on data centers uh and and you know getting to 97% as good um and what we are doing right now you know leveraging uh you know hundreds of billions of dollars right you had last week uh Nvidia announced that they have a $500 billion deal with black rockck and carile and and Blackstone and so forth to to essentially securitize uh um uh chips and and and compute, right? This sounds a lot like the the subprime uh issues. >> Wow. Wait. So, so, so this is the most astonishing thing and I this is also to your credit Alvin the first time I've heard anyone basically analogize the the credit I I don't want to say bubble but

[01:53:00] the enormous amount of private credit that the west is allocating to to compute analogizing that to a reverse SDI Star Wars moment that could presumably what what you're gesturing at is that could lead to the Chinese the proverbial Chinese century and the collapse of of Western dominance. Is that the the thesis? >> Well, I I mean I I I hope it doesn't happen, but I I think we are we are pushing ourselves in that way. We're actually right now acting like USSR and and see what what what perpetuated the the arms race was this missile gap, right? And the idea that oh, they have more missiles, we have more. And at both cases, they were both having the wrong numbers being provided to the leadership. They say we need to build more because they have, you know, 30,000, we only have 20,000. how they go and it became we had 70 or 80,000 missiles between us that would have blown up the world you know hundreds of times what there was no need for any of that right and in some ways we're kind of doing the same thing right now with

[01:54:00] um with AI where we right now 45% of the US stock market value is in AI sector right that is a very very fragile place for us to be uh at the height the at the height of the internet bubble, I think around 30% of the stock market was um was internet companies, right? I don't know if you guys heard of something called the uh the Buffett indicator, right? The Buffet indicator is something that says a a market is healthy when your stock market uh is the same value as your uh GDP. Okay. uh and uh at the height of the internet bubble we were around 120% of the GDP was the stock market value right now we are at 240% of the GDP is the US stock market value right in fact the AI sector alone is worth more than the GDP of America today that that to me is a sign that we

[01:55:01] are in a very very fragile place and a economic correction is due I'm not saying it's going to happen tomorrow but Um, you know, uh, Buffett's a pretty smart guy and he's been doing this for a while. >> Are are you saying that the US is, uh, is the the Soviet Union, the USSR in 1988, or are you saying that the US is Japan in '89? >> Uh, well, I mean, I think they're they're they're two different. I I actually think that we are right now um the overinvestment that we've put into infrastructure for for AI especially when we both what we just talked about that it it is a an industry that will become commoditized not saying that AI is not amazing it's going to do amazing things but the companies who are investing it uh are not going to be the ones that profit from it right and and that is going to create uh a major instability in the economics of the country and if we don't manage it um it could create a a a major crisis of

[01:56:01] what happened in the USSR during the late 80s. >> So you think we're overvaluing the frontier labs and undervaluing the sort of China AI plus type rest of the economy that should be the applications. >> Yeah, I I I think that's I'll I'll take the positive side of this. you know, the the deployment velocity in China is actually quite a huge gift because it's forcing policy makers here to to solve the real bottlenecks, permitting, energy, manufacturing. So, it's that part is at least the good part. The bad part, I think, is what Alvin you've talked about where we're operating like the USSR and some of this industrial policy and it's not going to sustain. On behalf of my Moonshot mates and myself, I'm inviting you to join us at our inaugural Moonshots live event on September the 25th in downtown LA. Alex, Salem, Dave, and I will be hosting 1500 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

[01:57:01] 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 uh advising the US Treasury and uh the team that's going to the to the she negotiations or uh conversations on September 24th. What how are you advising them? Um I I think the key right now is that we we don't need to get to a solution on day one, right? Um I the the success factor of this discussion is not that we come out with a a massive framework that solves everything. And and by the way, I'm I'm this is just my personal representation, not a representation of of anything that's being discussed in in any of the other um organizations that >> disclaimer is noted. Yes, [laughter] >> I'm sure Alvin, you and Jacob Hellberg must be besties at this point.

[01:58:01] >> So, um, but um, >> no comment. >> Yeah, I think that's that's the the the thing is if 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 in the past I think people um there was a discussion in 2024 the dialogue um and there was a lot of disappointment because um you know China didn't didn't come with all their technology people and they didn't we didn't come up with a solution and so they're not really sincere in in the dialogue and I I think you need to understand just like you know Chinese works on decades for their strategic plans they also take a lot longer to prepare uh when they're doing these um kind of um uh diplomacy discussions, right? Um and you know, for example, for for the the May visit, uh there was really no discussion on any of this until the day before uh you know,

[01:59:00] between US and China and that 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? You know, what do you want to talk about?" Right? And the fact that we're now at least, you know, a month in advance of that having discussions, I think it's it's a good thing. It it's it's it's a start of of more proper dialogue. You know, when when Kissinger was was, you know, uh doing a lot of these, you know, crossber discussions, he would be there months in advance to talk to the Chinese before they had the visit with Nixon. Right. And >> as you're talking to them, as you're talking to them, can I suggest something? >> Yeah. Because I think it feels to me like the US is focused on having the best model whereas the real power will coming from will come from having the best ecosystem. >> Yeah. >> 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 I think they're they're you know the the discussions right now are really more around safety, right? Because the the ecosystem versus model thing is a

[02:00:01] competitiveness of how do we how do we become more competitive as a country or you know have have greater uh influence or capability. Um really the discussions that 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 you know AI is not being used by bad actors to to create instability around the world um that AI itself is not um you know potentially creating uh harm to to the world on a on a longer term right and and I think that's that's the and and you know there is also the kind of underlying idea of you know uh nation nation to nation um uh kind of aggression, right? And I I think, you know, those three different things are are are all being balanced. I I would say that the higher priority today uh would actually be the bad actor um you know non-state actor risk which is what what Besson said you know when he was interviewed the day after that discussion because he realizes that

[02:01:02] [clears throat] nationto-nation um aggression has been in balance you know between superpowers for eight decades right and and and that doesn't change uh with AI right and and in fact um you know Because if you if you use AI to hack into somebody's network and then you take down their power grid, uh that may give you a one or two day or 5day advantage, but then there's asymmetric responses to that, right? And and people realize this. So people in the actual national security space, you know, 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, so non-state actors is something that everybody should be worried about because it is going to happen. Uh, it's already happening. I think the the the the ransomware and cyber attacks is up two or three uh 100% in the last year or two, right? So, and then it's going to be even worse because of what we've seen

[02:02:01] when you start putting a thousand agents all trying to attack a network. At some point, they're going to find the hole, you know? So, so we need to find that that that sheer risk requires sheer 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. That's >> the good news could be that the the 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. >> Uh Alvin, that's exactly why I love your stags uh 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 five 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 you're taking that into the conversations with with China on the on the 24th. >> Yeah. >> I mean, it's existential on those diagonals. >> Yeah.

[02:03:01] We placed oursel in that diagonal. I think that's that's you know, it's a self self-imposed harm right now. So I 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 of all of USV China. And it would seem to me one of the the cruxes at the summit and otherwise is Taiwan and TSMC. And I I'd love Alvin your perspective. How does this end? Does this end in in in your in your geopolitical analysis? Does this end with China attempting during this geopolitical and demographic window to invade Taiwan and seize TSMC to gain leaning leading foundry [clears throat] 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 I think there there's there's there is an assumption right now that some people, you know, in in uh DC are

[02:04:01] 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 they're going to, you know, go and attack the island." the the the the reality is that if anybody attacks the island, there is nothing there to be had in terms of of um workable facts, right? Uh so I I I probably shouldn't be talking about this, but I I was having >> No, you that means you definitely should be talking about it. I was having breakfast with the uh the CTO for TSMC and also a former uh senior official from the CIA and and the the the TSMC guy goes, "Hey, I heard that you guys going to blow up our data centers if China attacks. Is that true?" And the CIA guy says, "Hey, I can't never confirm nor deny that." But what we do have is we have a thousand engineers of yours that we know we will fly out before anything happens. Right. So what what what America cares about right now is that

[02:05:01] they want to make sure that they can duplicate these capabilities to fabricate the latest ships in America. Um you know if anything happens, right? And this is part of what the chips act was. And that's so I think I think there are some 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 uh put into domestic um manufacturing for for semiconductors. I I was with Intel and IBM and we were at the time the global dominant player in in in semiconductors but over the last you know 20 30 years we've lost that we've g given it away right um now if China actually does attack um Taiwan they will not get these fabs and they realize that right and if you go there fabs by themselves um require materials from all over the world requires maintenance requires chemicals requires uh supplies And if if they did that, even if they

[02:06:01] don't blow up, even if the US doesn't blow up the data centers or or the or the the Taiwanese don't sabotage their own systems, after a little while, you run out of these supplies, right? They they realize that the reason China cares about Taiwan is not because of the FAPS. It it is absolutely because of a political history and and you know this, right? in the yes essentially in the 19 49 the the the movement of um >> of the Taiwanese uh the nationalist party to to Taiwan and and uh that to them is a a a uncompleted civil war and um you know the the the the two countries actually right now is recognized by America including 190 other countries as being one country right so this is kind of like uh Hawaii and the US or or maybe Puerto Puerto Rico and and the US, right? They're they're kind of pseudo part of an a a a a one national um uh structure, right? And and it is more of a political uh and

[02:07:02] and and um and to them a civilizational ending to a a long story. That's the the main focus. And they've multiple times every you know, ever since essentially Denia now have talked about peaceful reunification, right? So, so I I I don't think there there is a a interest uh or a rush to do any uh near-term attacks to try to uh get Taiwan because of chips. I I just don't see. >> So, you don't you don't think like there there's like a back 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 US maybe two to three more years to migrate leading edge node fab capabilities to Arizona or otherwise redomemesticate TSMC's capabilities and then China. Okay, fine. You you can retake Taiwan because we don't care anymore. Um so I I don't know if you had a a chance to read my uh my paper the the

[02:08:00] the great reckoning and the reconnecting but it it talks about um Taiwan in some aspect to say hey um just like in in during the 2008 great financial crisis actually China helped out the US 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 uh if China actually started to sell T- bills versus buying uh it could have completely destabilized the American financial system, right? And they kept buying. They kept buying at trillions of dollars. Um which which uh helped to keep interest rates down and so forth. we potentially might have a repeat of this situation I if there's a correction in the market uh due to what's happening right now with the overbuild and the overlever uh of the the AI sector right um and maybe at that point uh the the Americans uh or maybe Trump will will give a call to she and say hey can you help us out again um and you know maybe

[02:09:00] I I'll just be more handsoff or be more more clear instead of the ambiguity issue and and you know this me completely speculating. >> Wow. This is your this is your war game. Just just for for clarity, what what I hear you saying in your war game is sometime in the next two years, there's a private credit bubble that the US is using to finance its data center buildout. The bubble, assuming it exists, pops and then the US asks China to help financially in return for what a quid proquo regarding Taiwan. Well, not not not in the sense of here's Taiwan, but to say, "Hey, um, as long as you agree to some kind of a peaceful thing and over, you know, mutually agreed uh term that we're going to stay out of it, right? Because we we've been very involved in uh in the the kind of Chinese political or the Taiwanese political sphere for a long time. We've been selling weapons to them for the last, you know, 40 or 50 years, right? Um and at one point we used to have you know uh soldiers based on in Taiwan. Um

[02:10:00] we still have advisers right now, military advisers uh based in Taiwan, right? So so um you know this is like saying if if uh if Chinese were selling weapons to Puerto Rico and and they were they were you know helping fund them uh what would America do? And and look at what happened in 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 uh 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 for for in any negotiation discussion to understand the other side >> there. There's important for everybody to understand that uh the US 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. Uh where do you how do you think this next 3 four years goes? What are the two

[02:11:03] couple of big paths that we think we you think will kind of we have to pick one or the other? Uh how do you see this next few years playing out? >> You talking about for just the AI space in general, >> AI and the global transformation. Yeah. [clears throat] So, so that's actually the the the whole um narrative in that that Great Reckoning paper is is how the next few years plays out. And uh what I what I foresee is that we will soon find that um these AI companies uh their you know once once they go public or if they go public um their financial will become much more clear. People will start to realize that the value of the the AI innovation does not necessarily acrue to them. it it may in the near term, right? And then it has they're they're you know, one of the biggest beneficiaries, but um it was also that acrruel came from a period when you didn't really have the the open-source capabilities that that we have today. Um and in fact,

[02:12:00] if you look at the the the recent disclosures in terms of where uh Anthropics revenues are actually starting to flatten out a little bit, it's not, you know, earlier this year, they were growing like, you know, 10x over just, you know, a few months, right? And now they've essentially flattened out at the kind of AR in the 70 billion range. Although even though their AR is a 70 billion, their first two quarters was I think right now total less than 20 billion in revenue, right? But they're they're committed to hundreds of billions in camps in debt. Uh you know there is right now $1.6 trillion of offthebook debt uh of the major uh hyperscalers today. 1.7 trillion. Okay. um during Enron days there was a $200 million of off the book debt. Okay. So just to give some some context of the scale of the kind of problems that that we are we are looking at and um if if that happens um I think people will actually slow down uh the the construction because the

[02:13:00] construction right now is all based on the idea that these companies will continue to make money, continue to be able to fund uh and and and service their debt. Um you know if you look at at Amazon and uh and open AAI revenues when they talk about AI most of that probably more than half of it comes from two companies right so that is not a very uh diversified revenue base uh and as more and more of the capabilities uh move to open source move to move to edge computing um the the dependency on cloud-based premium services will continue to to erode I think that >> we'll yeah >> we'll shift the bottleneck down the stack to compute, electricity, power, etc. >> Yeah. And >> 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 m sort of provincial revenues because the

[02:14:01] provinces are are really selling off the real estate or had been selling off the real estate to generate their own local revenue. uh h how is it I mean I don't want to overanalogize but I isn't it isn't there sort of a striking parallel between Alvin you're pointing to the west maybe overleveraging compute uh and data center infrared development and China perhaps overleveraging or overindexing on for humans real estate development >> so so actually you make a really good point and I think they they did the hard thing right over the last three years there's been about a 30% deflation in in total real estate value in in China, right? And and they managed it in a way that it was not a crisis, right? We we need to do a a soft landing for these things so that it does not create a crisis, right? So I >> we're building the houses for the AIS and China was building the houses for ghosts. >> Yeah. Well, I mean, no, I think that the

[02:15:00] the ghost town thing, there may be a few, but the reality is that the the home ownership right now is something like 70% in in China, and it's probably less than 50% in America, right? >> So, um, so, so I I I I I don't think we want to, um, over, um, I guess parallel these two things, but I I think what we can learn is that when the crisis happens, you need to be willing to take some near-term pain. And they did, right? they they they took major hits in their GDP slowdown. Uh you know they were growing at you know 8 9 10% and now they're growing at four or 5% uh per year GDP mostly because the real estate sector stopped growing. In fact it's it started to to decline and they had to make up for it with other types of of uh of industries. Right. So global policy folks are talking about this China [clears throat] this managed uh crisis that they've done as a as a hallmark case study on how to do it in the future. >> Yeah. Let me let me ask you about the managed crisis actually because China >> Guys, we're going to take one last comment. Dave, last question and then we

[02:16:01] got to wrap it up. So 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 >> So China is clearly a country coming into a crisis because of the birth rate. um you know the the one uh child per family is catching up in a huge way. The the population is aging like crazy. It's it's it's a crisis and that's why the country is so focused on robotics because they're going to need it more than more than anyone. Um 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 the this is at the height of the cold war. The US is over here, the Soviet Union's over there. And it's a p it's a prisoner's dilemma. And so as nuclear weapons get more and more efficient, inevitably the prisoners dilemma gets more acute. Sooner or later, one country or the other is going to have a a ability to destroy the other country with no retribution whatsoever. This is this is going to destroy the

[02:17:00] world. In reality, you know, this is where I lost faith in poly classes. In reality, it didn't play out that way at all. Uh, 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 and now we have Ukraine and Russia, you know, both part of the Soviet Union in a 5-year long catastrophic, devastating war. And the other satellite entities don't even speak, you know, Russian. You know, they're off on the So, I don't have any idea. What is China like? Is it is it truly unified like the United States? Is it fragmented? Yeah, I mean I think this is one thing that China um has that's very different than a lot of rest. It's very homogeneous, right? It's probably like 95% of the population is Hanchinese, right? And and everybody speaks Mandarin. And you know, they may speak other local dialects because there there are, you know, hundreds of local dialects, but they all speak Mandarin and they all uh their written script is

[02:18:00] the same across all the all the different pro provinces, right? So, uh, I don't think we're going to see the type of issues that you saw, uh, with, um, the USSR, even if there was a a major economic crisis. And, um, I think they they've managed it very well. In fact, um, they've learned a lot of lessons from from the disintegration of the Soviet Union to to say we cannot let that happen because that would mean hundreds of millions of people would suffer or die, right? and and um you know that that is their biggest priority is is social stability uh political stability economic stability. I I I should just add you maybe the fine point the the wagers uh ethnic Muslims ethnic Turks may differ with that assessment uh regarding ethnic homogeneity and the the unity of of approach sort of everyone's Han type characterization of China for the record >> I said 95% right and so so I I think they're there definitely a few percent but it it is a relatively minority right

[02:19:02] um and and even even the the the Weaguers or whoever, they all speak Chinese, they all read Chinese. So, so there is a common language. Um, and you know, I I I I think that um the the desire to to succeed uh is not as uh prevalent as we um tend to portray it in in in US press. So, >> we'll have to save that for next time, I guess. >> Yes, I'll save that for next time. I'll say one thing. Uh I spent a few months traveling around China and I my conclusion was the native entrepreneurship in the Chinese people is higher than any other country I've ever seen just latently and and therefore if we believe entrepreneurship is 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. Um I do I do want to end with one thing is um go ahead. >> For for I think for America to actually

[02:20:00] uh be successful in in this industry and and take advantage of what we've created is actually to think about creating something that's akin to a um AI Marshall plan. Right. So the the the Marshall plan post World War II we spent around I think 15 to 18 billion dollars rebuilding um most or much of of Europe and some parts of Asia. And that created a ally uh >> created markets for our goods also >> created giant markets for our goods because you know at that time we have 50% in manufacturing um and it created loyalty and and and you know allies for for eight decades right we we need to be thinking more about that type of thing today uh but on with uh with AI data centers with AI technology the same thing that that what China is actually doing right now their their Waco the you know AI corporation organization essentially is is their AI marshall plan. We should be either doing something like that or we should be working with them to do it together,

[02:21:00] right? >> That's great advice. >> Yeah, if if we did that, uh we would have a big market to sell our chips when when we stop building data centers here, which we probably will at some point when people stop able to finance it. Um we're going to need to sell those NVIDIA chips in other places, right? So, uh, we're we're going to need to, uh, sell services. We're going to need to, you know, there's a lot of good things that that can be had. And we also hopefully will then have a place to to, uh, for the the open source models that we create. We start to create frontier open source models, safe ones that both both countries agree, both countries start testing, uh, and and have standards for. um then then this 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 you 95% of China is ethnic Han and we the US needs a Marshall plan but the US has this incredible advantage

[02:22:00] in that there's no single ethnicity of America. It's a complete grabag 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 slaver cleansing. They'd much rather work with the United States that if we just give them a chance. >> Yeah. And we're telling the world the opposite story right now, right? [laughter] >> There there's the advice that we'll end on. Stop shooting foot. Um All right. Uh Alvin, it's been awesome to have you on. I I I speak I think for all of us on the pod and the viewers when I say uh it's really great that you're in the middle of these discussions. So uh push your ideas as hard as you can. We'll do the same on your behalf. Um we'll definitely would love to have you back again sometime. And on that note, uh thank you for being with us and we'll wrap it up for today. Right.

[02:23:00] >> Thank you all and Alex will be back next time questions, Alex. So, >> you know what? Someone has to ask them. I 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 a good humored recipient of the balls and strike calls. >> No. Okay. Thank you. >> All right. Don't don't go away. Don't go away. Don't go away. Don't go away. [music]