Hivemind：市场现状、Saylor 的终局与中国 AI 增长

Nothing said on the hivemind is a recommendation to buy or sell securities or tokens. This podcast is strictly for informational purposes only and any views expressed by anyone on the show are solely our own opinions. Not financial advice.

1. State of The Market

Hello everyone. Welcome back to another episode of the hivemind podcast. We were off last week, partly because the founders were in China doing a little AI trip. We're going to talk about markets and that trip a decent amount today. As always on the podcast we have Yan Ventures partner. We have Jason head of markets at Delphi. And we have Jose head of labs, head of Delphi labs. Obviously myself Cedrus head of research at Delphi and today we're also bringing on Rossin who does AI investing at Delphi Ventures. And hopefully you've read a bunch of his stuff on Twitter. I always post really insightful bangers. So today is Wednesday March 25th 10:45 a.m. Bitcoin is up to like 71.4K now. There's been like a bit of relief over the past few days. You know there's these rumors of ceasefire and everything.

How are people feeling about the markets here? Maybe Jason, you want to start?

It's a tough question. These markets are really tough, especially when you get headlines every day or day and a half, which whipsaw things. It's definitely not a market you want to be in. If you are high-leverage perp trading this market, it's got to be tough.

It's also hard to have a real good understanding of what the impacts of what's already happened are going to be over the next 3 to 6 months. I'm clearly not a geopolitical, oil, or energy expert, even though energy is a core input to everything throughout the economy. Trying to gauge the actual implications over the next 3 to 6 months—with disruption, production, all of that stuff—even if there is a ceasefire, there still are going to be ramifications for that.

I'm still trying to figure out and underwrite what that is. To me, it seems like equities are still pretty priced to perfection right now. I still feel like equities haven't really priced in a lot of what's going on and what's already happened. You've already seen it start to manifest in a couple of other areas.

I don't know. I'm still super cautious right now. I don't really trust either side at this point. I'm looking at the market to give me an idea of what it's thinking. Oil's still elevated, right? It's not really moving too much after these ceasefire headlines. It was down, obviously, initially on Trump's tweet yesterday.

I don't have a great read on it. I still think there's probably a higher likelihood of downside in the short to near term than upside. If I had a bunch of things that were down a couple of percent and we rallied, I'd probably sell and look at it as a nice gift to get out closer to break-even, then reassess the situation. It's definitely not clear how this is going to play out, in my opinion.

It's a tough one. It's a really tough market environment. I'm curious what everybody else thinks, and also, with Rossin being in China, what the view is from people over there. I only really get the view of Twitter and New York, so I'm curious what the views are from elsewhere on what's going on.

Yeah, I would say my news diet is still fairly U.S.-centric, and a large majority of it is from non-domestic Chinese sources. But you do see some viral videos and posts. There was one that went viral—I think it was retweeted by one of the affiliates of Chinese state media—that was a cartoon of Persian cats basically making fun of Trump for being fairly erratic and causing problems where it wasn't necessary. But they would find ways around their traditional paths to energy.

I think this is one of the reasons the Chinese are investing so heavily in a lot of their domestic inputs and supply chains, realizing that in a globalized world, it's increasingly going to be difficult to depend on traditional trading partners, and they don't want those vulnerabilities. They need their own semiconductors. They need to invest really heavily in renewables and nuclear.

They are quite dependent on imports of oil and food, so those are real dependencies that I'm sure people in the Chinese government are quite worried about—knocking 10% of global oil offline that they were getting at a discount. I don't know exactly what the views on the ground are. I'm sure they're pretty diverse, but I would assume folks are generally unhappy with one of their sizeable energy import partners being in such a chaotic situation.

I think, specifically in crypto, it's pretty promising to see Bitcoin not nuking while Saylor isn't bidding with STRC proceeds. Perhaps the reason is that people are pricing in him turning that faucet back on. It seems like that game can go on for quite some time if you look at the dividend liability and the offsetting amount of Bitcoin you can buy with each amount of issuance.

The idea there is that he's basically selling these preferreds in the market and guaranteeing 11%. As long as the market keeps believing in his ability to repay that in the long run—which he definitely has the ability to do, because he can sell other assets, namely Strategy shares, to supplement it—that game can go on for quite some time. I think that's going to be the next, or continued, catalyst on the BTC side.

That will return next week with the dividend payment going out. It's a bit different from what you see with normal stocks going ex, because you have this ATM component.

On the alt side, you have these spurts of strength, but then you get this periodic washout. I think it's pretty evident that there's not that much money sloshing around, because you have BTC strength and then a handful of alts doing well. It's just been the hot narrative at the time.

You have TAO with the All-In guys shilling it, and then you see some other alts catching a bid, namely the AI ones in Venice and Grass. You're still seeing a lot of strength in Hyperliquid as you get more alternative assets trading there and it becomes less crypto-centric. Outside of that, it's been a pretty meager market.

I think the equities component is tricky. You definitely have somewhat of a slowdown in the U.S. consumer, and I agree there's quite a bit priced in. But we've seen the market hold up regardless of consumer strength. You can maybe expect that to continue, but at a certain point, I think that can realistically break if you have continued weakness.

The question is just going to be whether there will be some kind of stimmy activity going into the midterms, or whether something will be done on that front to create an environment that's more favorable when people are heading to the voting booths. But I agree, it's tricky.

I feel like Bitcoin's in a decent spot with how much it's been flushed out and the fact that it's holding up while the bid is offline. I think it'll be telling to see how much that continues. If you don't have any massive escalations in the situation—even just now, when Iran came out and said these demands are very far away—equities have actually held up well too, right?

What you want to look for is strength in the face of bad news. I think there's some signal in that.

What do you think the actual end game is for Saylor?

I don't know if you want to take it in another direction, but I just feel like this is an obvious place to be hedged and not take too much risk. Again, maybe I'm happy to be wrong here, but it does feel—now less so, because it's more consensus—that markets still haven't dipped that much.

It feels a bit like COVID, when there were 1,000 cases and our Twitter bubble was insanely loud about it, right? It seemed obvious to the Twitter bubble if you were following biology, or even Selkis back then, and a bunch of people who were talking about it. But outside that bubble, no one really knew about it.

Something with potentially dire, asymmetric consequences doesn't really seem priced in. Even if there were a ceasefire announced, markets are down 2% or 3%. I think the S&P is down less than 5%, for sure.

It feels like a place where I would rather buy once this stuff is fully cleared out, because I'm clearly not an expert on this. I think no one is really an expert on geopolitics. Everyone I've ever listened to over the years has been deeply wrong and deeply certain that they were right at the same time, which is a deadly combination for geopolitical experts—Peter Zeihan being the poster boy.

There are a lot of second- and third-order consequences from having the Strait closed for this long, from these oil facilities getting bombed and the natural-gas facilities getting bombed. We don’t have much information, but it seems like some of this might be offline for years. The Russia plant got bombed yesterday or the day before, too. We’ve never seen this kind of level of supply destruction, at the same time as the closing of the Strait of Hormuz, which affects fertilizers and oil, and oil affects everything else. It seeps into CPI and makes rate cuts way less likely, obviously, and I think it affects risk on the margin.

So, to me, it seems like a pretty clear spot where being defensive makes sense and just buying higher in a few months, once this stuff clears—if that’s the case. I’m kind of in line with that.

I mean, yeah, I’ve slowly moved in that direction over the last 2–3 weeks. I started off coming back from PR. The day before, I was looking at Polymarket, and the odds that we would strike Iran by the end of March were 9% or something. The next day, we’re at war with Iran. I’m like, all right, well, nothing ever happens, right?

Slowly, I’ve started to make this progression toward what Joe laid out. I have long-term holdings, core portfolios that are multiyear things that I’m not touching. But in terms of everything else—short-, medium-term, tactical, whatever it is—it’s very defensive for me. I just don’t have a great read, which means I can’t justify being super risk-on.

The intuitive thing for me is what Joe laid out. I’m not an expert on this stuff, but if the core input to everything—literally everything in economic growth, from food to pretty much everything—is disrupted this badly, and we don’t have a clear view on what it looks like to get this stuff back online or what those timelines are, it could get worse. More things could get bombed. I can’t justify being super risk-on at all.

To that note, what would I want to hedge my stuff with? I wouldn’t actually want to hedge with BTC, for all of the reasons Yianni laid out. I’m not trying to hedge with Bitcoin and then have Saylor come in and dump $10 billion in market buys over the next month or something. I’m just not thinking that’s ideal. You have to look at equities, and equities are clearly the main risk asset that hasn’t priced in at least what’s already happened, in my opinion, let alone what the actual timeline for this stuff coming back online could be.

I do think oil is probably going to be higher. Oil isn’t something I would trade; I have no edge in oil. But the things that oil impacts seem pretty clear. I don’t know if oil is going to level off here, at $90, $100, or $70. I don’t know. But what I do know is that it’s most likely going to be higher than it was before.

So I’m thinking, if oil is generally going to be higher over the next 1–2 years, what things should be higher because of that? You don’t really have to make a huge directional bet on where oil ends up, just that it’s going to be higher. That’s the way I’m trying to think about it and reason out some tactical positions, at least. It’s a super-tough market environment, for sure.

I was just going to say, I think in the 1970s, the oil embargo took 6% or 7% offline, and the Strait of Hormuz is easily 10%. Obviously, what happened in the 1970s was pretty catastrophic in terms of inflation and energy prices. I think it was offline for 6 months or something. The U.S. also has a lot less space from a fiscal perspective to deal with it this time around.

I think one difference is that I was reading something today or yesterday about how much less oil is actually responsible for GDP versus the 1970s. It’s obviously going to be bad, but in a pure 1970 situation, we’re not as dependent on oil as we were then.

What’s also U.S. production capacity now versus then?

I think it’s quite a bit higher. Now we’re roughly 2 times the second-largest producer. We’re at around 20 million barrels. The U.S. situation has changed a lot, but globally, there are still a lot of people who need oil for their economies to function.

It’s notoriously inelastic in terms of demand and supply. If you have an extended period of time where you take 10% of oil offline, that would be quite destructive to demand across a whole host of other areas.

2. What’s Saylor’s End Game?

Can we talk about Saylor quickly? What do you guys think the end game is here? Jeff Dorman had a good tweet on this. Basically, somebody needs to lose, right? MSTR shareholders think they’re safe. STRC owners think they’re safe. The other ones in the stack think they’re safe.

This STRC thing he has is basically a soft peg to $100, right? Then he changes the dividend yield to get that back to $100. But he doesn’t actually have an obligation to continue pegging it to $100 or even to continue paying out the dividends.

I think the dividends are also very tax-beneficial. I think they’re distributed as return of capital, so they’re actually tax-free, or there’s something like that. If you look at the dashboard that Flood posted the other day, the tax-equivalent yield of STRC is actually 30% or something right now.

How can it be 3 times?

Because I think they’re saying that the yield is all tax-free, and they’re backing into what the pre-tax yield would have to be to get the same after-tax yield.

Wouldn’t it be a 30% tax rate? It would be 14%, not 30%.

No, no, no, you’re right on that: 14.5%. I have to look into that more deeply. But I don’t know what the end game is here, because the only way this works forever is if Bitcoin’s CAGR is 30% a year, basically. That’s what Saylor believes in, in everything.

I don’t know. Part of me feels like maybe there comes a point where he decides that he has enough Bitcoin, and then he just rugs one part of the capital stack. I don’t know. Something has to lose.

His target is $1 million. I forget what he’s at now. It’s like 750k or even more than that. In that sense, he’s not terribly far away. This is always a concern, and there are a handful of ways this whole thing plays out.

You can argue that, in the current scenario, it probably increases the probability of success. You don’t necessarily want to be constantly buying new highs, because eventually that comes back to really bite you, and that’s been his MO for some time. I think the probability of failure is actually higher if you keep buying new highs, because ultimately, when you get the pullback, you’re in a much worse spot.

That’s kind of what we saw with him now, where his average price is roughly where we are now. I think dunking on that is dumb, because the real comparison is Strategy’s price over that time period, but that’s a different story.

If you’re still bullish on this asset, then the probability of success actually improves if he’s able to deploy at this lull and the bid comes back more broadly, versus the bid disappearing as he’s just breaking out and basically creating new tops each time. But, yeah, I agree. One of these things has to rug.

You probably see the dividend go away first. Then, if there are liquidity issues, Strategy probably takes the biggest hit, because selling BTC is obviously really tricky—unless he telegraphs it and says, “We will sell this finite amount, and then nothing else, because after selling this, we’re good on our debt coverage.” Then it becomes, how do we try to monetize this balance sheet of Bitcoin, get yield off of it, and grow things that way?

There’s a world where something has to lose, but there are a lot of things that lose before he has to sell Bitcoin, basically.

If you look at this dashboard—great dashboard, Bitcoin client—he’s basically just become super focused on STRC recently. It’s the only thing he’s really issuing outside of some at-the-market offerings. If you look at the volume over the past few months, he’s predominantly issuing STRC now.

That effective yield is 25%, which makes more sense. They’re taking the 55.2% maximum tax rate and the dividend of roughly 11%. You can see why it’s juicy to investors, but it also feels like, if you’re holding this, you’re thinking that you’ll eventually be able to get out in time.

Obviously, if he stops honoring any part of the capital stack, the mNAV premium will go down.

He won't be able to issue anything new, but if he hits a point on Bitcoin that he likes, does he really care that MSTR trades at a discount? An mNAV discount? If he's the majority shareholder and kind of owns all the Bitcoin in there anyway through MSTR, maybe that's partly the end game in his head: once he gets enough Bitcoin, or the perpetual-motion machine kind of runs out, he's okay with everything else failing and MSTR just trading at a discount.

But yeah, I don't know. That's the scenario where I think that's the one I want, because that's the one where Bitcoin doesn't get really impacted. You stop getting the inflows, but he's not selling Bitcoin, because it doesn't seem like he would ever sell Bitcoin to cover the dividend payment.

Right? That kind of thing is antithetical to the whole point of this whole game. So, as that dividend expense gets bigger and bigger, it's something to keep watching. Obviously, this is one of the major market movers, and it has been over the last few years.

I think, at least in the short term, it'll probably only be accretive, considering the size of them, but we'll see. If he actually is able to issue $20 billion more of this, then it becomes serious. I'd be surprised if he's able to—

Surprised if he's able to, for sure, for sure. But even $5 billion more is still a lot. If he's paying 11% right now, it's a lot of cash. That interest expense can get up to $1 billion a year at some point, right? It's a lot.

3. AI Growth in China

So yeah, that's markets. I don't really have anything different to say from what you guys have said overall. I think crypto has looked good relative to everything else, and crypto does have a history of bottoming first. That doesn't mean this is the bottom or anything, but it seems like a lot of people who wanted to de-risk their crypto did most of that.

We'll see. Obviously, if something really bad happens, that changes things. But yeah, I'd have a lot less conviction in BTC right now if it wasn't for STRC. Yeah, yeah. Let's go to China, because we're already 25 minutes in. How was China, guys? Talk.

So, yeah, how was China? I think, Russell, you want to start? By the way, you were a lot of the instigator for this. You've been sort of shilling China to us for a while, and we've been talking to a bunch of managers there, getting excited about the ecosystem. Maybe you want to kick us off.

So I guess it wasn't shilling China; it was just living here, being on the ground, and seeing the density of AI talent, deep-tech talent, and robotics talent, and the dissonance between that and market sentiment toward China after I got here in early 2024.

It was the classic, “China's uninvestable. It's not a market economy. They're cracking down on all of their internet companies.” Obviously, that did happen. Then there was a massive real-estate bust, so it seems like the Chinese economy did go through 4 or 5 years where there was massive deleveraging, and a lot of foreign capital fled, especially institutional capital.

But then in 2024, I was here on the ground and started talking with people. Obviously, the AI revolution was happening in the US, and I think sentiment was just a little bit disconnected from the actual talent on the ground, which was some of the best in the world.

Over the course of 2025, that started to shift a little bit with the DeepSeek moment, then obviously Chinese humanoids, and then the rise of Chinese open source. Instead of just reading talking points as they come through in the news, which tend to bifurcate between “China is uninvestable” and TikTok videos saying that China is taking over the world, there's actually a lot of nuance.

There are things that are structurally very difficult about China, but there are also a lot of positives in terms of human capital, infrastructure, supply chains, and so on. Actually coming to visit, meeting with entrepreneurs on the ground, and meeting with investors is the best way to collect those nuanced data points for yourself.

I was excited to host the group as they came to meet a good number of folks that I've been chatting with here for probably the last 6 months, when we started getting a little bit more serious.

But yeah, as Jose mentioned in his note, over the last 3 months in particular, valuation has started to accelerate quite dramatically. You've seen the Hong Kong markets obviously starting to get very excited, particularly around AI and robotics names. A lot of early-stage funds that we've been looking at are seeing their portfolio companies marked up by some of the sizable growth funds in China.

I do think the ecosystem in terms of valuation is evolving quite quickly. If you look at indices across the board, it's still materially below the US, and I think there are reasons for that. There are less-deep capital markets, regulatory overhang, and geopolitical overhang. But at the end of the day, I also think there are basically 2 markets that have a full vertical AI stack.

One of them is the US, which obviously has a very robust and deep capital market, and a lot of those names have been bid up pretty tremendously. Now, increasingly, there's China, which is number 2.

Obviously, it's a complicated market, and there are reasons why you should be cautious, but it's also one that's worth looking at and evaluating for yourself before just using secondhand hearsay to come to your own conclusion.

Yeah, by the way, I got an email from Danjuran[?] that the Hong Kong Stock Exchange became the biggest in the world in 2025. I didn't realize this, but it was 119 new listings and more than $30 billion in IPO proceeds, which apparently is more than both the New York Stock Exchange and Nasdaq. That's kind of crazy, and you could definitely see that on the ground.

There was something very crypto going on. I mean, Zhipu and MiniMax are the 2 big model companies there, and they both went public and are trading at $40 billion and $50 billion, basically, on less than $100 million of revenue for both of them. That's kind of wild, because it's 5 to 10 times the multiple that OpenAI or Anthropic ever raised at. I would definitely pay more per dollar of revenue for Anthropic or OpenAI than I would for these Chinese model companies.

It's very frothy, and then you're seeing the private markets raise using these public marks as comps. Obviously, the public markets are pre-unlocks, so it's very crypto. I think my crypto experience helped me dodge it, actually, because I was getting pretty excited at one point about some of these private opportunities. But it's tough when the public mark is pre-unlocks and you never know when you can go public either.

I also think they're just the only 2 ways to bet on Chinese AI right now. The pure plays are Zhipu and MiniMax. I think once things go public, it sort of gets diluted.

What I would say is, software-wise, I was kind of underwhelmed. I wrote an update on Twitter, and I think Ross is going to put one out, too. I'm curious to hear what Yian thought of everything, but I was underwhelmed by software. I was a bit underwhelmed by the founder quality overall.

But I was really impressed with hardware. I think it's really hard for anyone to compete on hardware with China, given some of the stuff we saw in Shenzhen and some of the stuff we heard. The iteration cycles are much faster when you have 70% of your inputs coming from the Greater Bay Area, and pretty much everything comes from China.

We saw a lot of robotics companies and a lot of consumer hardware. I think it's very hard—if you're building that, not building it in China is probably going to be pretty tough. It's an obvious conclusion, not particularly contrarian, but I'm curious, Yian, if you had any takeaways.

Yeah, no, it was interesting to see the consensus from the perspective of builders and investors there around the ideal archetype of what they're trying to invest in, which is a combination of Chinese hardware-manufacturing know-how, talent, and engineering talent, and then trying to service Western markets.

In order to do that, you need someone Western on the front end and Chinese on the back end. Even all the hardware that's being built in the US is still sourced originally from China. So, exactly what you're saying: you need someone there to shorten the iteration cycle.

There's broad consensus there that the Chinese consumer market is pretty weak for a couple of reasons. One thing I was pretty surprised to hear was the degree of youth unemployment, and particularly that it's probably 2 times the headline number.

Yeah, that's right. A guy said 25%, but looking around, everything says that's probably overstated. So let's just call it 20%, with the idea being that it's actually probably understated by half, because so many people are just going to grad school after university because they can't really find positions.

And so, from a consumer perspective, you have that, and then you also have the low propensity to spend on products from Chinese consumers. I think it’s somewhat of a cultural thing: they’ve just been able to produce things at a fraction of the cost and specialize in copying and knockoffs, and that permeates through the consumer spending patterns. Their goal really is to target Western markets.

We went in there obviously very open-minded about trying to partner and invest there, but I think if you question Americans on US-China relations, they would think of it as a bit more adversarial, whereas everyone we spoke to was really welcoming. Everything was focused on how to work together. There wasn’t this us-versus-them component. Maybe it’s just the appropriate strategy because that is what serves them best, but it still felt very genuine, and everyone was really welcoming.

They realized that the American and Western consumer, more broadly, is where you want to be selling. It’s positioning yourself to find a way to leverage the know-how of selling to Western markets with the know-how of producing in Eastern ones.

You also just had a lot of founders who were so impressive on paper. They’d been at the top university in China, which is literally the top university in the world, right? And I think it’s actually the most selective place in the world. I think it’s more selective than the IITs, even, just given the sheer number of people who want to go there.

Then they would have a stint at DJI or ByteDance, managing 100 people on products with hundreds of millions of users. But when they were pitching, it just lacked a certain vision. In the US, you have the opposite problem: a bunch of people pitch really well, but actually there’s no substance. They sell you these big dreams. In China, you definitely have the reverse of that. They were just not, in general, good pitchers. Obviously, there’s a language barrier, but I’d say they weren’t good pitchers and weren’t super visionary about the stuff that was being built.

It was always something that already existed. There wasn’t a lot of—when we went to SF together last time, there was the dude drilling into the earth to terraform, and people trying to build new models and stuff like this. In China, it’s much more like, “We’re doing Lovable, but we’re going to execute faster,” or, “We’re doing lawnmower robots, but they’re way cheaper and better.”

There was a lot of 1-to-X versus 0-to-1, which is maybe just cultural, right? If you have to compete on these really hardcore grades and stuff your whole life, you just don’t have the time to do that, and it isn’t rewarded. Having these—

Yeah, I don’t know. I mean, I guess to do that, you have to be rewarded on valuation, because the ceilings are much smaller, I think. Most of the seed rounds were being done at $30 million or more, and then there was a bunch of pre-launch consumer stuff at $100 million or $200 million. It wasn’t very compelling, I think, on the startup side right now. We didn’t see too much that was compelling.

Yeah. And then as soon as we left, towards the tail end of the trip, some more negative regulatory headlines hit in terms of basically China being less reliant on Western capital in its markets and basically gatekeeping and making it difficult for Americans in particular to invest in these markets at various stages.

I think it’s reasonable to expect that rhetoric to continue in this direction from both sides, the US and China. But I think what that translates to is probably a lower ceiling on some of these Chinese markets, just because they are going to be gated from Western and American capital in particular.

Right now, people are looking for good opportunities, and you’re seeing a lot of activity in late-stage secondary markets for pre-IPO companies. China seems like—or is very clearly—the other big global player, and so I think there was going to be a shift in investing and capital shifting there. I’m sure it’ll still happen, but to a much smaller degree. That has knock-on effects, right? If this money’s not coming in, then some other pocket will follow. I think what that basically does is reduce the ceiling on the opportunities and the multiples that some of these can trade at, for the time being.

Yeah, you can see that in public markets. It trades at like 60% on a P/E multiple, and the size generally is much smaller than the US markets. But it’s tough, because in SF you do have, for the talent that comes out of OpenAI and Anthropic and stuff, people who sometimes raise $4 billion seed rounds, right? You don’t have that in China. Even the really expensive seed rounds aren’t $4 billion, but they are $100 million or $200 million, right? So it’s definitely cheaper at the top level, but I’d say the median valuation is probably fairly similar, which isn’t the case in public markets.

But one thing that stood out, just to also say some positive stuff, because it was a super impressive trip, was how hard they work. We had meetings at all hours of the day, all weekend; the people made no distinction between weekends and weekdays. One founder came to meet us on the day his wife gave birth, which is pretty chad, depending on how you want to look at it, I guess.

Debatably chad. Debatably chad. I mean, he was like, “I’m not a doctor. What am I going to do?”

Which I thought his wife loved. Yeah, and insanely qualified, insanely hard-working, really nice people too—extremely welcoming.

I do think there’s some alpha in finding founders who are sort of more divergent there, and finding managers who are looking for that, because I think there’s a lot of pattern matching going on and people looking for people who have been to ByteDance and stuff like that.

On the software side, it’s just hard to invest in software right now. I don’t know. I used to be really excited about applications and vertical AI. I kind of talked about it here. Now I’m not even sure. I still think there are going to be opportunities there, but I don’t even know if we’re going to have UIs, right? If agents get good enough, you don’t need a UI. You can just ask for what you want and receive it generatively and on the go. So, yeah, I don’t know.

I think those are fair critiques, and I do think that over the next 24 months, the US model providers will continue to pull ahead, just based on the sheer breadth of compute that they’re going to bring to bear. But at the same time, I think the US economy and other Western economies are very exposed to rapid AI disruption. The entire economic model is generally services-based, and obviously China has a big component of that too, but less so, considering it’s more of a manufacturing-led economy. So I think that presents a pretty big question mark.

The other thing is, I feel the US is trying to have its cake and eat it too: “We’re going to reindustrialize, but we still want to have the reserve currency of the world and the deepest capital markets in the world.” Those things are often in tension, right? Do you want to depreciate your currency and invest really heavily in being competitive on exports and reinvest in infrastructure, while simultaneously having a strong dollar, while simultaneously having the rest of the world invest in Treasuries, which is propping up the stock market? At some point, you kind of have to choose which path you want to go down.

In my mind, in the event that the US does pull back from its role as global hegemon, other markets and other equity markets are likely to need to be relatively inflated compared with the US, which has basically been the capital market for the globe for an extended period of time. I’m not saying that won’t happen, but I am saying that in the event that the US does choose to pull back from its role, which began in 1945, I do think there will be ramifications for how people choose to allocate capital.

Will you continue to have 65% or more of the relative market weight invested in US indices 15 years from now? I don’t know. I’m still really bullish on Chinese public markets, honestly, at these levels.

And I think Hong Kong is going to keep doing well, although I don't know that thing well enough. There are some people telling us about shenanigans that go on there that reminded me of crypto, and I don't know how true that is with these IPOs. But I do think the public markets are going to do well.

It's also hard to understand what the Chinese government wants. You kind of need to know that before investing in China, in a way that you don't with the US. In the US, Trump can do and say as much dumb stuff as he wants, but the courts—there's rule of law, right? He probably won't get a third term, the tariffs get ruled unconstitutional, whatever. That just doesn't happen in China.

Today you find out the Mana's founders are detained. And then they did this thing where they don't want offshore vehicles. Yiannis is consulting me in chat.

So it's hard to say what they want. It seems like it would make sense for them to keep the private markets pretty open, let Chinese nationals benefit from those, and definitely not let their talent leave and start companies elsewhere. They also don't want to arbitrage Chinese talent, because there's a lot of people trying to do that, right? They'll have our hardware in China and engineering here, but set up a company in Singapore, Hong Kong, or even the US.

I think that's clearly something they don't want, right? But with the public markets, it seems like it would benefit them to have really liquid, rich public markets where you can dump on foreigners. I don't know.

Yeah, but it's not clear to me that they want that.

Well, it seems like they kind of do, because they're trying to block that late stage, based on that regulation with red chips, right? They're okay with you buying equity in public markets, where Chinese investors can get liquidity and effectively dump on foreigners. It's a protectionist policy; I get it.

Didn't they want to try to encourage people to go public in Shanghai rather than Hong Kong?

No, it's basically this particular structure that they were less excited about you using to list in Hong Kong. As opposed to having the Moonshot structure, they preferred the Zhipu structure, which gives them more domestic control. It's not the VIE entity, the shadow entity that's actually in Cayman. I think it's just the Chinese government's way to ensure that they actually have more control over a lot of the larger strategic AI companies.

But yeah, relative to the US, the upside for capital is capped, right? You'll never have companies that challenge the state in China. Simultaneously, I feel like there is a period where having a relatively—not a crazy stock market, but an inflating stock market—can fill the gap from a massive deflationary rug pull in the real estate market over the last 3 years.

They do need pieces of the economy to start inflating and rotating the store of wealth away from real estate, which got very unhealthy in terms of cost of living and obviously had an impact on demographics. Moving toward an inflating equity market to get out of this deflationary spiral seems fairly palatable from a strategic perspective.

So yeah, I guess I do see the Chinese government tentatively starting to get excited about a long-term structural bull market that doesn't get too crazy, because it is an avenue out of the current deflationary environment, which is obviously a pretty good tailwind for equities.

So are there pockets where things seem super frothy, like MiniMax and Zhipu?

Yes, but at the same time, holistically, I would expect the next 5 years to see pretty— not crazy, but structural—appreciation in the RMB and a relatively healthy, slowly growing equity market, just to help them get out of this deflationary event that they've been in for the last 4 years.

And then China itself—the first city we spent about a weekend in was Shenzhen. It's a metropolis of about 20 million people, but it was a fishing village 45 years ago. Everything is brand new and clean. It's really a sight to behold, especially relative to the decrepit state of a lot of Western blue cities and what's happening here.

You just feel safe there. Everything is orderly. You can actually see drones doing deliveries. It definitely feels futuristic, and it's kind of the headline you heard going into it.

4. Takeaways From Visiting China

Shanghai was really beautiful, too. You see elements where one area looks like Paris, another looks like London, and another looks like New York. It's just a beautiful, modern city.

What was the coolest thing? Let's do something more fun. Not that this hasn't been great, but I saw some things from Anil. What was the coolest thing you guys saw on the robotics side? Or what was your favorite?

I had to leave early, so I missed the dark factory tour. I'm pretty salty about missing that. The coolest thing for me was the humanoids, but I imagine for those guys, that was probably it.

The Xiaomi dark factory was pretty absurd. What size was it? Do you remember, Rossin?

It's 500,000 square feet. I'm a words guy; I'm not a numbers guy.

It was just robots—robot arms and automated transport vehicles. I think they're producing a car every 90 seconds or something insane like that.

It's pretty crazy. I don't think we saw the dark one. There were some humans there helping the machines or fixing the machines, but the scale of it was mind-blowing. It was totally insane.

Also, Xiaomi went from an idea to shipping its first car in 18 months, and to mass production in 3 years. It's something that could only happen in China. It's absurd, right? Apple spent 10 years and $10 billion or something and couldn't get a single car shipped.

I do think you guys had some solid critiques that I agreed with, but the one thing that shouldn't be underestimated is the Andy Grove line of reasoning. Not just Only the Paranoid Survive, but if you lose proximity to the forces of production and the production line, then you also start to lose the ability to innovate on those things, right?

I think Xiaomi is a really good example. It's close to the production of a ton of different types of electronics, so it can spin up a beautiful car within 1.5 years, whereas Apple outsourced all of those things over the course of decades and is now much further from a lot of the processes needed to build a beautiful car.

Dan Wang talks about this a lot, but there is a high cost of innovation to not being close to the actual production and just doing design and financialization. Eventually, that does come back to bite your ability to innovate, especially in deep tech, manufacturing, and a lot of these other domains.

So maybe you have fewer single-founder visionaries, but you do have an ecosystem that is continuously and rapidly evolving. If you look at things like DJI and its tech stack, what's going to come in robotics, and what we've seen in EVs, I would expect that stuff isn't going to slow down.

I think one funny point was that basically everyone we spoke to was very smart, humble, measured, and not hyperbolic at all. But the one thing everyone was also highly convicted in was that the global market share will be 70% China, 30% US, and 0% Europe. Everyone is convinced Europe is cooked.

It was pretty funny. It was the number-one bonding experience between the Americans and the Chinese. You'd have this stately elderly Chinese man saying, "Obviously, in humanoids, China will have 70% share, the Americans will have 30% share, and the Europeans will have nothing." Yan and I would look at each other like, "Oh yeah. Yes, of course."

We'd give Jose a quick sideways glance. But I honestly do think it will be very difficult for places like Europe in terms of competition. I think the US has its own growth engines, but Europe is more directly in the line of fire, because China is going after a lot of its industries and doing it better.

They're doing it better than a lot of US companies, too. We just have protectionist policies and a strong consumer. We have a large enough market that can insulate itself in certain areas, so they don't export some of these products.

Like the Chinese electric vehicles that we took an Uber in were all really nice. Didn’t you guys feel, while you were there, that the quality of life relative to the cost was abnormally high? To me, at least, China just feels like the exchange rate is too low.

Yeah, there are just some trade-offs, though. I would jaywalk, and I would look at you and watch you patiently waiting for your green light.

I live here, man.

And you’re driving, and you’re getting photographed every 100.

Every car that drives by is photographing you. You’re just getting photographed.

Really? You can see the flashes?

Mhm.

Oh, wow. So, yeah, I’m not saying there aren’t trade-offs. I basically was saying that, relative to the material quality, it seems like the RMB is undervalued. They should let it float, and it would allow the rest of the globe to be more competitive on exports. It would also solve a lot of the trade imbalances.

This is basically what the U.S. wants, right? Can you let the RMB appreciate slightly so we have less of a trade imbalance? We don’t want decoupling. If you look at Scott Bessent’s latest speeches, we just want adjustment because this is unsustainable. A lot of people would say China has an unsustainable trading relationship with the rest of the globe because they export so much, and a lot of it has to do with the fact that the exchange rate is relatively low. I think a lot of those issues would be assuaged pretty materially if you just saw the RMB appreciate by 20% overnight. That has other implications.

They’ve been saying that for years, though. I feel like I’ve been hearing this for years—that they should do this, that they’re going to do this. Why haven’t they done it yet?

I’m not sure I want to say this on a podcast. I think it’s basically that there are strategic industries that China wants to be very competitive in. Also, having a ton of exports allows them, as the state, to direct resources as they see fit, strategically compounding in areas that are nationally strategic to them.

Whereas if you appreciate the RMB, it’s effectively giving more of the wages to consumers and workers, and they would then be able to direct that capital individually, as opposed to having more centralized control over how it’s spent, basically.

Agreed. Guys, I have to jump to another call. I’m already 10 minutes late. See you, Jose.

See you.

Yeah, this is good. We’re up on an hour now. We’ll probably wrap it anyway. I thought that was a good talk. I don’t know if anybody has anything else they want to bring up.

No, that was great. I loved listening to the China takeaways. It was great. Thank you, guys.

Yeah, it was a cool trip. You guys got to organize it and explore it yourselves. What was the food like?

How was the food, Yan?

Yeah. I’m a bit of a garbage disposal when it comes to food. I’ll eat everything and anything, so I liked it. What ended up happening is there’s basically Chinese food and Cantonese food. Ross can definitely provide better detail on this.

Cantonese food is a bit more exotic. It’s a bit more squishy. “Squishy” is probably one of the best words to describe it. It’s squiddy and kind of just weird fish and enterprises that you can’t really figure out. But it was good.

The M.O. is that if it’s fancy, it usually skews Cantonese. Every dinner that we went to, and even some lunches, ended up being fancy, so it was 80% of our meals. The Chinese food, though, that we had a couple of times was really good. I liked all of it. I think we ended up having 90% Cantonese, and I think that ratio was the issue.

Squishy. Good word, huh?

Yeah, sorry, my video cut out, but I would say we spent a good amount of time in Shenzhen, which is obviously in Guangdong, close to Hong Kong. There’s a lot of Cantonese cuisine, which at the high end is pretty fancy and a little more exotic. The guys did well being adventurous in some of their meals.

I’m actually partial. I feel like Chinese cuisine is one of those where the everyday street stalls tend to be better and easier for my palate than the super high-end stuff, which starts to get a little more creative. When in doubt, I feel like the blue-collar street-stall options tend to be the stuff I really like about Chinese cuisine.

Chinese cuisine is super diverse. If you’re in Yunnan, you’ll get something very different from what you get in Xinjiang, Shenzhen, or Shanghai. It’s pretty diverse, but I thought the guys did well with some more creative dishes.

Honestly, the most difficult thing to consume—and I’m a pretty big drinker—was this booze that they have there. Holy hell. What is it called again?

Baijiu.

I thought you were a huge baijiu fan by the end of the trip, Yan.

It’s probably one of the harder alcohols I’ve ever tried to consume.

Yeah, it’s definitely an acquired taste. It’s all the rage, especially with the older cohort in China. If you go to lunch with anyone over 45, I think there’ll be a good amount of baijiu consumed. It’s a pretty big part of the culture.

In general, if you come to China, I’ve been very impressed by the hospitality living here. I feel like people in general, when they’re hosting, are exceptionally gracious and really, really welcoming. Hopefully, Yan and Jose and the guys felt that on this trip. Every time someone takes you out for a meal or invites you over, they typically go above and beyond. I feel like they’re actually a very generous people, despite what’s often portrayed in the media.

100%. Yeah, I couldn’t agree more. One element I really enjoyed was that every large dinner was at a round table. You basically had the food sitting in the middle, and the plate would rotate maybe once every 2 minutes, so you gradually picked off food and put it on your plate.

The other element was that even if you had a group of 6 or 15, it still managed to be one conversation for the most part that everyone was part of. In most other places, you have these rectangular tables, and it siphons off into 3 or 4 conversations and just becomes chaos.

Here, it was a really good setup for everyone to meet each other and all speak about one thing. I thought that was really productive because you didn’t have this issue where it was, “I don’t really get to meet this person or hear about them.” The whole point was to meet people and understand how they were thinking about Chinese markets, Western markets, and everything in between. This was a really fun way to do that.

Bullish on round tables. Well, it sounds like it was fun. I’ll have to go someday, Ross. I’ll have to hit you up.

Yeah, come on through, sir.

We’ll wrap it here for this week. Thanks, everyone, for coming on. Thanks, Ross, for making a guest appearance. I think that gave us some cool insights into the Chinese tech and AI market. We’ll see everybody else in a couple of weeks.