The World Now Studies China the Way China Once Studied America

“A lot of Westerners are still working from an assumption: if you’re doing better than us, you must have done something wrong — you either stole it or you took it.” — Robert, on what foreign investors believe before they arrive in China

Eight or nine billionaire-tier investors sit down at one table and order everything on the menu at Lao Xiang Ji, a Chinese fast-food chicken chain. Some dishes twice. The bill comes to 400 yuan — about $56. Robert asks the table to guess. The lowest guess is seven or eight hundred.

Robert Wu is the CEO of Baiguan, which helps global buy-side funds make investment decisions with better data on China. Its English-language newsletter has over 20,000 subscribers on Substack, reaching investors, corporate executives and senior government officials across many countries. He was the guest on episode four of this show, where we discussed the op-ed he wrote for the New York Times on the US-China trade war. Lately he has taken on a new role: tour guide. He brings the world’s top investors, founders and large-company executives to China to see MiniMax, Pony.ai, Black Lake and Fourier Intelligence — and to sleep at Atour and eat at Lao Xiang Ji.

This episode is a run of expectation gaps. What surprises American investors is not how advanced Chinese autonomous driving is, but that ordinary Chinese people can barely get into a driverless car. The questions European investors ask reveal an underlying premise: without us, you couldn’t survive. The robotics industry is somewhat overhyped — five stacked “maybes” add up to roughly zero. And the best indicator of Chinese consumer confidence may be buried in KFC’s quarterly numbers. The larger judgement fits in one sentence: the world is switching its frame for China from “big market” to “something to learn from.”

What follows is the full conversation, edited and condensed.

1. They pay thousands to come to China — and skip the Great Wall and the Forbidden City

Raymond: Since China relaxed its visa policy in 2023, #ChinaTravel has blown up on TikTok — foreigners visiting China and posting videos, enormous volumes of it both on Western platforms and on Xiaohongshu and Douyin. I’m curious how foreigners actually see China, especially investors and founders. What assumptions do they bring before they join one of your tours?

Robert: Our group is very different from the tourists on YouTube. Investors and corporate executives who are willing to pay this price have a very strong desire to understand something. Roughly two types. Most already know China relatively well, come here regularly, and just want more surface area of contact. But there genuinely are people who have never been to China before — and this is the striking part. Someone who has never set foot in China, who doesn’t climb the Great Wall or visit the Forbidden City, and goes straight to looking at Chinese startups.

Raymond: Is it a big culture shock for them?

Robert: Quite big. But the people who choose us tend not to arrive with strong preconceptions. They come with a learning mindset, fairly open-minded. People carrying strong prejudice scroll past and never sign up — they won’t spend the money. One participant offered a phrase that defines our service well: educational tourism.

Raymond: I’d heard of medical tourism. Now there’s educational tourism. Which countries do these investors mostly come from?

Robert: Fairly evenly distributed globally. Southeast Asia is a large source — on the last two trips, about a third came from Southeast Asia: Singapore, Indonesia, Thailand. Europe is a constant, Germany especially; both trips had a good number. The US is of course an important source, and we also get Mexico and other Latin American countries. It mirrors the distribution of our readership: the US is 30%, but it’s only 30%.

Raymond: Do people from different regions look at China differently?

Robert: Yes. The Southeast Asians are more familiar and will slot what they see into an existing framework as an increment — they’ve looked at another robotics company before, now they see this one, and they can run a comparison during the conversation. Singaporean investors have been here many times; they’ve seen Li Auto and NIO, they’re done with EVs, now they look at autonomous driving compliance and testing, and they start asking about very specific technical choices. The Europeans ask more basic questions. Our challenge is to serve the repeat customers and the beginners well in the same room in a very short window — teaching each according to their level.

Raymond: A bit like putting kindergarteners and university students in the same classroom.

2. Nobody says “the Chinese OpenAI” anymore

Raymond: I’ll put your Substack link in the show notes. A lot of your angles come out of collisions with overseas readers and visiting investors, and reading it I often find things I hadn’t considered. One of them: you say the defining trait of this generation of Chinese founders is confidence. Where does that confidence come from?

Robert: That line came from someone at a large-model company we visited. “This generation” mainly means those born in the mid-1980s and 1990s, as against the generation of Jack Ma and Pony Ma. I think the biggest difference is that this generation grew up broadly in sync with the world, without a huge information gap: the books they read as children, their understanding of how the internet developed, were synchronized with the most sophisticated people anywhere. Their parents’ generation was still led by a narrative of “I want to become the Chinese Siemens,” “the Chinese Google.” Now if you say “I want to become the Chinese OpenAI” or “the Chinese Palantir,” you wouldn’t even feel confident saying it. More and more people define themselves on their own terms, reasoning more from first principles.

Raymond: The early Chinese internet companies — that whole cohort from 2010 to 2020 — were basically Copy to China. I worked on the Renren IPO: China’s Facebook. Tudou, Youku, iQiyi: China’s Netflix.

Robert: Right, that generation was like that. Maybe TikTok is what made people feel the pattern had changed, and later Manus as well. One topic the group discussed a lot on this trip: the founding stories of China’s leading AI companies no longer follow the old template. MiniMax founder Yan Junjie started MiniMax before GPT came out; he committed to the Transformer direction very early, rather than going after it because GPT got hot. DeepSeek is an even clearer case. A lot of them think about problems from first principles instead of insisting on becoming “the Chinese something.”

Raymond: There’s another detail in your report that surprised me: many of this cohort of core founders were educated domestically, not abroad — and yet in that same setting, nearly all your meetings were conducted in English.

Robert: Let me separate the two. We visited quite a few startups, including ones inside Shanghai’s large-model incubator MoSpace — there are plenty of returnee-founded companies there, but the teams are not all returnees, and the level of localization is high. Large-model companies have a particular characteristic: the founders of essentially all of China’s leading large-model companies have no overseas-education background. Liang Wenfeng, Yan Junjie. Yang Zhilin’s first degree was from Tsinghua, trained domestically. Zhipu is Tsinghua all the way through. Nor is this like the previous generation of AI founders, who routinely spent eight or ten years at Microsoft. Of course, once a company reaches this scale, it’s perfectly normal for executives to be able to converse in English.

At one large-model company, a European participant asked a question the company initially didn’t understand, and I “translated” it twice — in quotation marks, because what I was translating was the intent behind it. First question: what percentage of your people studied overseas? Answer: one third returnees, two thirds domestic. Next question: if one day the US no longer allowed Chinese students to study there, would that have a major impact on you?

The company representative froze. I’ve talked with German friends, so I knew his premise: he assumed by default that these companies are heavily dependent on overseas education, overseas resources, overseas everything. The company’s answer was: first, two thirds of us aren’t returnees; second, let me correct something — a diverse talent pipeline doesn’t mean we can’t survive without returnees.

That premise objectively exists. Many Westerners start from “without us” — us being the entire greater West, from ASML to the education system — “you couldn’t do much of anything.” If you’re doing better than us, you must have done something wrong, whether you stole it or took it. A lot of people still haven’t fully accepted a reality: China can do it better.

Raymond: The best rebuttal is DeepSeek.

Robert: Right, DeepSeek is unambiguously all-domestic, and when R1 came out it clearly contained some very new ideas, a major innovation for the field — and open-sourced on top of that.

3. Robots learning to be human: a guy sitting there teleoperating a robot to close a wardrobe door, all day

Raymond: You took them to MiniMax, Pony.ai, MoSpace and others. Was there a single scene that hit them hardest?

Robert: Each had its own character. The one that stayed with me was Black Lake. Black Lake has no clear overseas comparable — work-order management and a full internal system for small and mid-sized manufacturing plants. The visitors were extremely interested; one former Hong Kong Stock Exchange director ended up strenuously urging Black Lake to go public. This kind of company is closely fitted to China’s realities: China’s manufacturing base is enormous, so the same business that might be a very niche one in the US can reach scale here. And once you hold large volumes of data, you can evolve toward agile manufacturing, which opens up a lot of room. This is not a business model you can easily see elsewhere.

Raymond: Let me add some context on Black Lake. The founder’s family has a factory background, and what they first built was making factory order systems more flexible and more mobile — a business that sounds grim but is actually very cool. Now they’ve added a lot of AI, using models to handle dispatch and lift overall plant efficiency. I need to get Black Lake’s founder on the podcast one of these days; he lives very close to me. What about MiniMax?

Robert: The questions concentrated on technical choices. MiniMax currently holds some non-consensus views: many large-model companies are moving toward coding, while it still puts heavy emphasis on multimodality. It’s hard to say that’s definitely wrong — the wind in AI changes direction every six months.

Raymond: What I’d imagined was that your group would see a giant Chinese factory.

Robert: The window was too short this time; we only started promoting it a month ahead and honestly thought it might not come together. In the end 19 people came, one short of our 20-person cap, so we only arranged things in Shanghai. But we very much do want to run a manufacturing trip.

Raymond: Can I tag along for free? I’ll carry the bags, I’ll shoot extra footage for you. Some of what’s happening in China’s auto industry is genuinely absurd — the German chancellor came to China recently and visited Unitree, and I imagine the Germans concluded that this life is a write-off.

Robert: Speaking of which — we went to Fourier Intelligence, one of Shanghai’s leading robotics companies. What made it special is that we went to its data collection center. Normally when you visit a robotics company you watch a performance, but many people don’t know that the main job of a lot of robots today is data collection.

Raymond: Walk me through that, I don’t really follow.

Robert: The brains of today’s robots are underdeveloped. Physically they’re already very capable, but they lack strong autonomy, and the biggest bottleneck for that independent judgement is insufficient data. Autonomous driving has millions of vehicles on the road collecting data to train the models. Robots? There might be ten or twenty thousand robots in commercial operation in China — and that number is already high — and most of the time they’re dormant, so the data volume is actually small. When a company buys a robot, it’s buying the body; it has to be trained in the actual setting to get better. And robotics companies themselves need to collect large amounts of data. How to collect real-world and synthetic data turns out to be the most core thing at many robotics companies.

What we saw on site: in a simulated home environment, some robots were folding clothes, some washing dishes, some closing wardrobe doors. A guy sat there teleoperating a robot to close a door, then close it again, and again, practicing over and over, recording large volumes of data throughout — the force used on the first close differs from the second, the joints load differently, how to get the door exactly shut without damaging it. A dozen or so operators in that workshop, each drilling one robot continuously. The scene was fairly science-fictional. We’d seen how robots move before, but we’d never seen how they get trained at the source.

Raymond: I already have the title for that video: watch robots learn to be human. In a sense this is also an important future category of employment. Folding clothes, cooking, closing doors, looking after children in a home environment is a finite task set, but each task has an infinite number of variations, and the better you do on each one, the greater the eventual commercial possibility.

4. What surprised American investors most: Chinese people can’t actually get into a driverless car

Raymond: How do you look at autonomous driving?

Robert: It was a major research focus this time. We invited an early investor in Pony.ai to give an industry briefing, and we took everyone to Shanghai’s pilot zone to experience autonomous driving. There’s a large expectation gap here. For a fair number of the American investors, Waymo is already very common in many US cities with high penetration; it’s part of daily life. But in China, penetration of L4 driverless vehicles is very low. The zone closest to central Shanghai is the area around Century Park — not far, not near either, but definitely not the busiest part of the city. The share of Chinese consumers using pure L4 driverless transport in daily life is, I think, extremely low. That surprised some American investors: they assumed that with Chinese technology this advanced, with robots supposedly everywhere, ordinary people should be riding autonomous vehicles daily — because the US is already like that, and plenty of people in San Francisco would rather hail a Waymo than an Uber.

So we analyzed why. This is clearly not a technology problem — Chinese L4 companies aren’t far off Waymo technically, and the experience is very good. Our read is that China faces a dilemma. On a ten- to twenty-year horizon, with an ageing population, there certainly won’t be enough drivers; autonomous driving, robots, substituting silicon for carbon — all of it has to happen. But on a three- to five-year horizon, employment pressure is significant, and the gig economy of ride-hailing driving functions as a social safety net occupation. So you can see the Chinese government in a posture of long-term support and deliberate short-term management — artificially keeping the technology from developing too fast. It knows this has to advance, but pushing too hard in the short term would create social problems. The group found this fascinating: it combines technology, policy and economics in a comprehensive calculation, which makes the industry look, on the surface, as though it isn’t developing as fast as in the US.

Raymond: This is different from what I’d assumed too. I used to think the reason I don’t see Waymo-type vehicles around me is that China hasn’t built one yet. The way you describe it, that zone was drawn where it was drawn. Apollo Go in Wuhan is also in a fairly peripheral district.

Robert: The zones aren’t actually small, but they’re all a bit out of the way. How they’re drawn is a policy question, not a technology question.

Raymond: Draw it in the outer districts and it doesn’t affect many people.

5. Robotics is somewhat overhyped: five stacked “maybes” add up to roughly zero

Raymond: Robotics and autonomous driving are the same category, and investors are certainly all watching China on both. Your notes mention that in robotics the body is currently running ahead of the brain, and data is the obvious bottleneck. Beyond data, what else worries investors about Chinese robotics?

Robert: A lot of investors are very focused on where the whole-body components come from, and they ask in great detail: the actuators — the joint part — who supplies those? Are the chips Nvidia’s or someone else’s? First, they want to understand how self-sufficient this Chinese industry is; second, they’re hunting for upstream opportunities. In AI broadly right now, what capital markets care most about is hardware and source components, and hardware inherently has the question of who has a hand on whose throat — memory is currently choking AI. Embodied intelligence hasn’t truly broken out yet, but many people are looking ahead: if it does break out, which links in the chain will bottleneck? All of it observed from a perfectly ordinary investment standpoint.

Raymond: What did you personally find interesting? How self-sufficient is Chinese robotics really?

Robert: Honestly, I think robotics is somewhat overhyped right now. It’s more that people feel this will be a very big thing in the future, so a lot of force is being concentrated into fundraising to do it — but what it actually turns into, who can genuinely train the brain, when the so-called ChatGPT moment for robotics arrives, nobody can say precisely today. That question is guaranteed to come up in the group. And a lot of people ask Chinese robotics companies: do you think Tesla can pull it off, and when?

Raymond: What do they say?

Robert: Every kind of answer you can imagine. Everyone respects Tesla, but they also think Musk is frequently overselling and drawing pictures — Optimus is always about to roll off the line, and the date keeps moving. My view is that attention is high enough right now to generate commercial value on its own — even building performance robots is a form of demand. But actually entering millions of households and becoming a vast industry that changes how people live is still a long way off. So a lot of robotics companies are expanding fast and raising fast while the window of society-wide attention on AI is open. It’s a very early stage.

Raymond: I saw an investor post that total funding for Chinese robotics has already reached an absurd number. How does that math work? It could be even more brutal than EVs were. What they have to bet on is: very large shipment volumes, which create a large entry point for data collection, and better data may produce a better model — stack five “maybes” and you get roughly zero.

6. A problem raised in the morning is solved by lunchtime — and it made the Europeans envious

Raymond: You mention that local government here plays the role of a supportive platform rather than a central planner. What do incubators like MoSpace actually do?

Robert: The usual understanding of an incubator is space plus startup mentoring. But the tenants of these government-run incubators are mostly AI supply-chain companies. First, the support on tokens is fairly generous — effectively subsidizing them through token credits. Second, administrative efficiency is very high: approvals, taxes and so on are handled centrally, and there are staff who proactively ask, “do you have any problems? Bring them all to me and we’ll resolve them in one place.” On this point several European participants were fairly shocked — European regulation is very complex —

Raymond: Not complex. Malignant.

Robert: Right. So when they heard “a problem raised in the morning is solved by lunchtime,” that very strong service-oriented government posture both shocked them and made them envious. Of course Shanghai is a very advanced sample — an efficient government-business relationship with clear boundaries, a high-water mark many places are still trying to learn from. That sample is certainly biased.

Raymond: Here I might disagree slightly: Shanghai isn’t necessarily the most advanced government-business relationship in China. What you’re describing leans toward the Hong Kong model — efficient, flexible, a highly connected service-oriented civil service. But there are many other places in China, like Hefei, and the Hefei model is extraordinary. What the Europeans keep saying they dislike and object to may be precisely the Hefei model — your Hefei-style government has beaten the fear into us.

Robert: Then next time we should take them to Hefei and let them see how much a government can do.

Raymond: Hefei genuinely just took off over the past ten years; everyone knows how extreme its support for high-tech companies has been. Large models are a comparatively soft business, and Shanghai’s kind of support is already enough to make Europeans say, “I’m so envious, I’m moving here tomorrow.”

7. Sleeping at Atour, eating at Lao Xiang Ji: nobody guessed the meal was only 400 yuan

Raymond: Let’s talk consumption. Unitree has become a highly representative Chinese company; at least twenty or thirty heads of state and senior officials from various countries have wanted to visit. But Chinese consumption draws just as much attention: has consumption recovered? Are we still in a deflationary environment? Those questions say more about the underlying tone of the Chinese economy. You ran a consumption trip last year. What did you look at?

Robert: We went to some mid-sized consumer companies in Shanghai and Hangzhou: Atour Hotels, JNBY, M&G Stationery and others — companies with market caps in the tens of billions of yuan that are extremely dominant in their niches. M&G is close to the only Chinese stationery company you can name; it claims to have more stores than China has schools, because it makes sure there’s one outside every school, plus all the ones not outside schools. JNBY is headquartered at Tianmuli in Hangzhou, and it has spent thirty-odd years in an industry where many people can’t survive: fashion. I wouldn’t buy it myself, but it has somehow built a very deep relationship with its own users: sixty to seventy percent of its customers are in private channels, membership-style customers maintained by store managers in WeChat groups, who buy every seasonal drop. Many of the companies we showed had this quality: they became a small giant in one narrow field, and they don’t care about anyone outside their own customers. Atour is arguably the same; its brand recognition exceeds what you’d expect of an ordinary consumer company.

The intent was obvious: China’s consumer market genuinely hasn’t been hot these past few years, and there’s even been a chill. But even so, you can find excellent companies within niches. JNBY’s share price has multiplied several times over in recent years with a seven or eight percent annual dividend — very high quality. None of this is investment advice; do your own research.

Raymond: What did foreign investors expect Chinese consumption to look like before they came? Once they landed, did it exceed or fall short?

Robert: Definitely exceeded. Most people don’t arrive with high expectations; they’re broadly pessimistic about Chinese consumer confidence over the past few years. But after taking them through various micro-worlds within niches, their understanding shifts considerably. We even had them stay a night at Atour — many of them normally stay in five-star hotels — and some thought it was no worse than a starred hotel, at a dramatically lower price.

Here’s another example. We don’t only do the high-end; we also take them to experience the everyday. One evening we hadn’t scheduled a dinner, so I said it was free time, but if anyone was interested I’d take them for Chinese fast food — Lao Xiang Ji. Most of them came along, and we just picked a location in Hangzhou. You know, many of these are billionaire-tier people; eight or nine of them sat at one table and ordered every dish on the menu, some of them twice, plus drinks, for a total of 400 yuan. I asked everyone to guess the bill, and nobody landed near that price — the lowest guess was seven or eight hundred. At Lao Xiang Ji, forty yuan per person is already high, twice the average ticket. Everyone enjoyed themselves enormously.

Read enough foreign media and you’ll come away thinking Chinese consumers are miserable and despairing. But the actual feeling on the ground is: people are still consuming continuously, still obtaining high-quality goods and services. It’s just that wallets are tighter and spending is more careful, and merchants have started competing ferociously, so service quality is in fact improving. Expectations of falling prices objectively exist, and they do have adverse effects, but it isn’t that Chinese people have started living on air — the overall quality of what Chinese people consume has been rising these past few years. Contradictory, but objectively true. In a sense, this is building up potential energy for future consumption growth.

Raymond: I traveled abroad a couple of months ago and a friend there asked me one question: has sentiment come back at all? What I wanted to say was, it doesn’t actually seem that bad. The worst of the mood was probably 2022 and 2023 — the pandemic, plus real estate, with the Chinese economy under pressure. But over the past few years, the property that had to fall has finished falling, and people have gradually come back.

Robert: This is another thing that hasn’t fully landed overseas: the mentality of Chinese households has already normalized. What happened, happened, and when people make decisions they mostly weigh the current marginal change — and at the margin things have stabilized. Whether it’s official CPI or the micro data we see ourselves, it’s steadying and in places warming. An example we love using is Yum China’s results — they publish operating data every quarter, and we particularly like watching KFC’s average ticket. From the start of 2023 through the third quarter of last year, more than two years, the average ticket kept falling: revenue growing, volume growing, but unit price declining. There are many reasons — a rising delivery share and so on — but fundamentally, things got cheaper. Why watch KFC? It’s China’s largest restaurant chain group, so its cost control is certainly the best in the market. If it keeps cutting prices, nobody else dares raise them; conversely, if the whole market is going to warm up, it warms up first. We saw the average ticket begin to stabilize, and even recover somewhat, from the third and fourth quarters of last year into the first quarter of this year.

Raymond: That measure captures a great deal of what goes into CPI — labor, rent, food, transport are all in it. We should have you back sometime to do a dedicated analysis of KFC’s financials.

Robert: But many people in overseas markets, and in the domestic market too, don’t have this expectation yet. Across the whole consumer sector today, from both a capital-markets and a business standpoint, people are still operating on the expectation that “we haven’t come out of the old gloom.” I think there’s an expectation gap sitting in there.

Raymond: My homegrown method for reading Chinese consumer confidence: is there traffic when I go out, is there a queue at restaurants, are the planes full. Lately Shanghai has both traffic and queues, and the economy honestly looks pretty good. Two weeks ago I traveled in Henan; May is low season and it was still fairly lively.

8. Brazilians are buying counterfeit Labubu

Raymond: One more remarkable presence in consumption: Pop Mart. When I was traveling in Brazil, a girl of thirteen or fourteen explained the Labubu family tree to me — who is whose what — and I was stunned. Has the soft power really carried that far? Did your itinerary include Pop Mart?

Robert: No, but we included an anime-merchandise store — I really am getting old — we walked around a large one in Hangzhou to experience that culture. This is genuinely an area of intense overseas interest. On Brazil: I have a Brazilian reader, a macro fund manager, who came to China two months ago, and the main task his daughter gave him was to buy one particular series — a relatively obscure one, not Labubu. I was surprised at the time. The penetration is that strong.

Raymond: Let’s go into the Labubu reselling business in Brazil together, we’d crush it. Dear listeners, if you know Wang Ning, forward this episode to him: Brazil really loves Labubu. There are no official Pop Mart stores in Brazil, so electronics shops in Chinatown put out fake Labubu to draw customers in — if you genuinely don’t show up, the counterfeits become popular. Brazil is obviously a large market.

Robert: Designer-toy culture is still fairly rare in Europe and the US; Japan and Korea have more of it.

9. Where to take them next: Linfen, Lu Xun Park, and an ordinary Chinese factory

Raymond: If you were organizing the trip fresh today, beyond MiniMax and Atour, what would be unmissable? Something over ten percent of my listeners are outside China — I don’t know why either — so if a foreign investor wanted to come look at China, what would you recommend?

Robert: We are in fact thinking about this; the demand is large. Three things I particularly want to do right now. First, a median-city trip. Not the most famous cities — go to a place you’ve never heard of, sitting at the midpoint of all Chinese cities, representing the broadest mass of people. We discussed this on episode four; that city is Linfen. You get a real feel for the environment most Chinese people live in, and you judge for yourself whether they’re living well, what their mentality is, what the state of their lives is. Getting to Shanghai and Beijing is relatively easy. That middle place is one very few people go to.

Second, something to do with the silver economy. There are a great many founders and investors focused on China’s silver economy, and the newest cohort of retirees — people born in the 1960s and 1970s who have just stopped working — may be the wealthiest generation in China, holding the most assets and resources. On the last trip we nearly took the group to Lu Xun Park in Shanghai; I went and scouted it specifically for this. On weekends it’s extraordinary — dozens of groups across the whole park, choirs, dancing, instruments, all retired middle-aged and older people. You very rarely see this overseas.

Raymond: Lu Xun Park is Coachella for Shanghai’s elderly.

Robert: It’s a microcosm: the lives of older Chinese people are extremely rich, and they still have a lot of spiritual needs. Foreigners would love it. Third is manufacturing, a genuine foundation of the country. We want to take them to see more factories, even just ordinary factories — what we want to show is not the most cutting-edge thing but the more typical thing. The level of automation and management in an ordinary Chinese factory far exceeds what people overseas imagine. Ask any American about “a Chinese factory” and what’s in their head is still a sweatshop. That era is completely gone, but many people haven’t updated the picture in their minds.

Raymond: Here’s some first-hand feedback for you. Ever since you mentioned Linfen on episode four, whenever I travel or work in any Chinese city, I look up the local GDP — above Linfen is the better half, below Linfen is the lower half. People in tier-one cities may think Jing’an Temple is the whole of China. It isn’t. The overwhelming majority of China is not inside Jing’an Temple. So I use Linfen as the benchmark every time.

Robert: Have you since been anywhere below Linfen?

Raymond: No, but I check. For instance I went to a place near Zhengzhou and looked up roughly where it sits. Honestly it’s fairly hard to get to those places now — the cities you’d visit as a tourist never have low GDP, and places with no tourism draw, like Tianshui, you wouldn’t go out of your way for. I’ve done something similar in manufacturing: at my previous firm I took overseas investors to see China’s garment manufacturing, to Shisanhang in Guangzhou, to see how clothes are wholesaled and retailed. Once you walk into that building — never mind foreigners, Chinese people walk in there and are shocked.

Robert: Speaking of which, on our first trip last year there was one scene that made everyone go wow: a livestreaming company in Hangzhou, an entire eight-story mall, lots of storefronts, but no customers at all — each store is a livestream studio, with staff inside broadcasting, constructing the impression of selling goods in a shop. An empty mall with every storefront livestreaming. I was wowed myself.

Raymond: That kind of mall would be near Sijiqing; it used to be for wholesale buying. Sample garments hanging at the door: “this piece comes in S, M, L and five colors, pack up 150 units and ship them to France.” Nobody sells that way anymore, the mall lost its customer flow, and a normal young woman wouldn’t buy clothes there, so it turned into a livestreaming base. This is something even Chinese people find shocking.

10. The world is starting to look at China the way we once looked at America

Raymond: Let’s wrap up. Baiguan has a great many overseas investor readers, and paying attention to China is simply their job — everyone knows China plays an extremely important role at every level of the world. Using Baiguan’s readership as a thermometer, is the overall view more optimistic or more pessimistic? Two years ago many people thought China was a very difficult country to invest in. How has that judgement changed?

Robert: These past two years are when the overall paradigm and understanding have been switching. Earlier, the understanding of China was of a big market with a very large number of consumers — investors and multinationals both saw it that way. Over the last two years there’s been a clear trend: people increasingly treat China as something to learn from, which is also why there’s now demand for inbound study trips. That used to be how we looked at America: Chinese people organizing delegations to Silicon Valley, to Wall Street, and in recent years to Japan, to study business models and advanced companies. This has only just begun, but the world is starting to look at China much the way we ourselves once looked at America — as a place to learn from and draw inspiration from. Two frames are switching over.

On investing, China isn’t all that easy to invest in right now: consumption is still fairly depressed from a capital-markets standpoint, with no clarity on when a real recovery comes; and on tech, Chinese AI is genuinely hot, but large investable vehicles like the Hang Seng Tech index remain difficult. Even so, people are still paying close attention to China, because that attention has become comprehensive — it’s no longer just treating China as a market where “I can make money,” it’s now “I can upgrade my understanding, I can learn more.” That’s an enormous difference I’ve felt personally. As for optimism versus pessimism: when we started in 2023 and 2024, you could count on one hand the teams sharing information about China in English. Starting last year, all kinds of companies, teams and niche businesses have emerged. Behind that there must be more people paying attention. In 2025 expectations shifted very visibly.

Raymond: My personal feeling is that when someone holds a prejudice about China or fails to understand it, I often can’t be bothered to explain. I tend to think your stupidity isn’t my problem, and I won’t be that patient. But seeing your Substack, and the op-ed you wrote for the New York Times, moved me: we badly need people like you, who patiently take all the granular details and put them into language the other side can understand — and by “understand” I don’t mean English, I mean standing in the other party’s perspective, their starting point, their interests, so that there is a little more understanding in the world and a few fewer misunderstandings. Honestly, Chinese people as a whole are too reserved: do ten, say one. Some countries do one and say ten. We should moderate a little — when a Chinese person finally says two, that second one is your contribution.

Robert: In the actual doing of it, you find the demand really is large. At minimum it gives the people around the world who are genuinely interested and genuinely willing to understand a way to get information — that’s what we think our existence is for and worth. There are of course the idiots you mention, but there are also the non-idiots who sincerely want to learn and understand. What are they supposed to do?

Raymond: If nobody speaks, they’ll have to pay to join your tour.

Robert: Welcome, welcome.

Recommended by our guest: Baiguan’s English-language newsletter, “Baiguan,” on Substack.

If you're working on this too — or you think we've got it wrong — write to us at [email protected]; if you'd rather not write, just leave your email below.