I will personally be with these tours alongside Caixin, , , and many other illustrious partners of Baiguan.

But before I share more details with you, let me share a real story that shocked me recently.

Do you know Breakneck?

So a while ago, I hosted a sharing session about Baiguan’s experience with 40-plus Chinese media professionals working in international communication.

Here, “international communication” simply means they produce foreign-language content for a living.

In the sharing session, I mentioned the book Breakneck by Dan Wang, which went viral last year in overseas China-watching circles. Even many people who are not normally interested in China have read that book.

I wanted to use this book to make a point: when international audiences look at China, it’s not really China they are looking at; it is themselves.

Just like Breakneck, which is supposed to be the most popular book about China in recent years, the book is actually less about China and more about how the U.S. compares with China, and it’s the anxiety that Americans feel about deindustrialization that ultimately fuels the popularity of the book.

I covered this topic quite extensively here:

I brought up that book midway through my sharing and posed a quiz to those in attendance.

The quiz is simple: “How many of you have heard about this book?”

I was expecting maybe only a few of them to have heard of the book, so I would use that as a second-order proof of my point: we also know far less about the other side than we should. A lot of what we Chinese know about the world is based on what we want to know, and we are always projecting ourselves onto the world as well.

Only that I probably made my point too well.

It turned out, to my great shock, that of more than 40 Chinese media professionals whose main job is to produce English-language content for foreign audiences, and who are supposed to follow what the foreign readers are reading, NONE of them, absolutely ZERO of them, have heard about Breakneck.

Such is the state of our world. We live too much inside of our own bubbles. We imagine others in our own mode. We are so used to projecting ourselves onto others, to the point that we no longer have an accurate reading of what’s actually going on.

This is why we sometimes see a great sense of mission in what we at Baiguan are trying to do.

In a world where walls are growing ever taller, we are still naive enough to think that we can break the wall, at least for some people.

To make that happen, simply providing data and research, or writing newsletters, is far from enough. We have to make it real.

This is why we have come to see offline events and study tours as an increasingly important part of our business. After organizing several successful tours to China, we have come to appreciate the unique value of offline, human-to-human interactions in enhancing understanding among people.

Such is the context for our upcoming 2026 autumn tour. You can read about all the details here.

My personal favorites include:

• Meet the senior management of Moonshot (Kimi)

• Meet the Founding team of Zhipu (GLM)

• Have a rare glimpse into Chinese commercial space companies

• A rare glimpse into Chinese nuclear fusion companies

• Test-drive one of the best electric racecars out there

• Be accompanied by for a private tour

• Visit ByteDance, Alibaba, and Tencent’s cloud and AI arms

• Visit the twin capital of global commerce: Huaqiangbei and Yiwu

• Check out the emerging Chinese luxury brands, who are redefining global luxury

• Visit the place where Buddhism and nomadic culture fused into what eventually became today’s China

• Visit the former capital of China (and my birthplace!)

Baiguan - China Insights, Data, Context

RSVP Now: 2026 Autumn Baiguan China Tour “Marathon”

We are witnessing a paradigm shift in global interest around China…

Read more

5 days ago · 12 likes · Robert Wu

To ensure the quality of our tours, they are by application only. You can apply by filling in this form here.

Next Wednesday and Thursday, I will also hold two info sessions and live Q&As to answer any questions you have. Please choose the time slot that matches your time zone and make sure to mark your calendar!

Session 1

• Wednesday, Aug. 19

• 9 pm Singapore / 9 am ET / 2 pm London

• Meeting Link: https://us06web.zoom.us/j/85232827349?pwd=znYbS32pEsLsoE74SxuWarUh8KC15A.1

Session 2:

• Thursday, Aug. 20

• 4 pm Singapore / 9 am London / 4 am ET

• Meeting Link: https://us06web.zoom.us/j/82363623800?pwd=rrI8dkb8RoTWjVGRdWNTdVy7Lc1dS8.1

Subscribe now

On July 28, the Federal Communications Commission added foreign-produced “advanced robotic devices” to its Covered List, defining the category as mobile robots such as humanoids and quadrupeds, but also including household items like robot vacuums. Formally, the restriction applies to equipment manufactured in any foreign country. In practice, however, non-Chinese suppliers are expected to receive exemptions, making it an effective ban on new Chinese robot models.

Then, Reuters reported on August 4 that the FCC was drafting a measure to block imports of new Chinese optical modules. These components convert electrical signals into light and carry data over fiber between the servers, switches, and accelerators inside AI data centers. As optical networking becomes increasingly important to the construction of large AI systems, this sector has ushered in some of China’s most valuable companies, occupying a prominent position in Baiguan’s “China Global Champions 30” index.

The Chinese response was as swift as in last year’s trade war. The Chinese Ministry of Commerce responded on August 5 with a package of targeted countermeasures. Among other things, Beijing tightened licensing reviews for US-bound exports of controlled drones, key components and related technologies, removing the streamlined approval process previously available to these shipments.

Also in the last week, Forbes broke the supposedly “shocking” story that American AI data vendors have been selling training datasets to Chinese large language models.

This type of “accusation” is quite interesting, because at the same time, the same Chinese AI labs are accused of “distilling” from American large language models.

As very eloquently laid out in this article about why Chinese AI labs distill, distillation is mainly used as a shortcut to get high-quality training datasets. It’s a “data strategy”, as opposed to getting underlying algorithms and weights, which is impossible to distill.

Now, if distillation is not allowed, at least you will think buying data through proper channels is fine?

Turns out, it’s not fine. The logic of America today, both its government and its public opinion, appears to be that they should stop selling or giving ANYTHING that might potentially help the Chinese AI industry grow, even while Chinese open-weight models have been widely adopted in the US.

But why should we stop there? We should make it illegal for American firms to use Chinese open-weight models as well.

And it’s not just the US. We should ban all of our allies from using Chinese equipment, just as Chris McGuire, the senior fellow of the Council on Foreign Relations, proposed in response to the Reuters story that Samsung and SK Hynix are testing Chinese chip tools as a hedge against US risks.

At this point, it’s difficult not to do a linear extrapolation into the future and see that at the current rate, we will be heading toward a complete breakdown of the trade relationship between the US and China. No buying, no selling. Don’t touch me!

It’s also hard not to spot a pattern here: in these episodes of trade tension between China and the US, China is always the passive one, on the receiving end. The US introduced tariffs; China responded with equal measures. The US banned chips; China banned rare earths. The US banned robots; China struck back. The US always deals the “tit”, while China responds with that “tat”.

I struggle to think of any one instance in the past few years that China struck FIRST, while the US retaliated second.

It’s also clear who is embracing openness, and who is closing up. The US AI companies don’t allow their AI to be accessed by Chinese companies, but Chinese companies are announcing open-weight models every day.

Clearly, there is a total asymmetry of self-confidence, and of paranoia, between the US and China now.

Subscribe now

Asymmetry of Paranoia

For the US, China is an "adversary," which is just a slightly more charitable word than "enemy." In the Chinese business and technology circles I know, the language is different. The US is certainly a competitor. It is also seen as a power that is increasingly going mad. But it remains a great country from which China can learn.

There is paranoia in China’s security establishment, of course, but it has not penetrated ordinary business and technology conversations to the same extent. For instance, if you read what Liang Wenfeng said about the US, even behind closed doors, he did not sound as paranoid about the US as Dario is about China at all.

Why is today’s America so much more paranoid about China than the other way around?

Perhaps this is simply the psychology of an incumbent and a challenger. The incumbent experiences every exchange as a possible loss of advantage. The challenger experiences the same exchange as an opportunity to learn. One tries to preserve its lead by reducing contact, while the other tries to close the gap by increasing it.

But this explanation does not make the result any less ironic.

For most of the modern era, openness was central to America’s identity and its claim to superiority. Open markets, open scientific exchange, the free circulation of ideas and talent. These were evidence of American confidence. A powerful country could afford to remain open because it believed that it would gain more from exchange than it would lose.

Closed countries behaved differently. They banned foreign products, restricted information, pursued self-sufficiency and treated outside influence as a threat. Their fear of exchange was supposedly evidence of their weakness.

That is the deepest irony of this trade war. America’s opposition to China has become so total that it is gradually adopting the very characteristics it once presented as proof of the Chinese system’s inferiority, and perhaps is even on the path to surpass China in this regard.

I enjoy writing, not least because writing is a form of thinking exercise for me, and I try to write everything myself. However, writing is only my part-time job, alongside my full-time CEO job. To strike the right balance between writing on my own as much as I can and publishing in time anything at all, some articles in this newsletter are written by AI, but always with 1) my own core thesis and skeleton, 2) internalization of my style, and 3) my final proofreading of each word before hitting Publish]

DeepSeek’s Liang Wenfeng is probably the least publicly visible founder of any major AI company today. Sam Altman and Dario Amodei appear regularly on podcasts, at conferences, and before policymakers. We know their arguments, their mannerisms, and sometimes far more about their internal conflicts than we would like.

In contrast, Liang has given only a handful of interviews, mostly in writing. We don’t even know what Liang’s voice sounds like. Yet, DeepSeek has somehow become one of the most consequential AI companies in the world while its founder remains almost entirely silent.

That is why an alleged transcript of a nearly four-hour meeting between Liang and DeepSeek’s investors caused such a sensation when it circulated online late last month. The transcript contains answers on everything from artificial general intelligence to Huawei, Nvidia, open-weight models, pricing, corporate governance, and the future structure of the AI industry.

DeepSeek has not confirmed the transcript, so every quotation attributed to Liang should still be read with that qualification. I cannot prove that it is real either.

My instinct, however, as soon as I got my hands on it, is that it probably is.

The simplest reason is that it sounds remarkably consistent with everything DeepSeek has done, or with whatever little clues Liang has previously let out into the public domain. In a rare earlier interview, Liang described open source as a cultural behavior rather than a commercial calculation. The company has repeatedly released model weights, papers, and production-tested code. Its official model repositories allow commercial use of many releases. It has kept prices extraordinarily low even when demand would probably have supported much higher ones. Its own website still describes the company as being “dedicated to exploring the essence of AGI”.

Then, the plot thickens. A few days after the transcript appeared, Bloomberg reported that DeepSeek had paused its second fundraising round, driven in part by Liang’s frustration over the leak. A new alleged story piles on another.

Who is the “good guy” in the AI race?

The American debate about China and AI usually begins with a familiar premise: the United States must win because it cannot allow the Chinese government, or people working under the Chinese system, to control such a powerful technology.

Yet, if we compare Liang with some of the leading American AI founders purely in terms of disposition, restraint, and the moral values we would want behind the wheel of AGI, it becomes difficult to say who the “good guy” is.

Again, we know very little about Liang. Secrecy naturally creates suspicion, especially in the current geopolitical environment. Yet the limited evidence we do have presents a founder who already had enough money, preferred difficult scientific problems to a celebrity life, and built a company around a narrow mission. In the leaked transcript, he repeatedly says DeepSeek will avoid attractive businesses that do not advance the path toward AGI. The company does not use conventional KPIs for researchers. It wants employees to retain time for unscheduled exploration. Liang also famously describes restraint itself as a strategy.

This kind of person is a minority in China. But it’s a growing minority.

Most Chinese entrepreneurs, like most entrepreneurs everywhere, care a great deal about profits. But I increasingly see a newer generation of founders, who were born into relative comfort and for whom money becomes the means rather than the end. (POP Mart’s Wang Ning is another case in point.)

Liang’s self-made wealth as a quant boss gave him the freedom to start DeepSeek without asking investors for permission. More importantly, it gave him the freedom to define success in his own way.

That freedom shaped the company before outside capital arrived.

“Open weights” as the objective

Many technology companies use open weights models to recruit developers, win attention, or pull users toward a paid product. For them, openness is one stage in a commercial funnel.

DeepSeek, however, seems to judge success partly by how widely its models can be used, including by people who will never become DeepSeek customers.

The alleged transcript makes that philosophy unusually explicit. Liang says DeepSeek intends to release even its most advanced models, wants other companies to deploy them successfully, and does not keep a superior private version for itself.

What’s remarkable in the transcript is that he laid out a clear maximum profit goal (10-month payback and 6-fold return on investment), so the open-weights stance does not seem like an empty promise but an intentional, calculated bet. In his own telling, the more restrained company has a better chance of reaching AGI because it avoids wasting energy fighting for every adjacent source of revenue.

In other words, open weights are part of the mission itself.

This is close to what the name OpenAI originally meant. In its 2015 founding announcement, OpenAI described itself as a nonprofit intended to build value for everyone rather than shareholders. Its researchers would publish papers, code, and patents, and collaborate freely with other institutions. Its charter still says its mission is to ensure that AGI benefits all of humanity.

OpenAI has faced pressures and responsibilities that DeepSeek has not yet encountered, and DeepSeek’s own ideals may eventually collide with commercial reality. Still, judging from their present behavior, the Chinese company looks surprisingly close to the original OpenAI proposition.

The Chinese government has also begun to recognize the geopolitical value of this approach. At the 2026 World AI Conference in Shanghai, Xi Jinping emphasized openness, inclusive development, and support for AI capacity in the Global South. China’s Foreign Ministry subsequently described the country’s open-source AI ecosystem and AI-related international public goods as tools for narrowing the global AI divide.

That language is self-interested, of course. All soft power is. But cheap, downloadable Chinese models offer something tangible. A developer in Indonesia, Brazil, or Nigeria can run the weights locally, modify them, and avoid permanent dependence on a handful of American platforms. This is a more useful form of soft power than another nice speech about “win-win cooperation.”

Is “intelligence” a commercial product, or a public good?

It’s tempting to compare China’s AI strategy with its manufacturing playbook: enter at a low price, improve quality quickly, and eventually offer a product good enough that the price gap becomes decisive.

That’s the more friendly way to put it. The unfriendly way will call this strategy “dumping”.

The analogy is useful, but AI may go further. A vacuum cleaner or an electric vehicle is just a consumer product. But “intelligence” could become a basic input used by every person and every business, closer to what the internet, electricity, or water are today. If that happens, the social value of making intelligence extremely cheap may outweigh the value captured by any one model company.

On this topic, Liang sounded almost Hegelian. In the leaked transcript, he said:

This differs from open-sourcing a piece of software in the past, because that software’s market wasn’t so large, while AI is simply too big. If we tried to monopolize the benefit, we would certainly be discarded by history. Most fundamentally, I think this is an objective law — a view of history.

这跟之前开源一个软件可能是不一样的，因为那个软件的市场没那么大，而 AI 这件事实在太大了。如果我们想独占这个利益，那一定是要被历史抛弃的。我觉得最主要的是，这是一个客观规律，这是一种历史观。

DeepSeek’s pricing suggests that it is already operating with this possibility in mind. According to the leaked transcript, its API prices are designed to recover equipment investment in roughly ten months and earn what Liang considers a “reasonable return.” He says the company could charge substantially more without losing much usage, but chooses not to, and when DeepSeek once cut a model’s price to one quarter of its original level, the team “celebrated.”

The Nvidia triangle

The leaked transcript becomes more politically awkward when Liang turns to semiconductors. He reportedly says Nvidia is “digging its own grave” by allowing export controls to accelerate Chinese substitution. He also claims that DeepSeek used Nvidia hardware to train V3 while reducing its dependence on Nvidia’s software stack, and that the company is working closely with Huawei’s emerging ecosystem.

Some of the specific technical claims in the transcript, especially its comparisons between Huawei and Nvidia hardware, remain unverified. Yet, the direction is much easier to see. DeepSeek is helping Chinese models run efficiently on a technology stack that depends less on CUDA and advanced American chips.

I recently described this as the “galapagosization” of Chinese technology. Export controls create short-term pain, then give Chinese companies a powerful reason to develop alternative hardware, software, interconnects, and standards. DeepSeek sits on the software side of that emerging ecosystem.

The interests of the three main actors now point in different directions. China wants distance from Nvidia. The US government also wants Nvidia to keep its distance from China, but for a different reason. Nvidia itself wants to remain at the center of both ecosystems.

This is why Jensen Huang sounds so frustrated when American policymakers tell him the China market will be lost anyway. Jensen understands that technology competition is perpetual. If Chinese developers keep building on Nvidia, the American ecosystem retains scale, influence, and visibility. Forcing them away does not stop development. It helps create a second ecosystem that the United States can neither control nor easily observe.

Liang’s reported language is more combative than Jensen would like, but the two men appear to agree on the mechanism.

Capital comes after the mission

DeepSeek now faces a contradiction of its own. I previously argued that its success was possible precisely because the Chinese government did not create it and venture capital did not control it. I went further in Baiguan: had DeepSeek taken conventional VC funding at the beginning, investors might have pushed it toward user growth, monetization, and a faster exit before its research culture had time to mature.

DeepSeek has now taken outside money. According to The Information, as relayed by Reuters, its first round raised more than RMB 50 billion, or about $7.4 billion, at a valuation above $50 billion.

I do not think this invalidates my earlier argument. Sequence matters. DeepSeek accepted investors after its mission, culture, and technical credibility were established. The reported deal placed most investors into a limited partnership controlled by Liang, imposed a five-year lock-up, and gave them no voting rights. (China’s national AI investment fund was the notable exception.) These are unusual terms for an unusual company.

The money also serves a clear purpose. DeepSeek needs compute, and it needs to keep a small group of researchers who could command much higher compensation elsewhere. In an earlier account of the fundraising meeting, Liang’s most important demand to investors and large technology companies was simple: do not poach his people. The leaked transcript says team stability is the company’s main non-negotiable interest, and that financing has reduced the risk by giving employees meaningful equity.

Capital has become a means of protecting the mission. Whether it stays that way is now the central test.

Bloomberg reported that the paused second round could resume later and that DeepSeek had been seeking a valuation around RMB 500 billion. Reuters has also reported early deliberations about a possible listing on Shanghai’s STAR Market. Neither step is a confirmed company plan. Yet a domestic listing would make sense if DeepSeek needs liquid equity to retain talent and finance compute without surrendering strategic control.

The company does not need to remain profitless to remain mission-driven. Liang’s stated goal is a reasonable profit and a durable commercial foundation. The real question is what the company chooses to optimize when revenue, market share, research openness, and its AGI mission eventually pull in different directions.

That test has broken many idealistic organizations before.

When OpenAI was founded in 2015, it said a leading AI institution should prioritize a good outcome for everyone over its own self-interest. Nearly eleven years later, one of the companies making the most serious attempt to operationalize that idea is a secretive Chinese laboratory created by a quant billionaire and now financed partly by Chinese state capital.

DeepSeek may still disappoint me. Its culture could change as it grows. Political pressure could narrow its openness. The alleged transcript could contain errors or inventions. But the vision inside it is coherent with the company’s behavior: make the models open enough to spread, cheap enough to become infrastructure, and profitable enough to keep the research going.

I hope DeepSeek ends up becoming the real OpenAI. That is what I am rooting for.

Subscribe now

Subscribe now

The CXMT IPO

In a highly anticipated event, Chinese memory maker CXMT finally became a listed company. As expected, it became the most valuable company listed across both A-share and H-share on its first day of listing, surpassing ICBC and Tencent.

From an issuance and market-stability perspective, this was a major success.

Things could easily have gone much worse. The nightmare scenario was a repeat of PetroChina’s 2007 IPO: shares were immediately bid up to a stratospheric valuation, then began an excruciating decline that made the company a symbol of the market top. PetroChina, you may have forgotten, briefly became the world’s first trillion-dollar public company, at least on paper.

It is far too early to say that CXMT has escaped PetroChina’s eventual fate. But the immediate disaster many investors feared, a giant IPO draining liquidity and crushing the market around it, did not happen, despite the turmoil and blow-ups in the memory and semiconductor sectors internationally in the past few weeks, hastened by a certain German Wunderkind and some maximally leveraged Korean gamblers.

It's a success story for the Chinese securities regulator as well. It’s clear that ensuring a successful IPO for CXMT is of paramount importance for regulators. On the eve of the IPO, when market sentiment noticeably soured along with the broader fall in global AI sectors, securities regulators and the “National Team” came out hard with clear messaging and hard cash to support the market.

I have long observed that the capital market has transitioned from a secondary matter to a primary policy item, and Chinese regulators’ skill at managing market expectation has grown rapidly. CXMT’s debut is another data point supporting my claim.

Also in the spotlight is Hefei. Long overshadowed internationally by cities such as Shenzhen and Hangzhou, Hefei has gradually built a reputation as one of China’s most consequential technology and industrial-investment hubs. CXMT’s IPO has now pushed that reputation much further into global view.

Measured by the combined market capitalization of locally headquartered A-share companies, CXMT alone catapulted Hefei from 19th place to fourth, behind only Beijing, Shenzhen, and Shanghai.

The “Hefei Model” is quite unusual in that it’s where government venture capital places massive, concentrated, and long-duration bets on strategic industries, and several of these bets paid off spectacularly. This is actually amazing, because government funding tends to be risk-averse and isn’t usually associated with roaring success stories.

Naturally, people may ask, given Chinese local governments’ tendency to neijuan, will more cities in China follow suit?

Well, I wrote a long article about it two weeks ago, but my conclusion is simple: No. This is probably one of the rare cases when, although one city may have a stellar success, it’s not easy for other cities to follow suit and copy.

You can replicate success and scale it up if capital is the main constraint. But when it’s serendipity, people factor, and something as rudimentary as “having guts” that’s at stake, there is really no clear roadmap.

You can find more details about it here:

Baiguan - China Insights, Data, Context

The "Hefei Model" behind CXMT: a rule or an exception?

On July 15, ChangXin Memory Technologies, better known as CXMT, is scheduled to begin book-building for what may become China’s largest semiconductor IPO. Investor subscriptions will follow on July 16. The company plans to raise RMB 29.5 billion, or about US$4.3 billion, to upgrade its DRAM production lines and develop new memory technologies…

Read more

a month ago · 19 likes · 1 comment · Robert Wu

I want to do this kind of short, single-issue briefings more frequently. Here are some other recent events that I find interesting. Which ones do you want to hear my takes about?

At 12:15 p.m. on July 10, the Long March 10B lifted off from the Hainan Commercial Space Launch Site. It successfully delivered a satellite into orbit. About six minutes after separating from the second stage, its first-stage booster returned vertically toward Earth and was caught by a giant net stretched across a recovery platform at sea.

It was China’s first successful controlled recovery of an orbital-class rocket booster, and the first time anywhere in the world that such a booster had been recovered using a net. China is now the second country, and roughly the third organization after SpaceX and Blue Origin, to accomplish this feat. The recovered booster is expected to fly again before the end of this year. Only then will it graduate from “recovered” to genuinely “reused.”

When videos of the recovery began spreading across overseas social media, they were immediately met with familiar forms of skepticism. Some people assumed that the footage had been generated by AI. Others said that China had, once again, stolen the technology.

Let us first dispose of the stupidest part of this discussion: the event was real. The rocket launched. The satellite reached orbit. The booster returned. The recovery platform caught it. The event was documented from multiple angles and reported by Chinese and international media. We are now living in an age when extraordinary footage is presumed fake until proven otherwise, but disbelief is not evidence.

The more interesting accusation is that “China stole SpaceX's technology.”

This is also wrong, although for a more instructive reason.

SpaceX actually uses two different approaches to recover its rockets. Its workhorse Falcon 9 returns on deployable landing legs, touching down either on land or on an autonomous drone ship. For Starship, SpaceX eliminated the landing legs and instead catches the Super Heavy booster with two mechanical arms attached to the launch tower, the system popularly known as the “chopsticks.”

China’s Long March 10B uses neither.

From the footage, it seems that Long March 10B eliminates the Falcon-9-style heavy landing legs, but instead of requiring the booster to thread itself between two rigid mechanical arms like Super Heavy, the rocket deploys hooks near its lower section as it descends. These hooks engage with tensioned cables and a large flexible net installed on the sea platform. The net and its buffering system absorb much of the booster’s remaining kinetic energy.

The broad objective is the same as SpaceX’s: do not throw away the most expensive part of the rocket after each launch. But the engineering architecture is clearly different, and in fact may offer several advantages.

On the one hand, it shifts part of the recovery burden from the rocket to the platform. Without landing legs, the booster can save weight and devote more of its capacity to payload. On the other hand, a flexible net may also tolerate a larger landing error than a rigid platform, reducing the precision demanded of the rocket during the final seconds of descent.

The trade-off is that recovery depends on specialized infrastructure and introduces a different set of engineering problems involving sea conditions, cable dynamics, hooks, and the stabilization of a very tall object suspended in a net.

Of course, China learned from SpaceX. SpaceX demonstrated that a large orbital booster could return through the atmosphere, restart its engines, navigate toward a small target, and land in a condition suitable for reuse. China’s rocket follows the same basic sequence and is governed by the same physics. But physics does not belong to Elon Musk. Neither does the general ambition of making rockets reusable.

Nor is most of the underlying science secret. Rocket propulsion, aerodynamics, guidance, materials science, and control theory are all well-established fields. The difficulty lies in systems engineering: making thousands of components behave correctly simultaneously, under extreme heat and pressure, with almost no tolerance for error.

That requires enormous amounts of tinkering, testing, and failure. China is very good at that kind of engineering.

But there is one thing China did take from Musk, which also contributes to the lingering perception that China is copying him, and is the single most important contribution Musk made to China’s technological innovation. It is just one sentence:

“This can be done.”

What China really took from Musk

A striking aspect of the technological competition between China and the United States is that it sometimes feels like a competition between China and a single American.

Tesla proves that an electric car can become a globally desirable consumer product. China, already investing in electric vehicles, gains a powerful market benchmark and accelerates its efforts.

SpaceX proves that reusable rockets are commercially meaningful. China develops its own reusable rockets.

SpaceX builds Starlink. China begins constructing competing low-Earth-orbit satellite constellations.

This description is obviously exaggerated. The United States has many other important technology companies, while China’s industrial strategy is not created in response to one man’s Twitter feed. But the pattern is real enough to deserve examination.

Musk is unusually willing to pursue projects that initially look absurd. Once one of them works, China is unusually good at turning that proof into an industry.

Consider electric vehicles.

BYD was not created in response to Tesla. It began as a battery manufacturer in 1995 and entered the automobile industry in 2003, the same year Tesla was founded. China had already made electric-vehicle research a national priority in 2001, years before Tesla sold its first Roadster. BYD launched its first plug-in hybrid, the F3DM, in 2008.

It would therefore be historically wrong to say that Musk introduced China to electric vehicles.

What Tesla provided was something subtler and perhaps more consequential. It demonstrated that an electric car need not be an inferior substitute purchased out of environmental duty. In fact, an EV could be faster, more desirable, and more technologically advanced than a gasoline car. Tesla turned the electric vehicle from a fringe research program into an object of consumer aspiration.

The arrival of Tesla’s Shanghai factory in 2019 reinforced that effect. Chinese policymakers were willing to give Tesla unusually favorable treatment partly because they expected it to function as a “catfish,” forcing domestic manufacturers to swim faster. Tesla became a visible benchmark for vehicle software, manufacturing efficiency, direct sales, and consumer appeal.

Chinese companies did not simply reproduce that benchmark. They scaled and refined plug-in hybrids, built battery-swapping networks, integrated supply chains, lower-cost batteries, and a bewildering variety of models aimed at almost every price segment. They are now forcing Tesla to respond.

This is often how Chinese technological development works. China may begin as a follower, but it rarely remains a passive imitator. Once the direction has been validated, competition, manufacturing scale, and rapid iteration transform the original concept.

The value of external proof

China does not lack imaginative scientists, ambitious entrepreneurs, or technically difficult ideas. What it often lacks is permission to commit enormous resources to an idea before anyone knows whether it will work.

Any new technological frontier requires a coalition. Government agencies must approve it. Investors must finance it. Suppliers must reorganize around it. Local governments may need to provide land and infrastructure. Talented engineers must be persuaded that the project will survive long enough to justify years of their careers.

In China’s consensus-heavy system, the hardest part is not solving the engineering problem. It is persuading enough stakeholders that this particular engineering problem deserves to be solved.

If you really think about it, the downside of supporting an eccentric and unproven project is tremendous. If it fails, everyone involved may be asked why they wasted so much money on an obviously unrealistic idea. Everyone around them will be laughing at them. On the other hand, the upside is harder to claim in advance, so this whole dynamic naturally encourages people to wait.

However, external success changes the political economy of the decision.

Once SpaceX has landed hundreds of rockets, reusable launch vehicles will no longer be one billionaire’s fantasy. They become strategic infrastructure. The question changes from “Why should we attempt this?” to “Why are we still behind?” Budgets become easier to justify. Investors become more comfortable. Engineers become easier to recruit. Failure becomes acceptable because the destination itself will no longer be in doubt.

This is what Musk gives China: social proof.

The same dynamic may eventually play out with space data centers as well. Musk argues that orbiting computers powered by solar energy could escape the electricity, land, and water constraints faced by terrestrial AI infrastructure. SpaceX has even filed for permission to deploy a constellation that could eventually contain as many as one million computing satellites. For now, the economics remain deeply uncertain, and SpaceX itself has warned investors that orbital data centers may never become commercially viable.

China is not ignoring the concept. In fact, it launched the first 12 satellites of its “Three-Body Computing Constellation” in May 2025 and has already tested AI models and inter-satellite computing in orbit. China’s main space contractor has also proposed gigawatt-scale space computing infrastructure.

Still, I doubt China will rush to achieve it before SpaceX and would rather wait a bit. Yet, if SpaceX eventually proves that orbital computing works economically at scale, I assure you China’s response will be swift.

Where is China’s Elon Musk?

I believe China may eventually produce its own version of Elon Musk: someone bold enough to choose a frontier before the rest of society considers it sensible, and someone whom the system is willing to tolerate through repeated failures.

But that person does not fully exist today. This is not because Chinese entrepreneurs are less intelligent or less ambitious. It is because an Elon Musk requires more than a particular personality. He requires an environment in which one eccentric individual can attract extraordinary amounts of capital, ignore conventional opinion, survive spectacular failures, and continue setting goals that no government planning document has yet endorsed.

China has proven to be capable of accepting enormous execution risk once a goal has become a national priority. Yet it is less comfortable to allow one individual to decide, largely on his own, what the next national priority should be.

Thus, in this “China-Musk” tech competition, a clear dynamic has emerged: Musk acts as the adventurer-validator who selects a frontier, absorbs ridicule, and demonstrates that it can be crossed. China follows, organizes, and scales. It mobilizes engineers, supply chains, capital, and government support until the frontier becomes an industry.

This division of labor is not fixed. Followers are constantly learning, and scale often produces its own innovations. Eventually, the country that began by copying the direction may discover the next destination.

The Long March 10B is already an example. China did not invent reusable rockets, but it did not merely reproduce SpaceX’s landing system either. It took the validated objective and developed a different way to accomplish it.

I believe the world needs both kinds of players. It needs people willing to pursue bold ideas before they appear reasonable, and societies capable of turning those ideas from expensive demonstrations into affordable, widely deployed infrastructure.

China has proved that it can do the second. I hope that one day it will also give its most unconventional talents enough freedom to do the first.

Subscribe now

[I wrote the skeleton of this essay, with some details filled in by AI]

The idea that I consider completely ridiculous is, why should American companies go compete in foreign countries? You’re going to lose it anyway.

You’re going to lose it anyway, so why go?

Well, if you guys all applied that same philosophy, why wake up in the morning?

I don’t prescribe to ‘we are going to lose anyways.’ I don’t prescribe to that.

If you want me to lose, you’re going to have to deal it to me. But I’m going to have to put up a good fight.

And I put up a lot of fights over the years.

I am doing okay.

- Jensen Huang, Stanford University CS153 guest lecture, May 2026

We have to keep innovating and… The premise that even if we competed in China, that we’re going to lose that market anyways… You’re not talking to somebody who woke up a loser. That loser attitude, that loser premise makes no sense to me.

- Jensen Huang, on Dwarkesh Patel Podcast, April 2026

In the last post on China Translated, I went through how, in response to the US containment policy, Chinese tech leaders like Huawei and DeepSeek are apparently developing along alternative technological paths by bypassing their relative weaknesses, which have the potential to eventually create a whole new ecosystem across hardware and software that’s independent of the American tech stack.

I also drew on the experience with Chinese EVs to make my point. I reminded you that China tried but failed to surpass the West in traditional ICE cars. Instead, China chose the alternative path of electrification, rendering more than a century of accumulated ICE IP irrelevant and overtaking the Western automobile industry “on the bend.”

It’s very likely that the same EV experience would repeat itself in the semiconductor world, which begs the question of whether US export controls will work at all.

A core assumption behind the export control policies is that if China becomes a more technologically advanced power, it will use that power to hurt the US. In the Dwarkesh Patel podcast episode with Jensen Huang that I kept revisiting for the last essay, Dwarkesh kept invoking the possibility that China would use “Mythos-level” power to attack the US, and a direct analogy was made between AI and nuclear weapons.

Let’s just set aside how ridiculous this assumption is for now, despite the fact that China is not a terrorist power seeking war and conquest. Let’s just assume it’s ALL true. Let’s assume China’s ultimate goal is to destroy the US, just for argument’s sake.

What choices does the US have? Option one is export controls, and doubling down on them (not in the half-hearted way that is reportedly being run right now). Doing this will for sure cause short-term pain for China, but will also force the “galapagosization” I mentioned. So not only can the US not prevent China from having the ability to launch attacks, but there’s also the risk that China can attack using a tech stack that’s entirely foreign to the US, not controlled by the US. Even worse for the US, they will have zero visibility into it.

Obviously, Option 1 is a net loss for the US, however you look at it.

Option 2: trying to get China hooked to the American tech stack. China will develop, but so long as the US itself is also developing at a similar pace, China will at least be developing within the US's control, influence, and visibility. Option 2 does not guarantee US success forever, but at least it won’t guarantee failure as Option 1 does.

This is all very simple logic to me, with a very small margin of error. What baffles me is why this is not common knowledge for everyone, and that when folks like Jensen Huang make similar arguments, he is always met with skepticism and criticism from his home country.

Fighter vs talker

Time to put up a disclaimer here: I am not an American. So what comes next in this essay should be taken with a grain of salt. Furthermore, this newsletter is about explaining China, and so what comes next is not what it should normally cover.

Just treat this as a small role-play, with me as a strategist for the US, thinking in the US's best interests, trying to understand why the US has chosen to sabotage itself, not heeding the call from its best and brightest, like Jensen.

Dwarkesh put Jensen on the edge in that podcast episode, and Jensen appeared unusually combative and started to call people “loser”. The feedback to Jensen from that episode is overwhelmingly negative. It’s like a public trial, where Jensen had a chance to explain himself to America. But in a culture that prizes rhetoric, his blunt language and aggressive behavior have won him more haters than lovers.

But here is the thing, and I talk about this from my perspective as a fellow entrepreneur myself here: Jensen and the people Dwarkesh represents are two different types of animals, with entirely different life experiences and language systems.

Jensen is a fighter. The people who don’t get him are losers — I am sorry, I mean “talkers.”

A fighter fights. A talker talks. A fighter has a winner’s mindset: to compete, to survive, and to be better than everyone else. A talker’s main drive is to get comfortable in the palace of circular logic they build for themselves.

I see most of the criticisms of Jensen come from the talkers - DC think tankers, journalists, and academics - who often lack real-world knowledge of how things work.

When it comes to the world of technology, there are at least 3 basic things a talker doesn’t understand:

1. Technology is never a moat.

Business strategist and investor Hamilton Helmer has a theory of “7 Powers” that explains where business moats come from. They are:

• Scale Economies: the ability to spread massive fixed costs over a vast customer base, meaning a competitor with the exact same technology must charge exponentially more per user just to break even. (e.g., Walmart, Cargill)

• Network Economies: the entrenched user base; even if a rival builds a technologically superior platform, users will not switch if all their friends, suppliers, or customers remain on yours. (Facebook, MasterCard)

• Counter-Positioning: the situation where the incumbents might understand your tech perfectly, but they cannot adopt it without cannibalizing their highly profitable legacy revenue streams. (How Apple beat Microsoft and Nokia in the age of mobile internet)

• Switching Costs: the operational friction and financial pain required to replace your ecosystem; customers will stick with your product if moving to a competitor means retraining their entire workforce or rewriting custom code. (Imagine the pains you will take to switch out of an iPhone to Android, or vice versa)

• Branding: the compounding trust and emotional resonance built over time, allowing you to consistently charge a premium over competitors who offer identical or even better technical specifications. (Coca-cola, S&P)

• Cornered Resource: the preferential, exclusive access to a scarce asset (Think any natural monopolies that need special licenses to operate)

• Process Power: the opaque, complex, often undocumented web of institutional knowledge and hard-to-replicate operational processes that lead to lower costs or superior products. (ASML, TSMC)

It’s very important to note here that none of the 7 powers of business moats comes from technology!

Why? Because, although technology is the spark that creates temporary value, it never stays. Because talents eventually flow. Knowledge eventually diffuses.

Technology can be easily reverse-engineered, copied, or leapfrogged by the next iteration. If your only advantage is having the best algorithm, the fastest chip, or the cleanest code, your margins will eventually collapse as competitors eventually catch up - and competitors always catch up.

The real moat is something that even if you spill your secret formula all over the place, you are still unbeatable. For instance, it will be naive to believe that Coca-Cola’s business power comes from its “high-tech”, “secret” recipe. No, its power comes from its tight grip over consumer consciousness and a massive distribution network it has built over the century.

This is why forcing Nvidia to stop selling to China on the grounds of superior technology is shortsighted. It protects something that is unworthy to be protected in the first place, while creating substantial harm to at least 6 actual superior structural powers that Nvidia (and by extension the US) enjoys:

• Forcing Nvidia out of China makes it relatively subscale compared to the ideal scale it otherwise could have enjoyed, preventing it from spreading its R&D costs across a larger revenue base to the fullest extent and eventually hurting Nvidia’s long-term ability to set prices in the most competitive way.

• The real strength of Nvidia is not just a chip, but a web of ecosystem consisting of hardware, software, networking standards, and knowledge that constitutes a real network effect. If you want to change one thing, you have to change everything else, making it extremely sticky. Now, the US is forcing China to develop its own ecosystem that keeps Nvidia out. At some point, Nvidia will no longer be the pre-eminent ecosystem in the world. An alternative ecosystem may rise, challenging Nvidia’s strategic position in the rest of the world.

• Giving the competing ecosystem a chance to grow naturally lowers switching costs away from the incumbent ecosystem and diminishes the brand equity that Nvidia would otherwise enjoy.

• China used not to have “counter-positioning” against the US. But now, by breaking the export link, Chinese firms like Huawei enjoy a unique counter-positioning to develop on their own, while Nvidia can do nothing to disrupt them.

2. Technology is not on a piece of paper.

I mentioned that none of the “7 powers” has anything to do with technology. I was not 100% correct.

I also mentioned that there are 6 powers at risk for Nvidia, but I have discussed only 5 above. The final one I haven’t mentioned is “process power”.

Process power can also be understood as some form of technology. In fact, it’s the only type of technology that’s worth a place on the list of Helmer’s 7 Powers.

People who live in the fanciful world of 007 movies like to imagine technology as a magical piece of information stored somewhere. Accessing it, you will unlock all the powers of the world. This is much like how medieval kings viewed gold as a fixed asset that you could lock in a vault and keep away from rivals forever.

The real world is much more boring.

Take TSMC as an example. All the physics in chip-making are known. There is nothing mysterious. But knowing the physics doesn’t mean you can run the factory.

Process power is derived from the factory floor. It is the opaque, infinitely complex web of institutional muscle memory that accumulates over decades. TSMC’s moat is not that they bought the best EUV lithography equipment, but that they have spent twenty years fine-tuning millions of interconnected operational steps to achieve a massive, profitable yield on a silicon wafer where a competitor with the exact same equipment would fail.

You cannot copy such power by stealing a blueprint. This is not something that you can fit on a USB drive. You cannot hire away one chief engineer and expect them to miraculously transplant it into your own company. This power, this “technology”, is deeply embedded in the culture, the daily habits, the unspoken routines, the suffering through the grueling and unglamorous years of building from scratch, and the trial-and-error scars of tens of thousands of mid-level engineers working in concert. It is operational excellence compounded by history.

So what do export controls do for process power? First, it gives Chinese companies a chance to develop their own process powers. And remember, process power is the only form of technology that matters. Once Chinese firms have built that power, export controls cannot take it away. You can restrict the sale of a physical chip, but you cannot embargo an entire organization’s fundamental ability to execute.

Second, it stops Western firms like Nvidia from participating in the same process of building up this new process power. It will be China's process power. It will be China’s technology, China’s moat.

3. No advantage stays forever. The game is perpetual.

Talkers believe that strategy is a board game. You capture the territory, you build a wall around it, and you hold it forever. They view American technological dominance as a static fortress that simply needs to be defended with lawyers, tariffs, and export decrees.

Fighters like Jensen know that technology is not a fortress. The past is not the future. Everything is in flux. No winning should be taken for granted. It’s a constant struggle. There is no finish line, and there is never a moment when you can sit back and relax. The moment you stop sprinting and instead choose to build a wall, you begin to lose, because you are refusing to join this constant struggle.

Winning is also a mindset. The crucial first step for a winner is to believe you can win, then compete with everything you've got, and never settle for whatever past glory you have achieved.

In this perpetual game, the only thing you can do to retain a plausible chance of winning is to keep running and leverage every tool at your disposal to add to your scale, your network effect, and process power to ensure that by the time your rival finally reverse-engineers today’s technology, you are already generations ahead.

I don’t know whether I will hope the US to take this advice or not.

Subscribe now

AI is a parallel computing problem, isn’t it? Why can’t they just put 4x, 10x, as many chips together… If they wanted to, they just gang up more chips, even if they’re 7nm.

…You could gang them together, just like we gang them together with NVL72. They’ve already demonstrated silicon photonics, connecting all of this compute together into one giant supercomputer.

… DeepSeek is not an inconsequential advance. The day that DeepSeek comes out on Huawei first, that is a horrible outcome for our nation.

- Jensen Huang, on Dwarkesh Patel Podcast, April 2026

At the May 2026 IEEE International Symposium on Circuits and Systems, Huawei unveiled the τ (tau) Scaling Law, sending shockwaves around the world.

For 50 years, the global semiconductor industry has relied on Moore’s Law: shrinking the physical size of transistors to cram more computing power onto a chip. Because US export controls have cut Huawei off from the advanced ASML EUV lithography machines required to shrink transistors past 5nm, Huawei seems to be shifting the goalposts. Instead of optimizing for size, the Tau Scaling Law optimizes for time: reducing the signal propagation delay (represented in physics by the Greek letter τ ) across devices and systems through LogicFolding technology, a 3D architecture that folds and stacks logic circuits vertically. Huawei states it will debut this architecture in its fall 2026 Kirin mobile processors and claims it will achieve 1.4nm-equivalent performance and density by 2031, all using older, legally available manufacturing equipment.

Let’s make it clear first: a lot of this is clever marketing, and Huawei has always known how to craft a compelling narrative.

What Huawei is proposing is not revolutionary in itself. It’s well known in the semiconductor industry that Moore’s Law has lost power. The world has hit the wall of physics in its quest to create ever-smaller transistors. To scale, the industry is shifting from “how small can we make a transistor?” to “how efficiently can we stack and connect them?” To save time and costs for moving data around, sophisticated stacking, advanced packaging, and system-level communication technologies increasingly matter more than merely carving out the tiniest transistors.

Shortly after the Huawei announcement, Jensen Huang noted that while tau scaling is an impressive breakthrough for Huawei, it does not threaten global foundries. TSMC has been pioneering advanced 3D packaging, chiplet stacking, and hybrid bonding for nearly a decade. Huawei is simply applying these existing concepts hyper-aggressively to compensate for its lack of raw node leadership. Yet their peers, through similar and perhaps currently more advanced packaging on top of their certainly more advanced nodes, still maintain a formidable edge over Huawei.

There is also the problem of thermodynamics. Stacking logic circuits directly on top of each other creates immense thermal management challenges. Dissipating heat from the center of a 3D folded chip is notoriously difficult, which will significantly complicate manufacturing processes, lower yields, and make mass production incredibly expensive. As of now, Huawei hasn’t yet demonstrated it can solve this problem with the right economics.

Furthermore, as points out, even if Huawei succeeds here, it’s very likely that Huawei still relies on Western bond manufacturers, such as Besi, a Dutch company, to make it work and is thus not free from potential US sanctions. After all, the Dutch are firmly within the orbit of US influence.

I am not belittling Huawei’s achievements here. I put out all these caveats first to suggest one thing: the actual significance of Huawei’s Tau scaling announcement is not that it has broken free from U.S. sanctions (it clearly hasn’t). The significance is that this announcement provides a roadmap for the future, and it’s a future that’s been played out many times already in history.

Overtaking on a bend - the EV experience

It’s extremely hard for a newcomer to fight against a formidable incumbent. To win, it’s often more desirable for a newcomer to enter a new game - a new paradigm - where everyone has a similar starting line.

The best example is China’s EV sector.

Automobiles are the crown jewel of modern manufacturing that any country with any decent dream of becoming a manufacturing powerhouse would have to win. But it’s also a sector dominated by century-old incumbents with deep, formidable institutional knowledge, complex process know-how, and brand moats that are nearly impossible to overcome.

China tried it before, but failed. Despite commonly held misconceptions about “forced technology transfer”, decades after establishing automobile joint ventures with Western giants like Volkswagen and Ford, China still failed to produce competitive cars powered by internal combustion engines.

A main reason is that although Chinese automakers can develop, their foreign peers are developing as well, and they are developing on top of an already more advanced foundation built through decades of iterations. In the end, for China, it became a desperate race with a target that is constantly moving further ahead. If China kept playing this game, it risked being forever stuck in a catch-up role.

Understandably, Chinese companies found this game to be both tiring and futile. So what did they do? They quit that game. They tried to open a new one.

The often-cited term for this in China is “弯道超车Overtaking on a bend”. In auto racing, overtaking on a “straightaway” requires sheer horsepower to beat an opponent who is already ahead. Overtaking on a curve, however, relies on taking a different line, using agility, and exploiting a change in the track's dynamics to bypass the leader.

The “straightaway” for China is the traditional internal combustion engine (ICE), and the “bend” is the electric vehicle.

China knows that the future of mobility belongs to electrification, for both environmental and energy security reasons. Yet, ICE incumbents indulged too much in their past glories and would be slow to adapt to this future.

And the playing field is level in this new game. Back in 2010, neither Toyota nor Volkswagen had a massive head start in lithium-ion battery chemistry, motor controllers, or smart cabin software. The century of ICE IP became suddenly irrelevant. On the other hand, China already had a massive consumer electronics manufacturing base and control over the refining of critical minerals (lithium, cobalt, rare earths), so it arguably enjoyed even some advantages.

What China eventually did is now well-known. It doubled down on this new game by nurturing champions and culling the laggards, building an entire ecosystem of supply chains around it, and eventually establishing an entirely new set of standards and benchmarks. When, finally, China emerged from Covid, the world suddenly woke to a new reality: Chinese EVs have overtaken legacy gasoline carmakers, on the bend.

What keeps Jensen Huang awake at night

Overtaking on the bend is exactly what Huawei is attempting here.

If the bottleneck of the semiconductor industry has shifted from inside the chip to system-level communication, why bother winning the old game? Why not enter this new game directly? After all, China’s disadvantage in communication and advanced packaging technology is a lot smaller than its gap with state-of-the-art lithography. Making the most advanced chips is the “straightaway”; bundling chips more tightly and making them communicate with each other at the speed of light is the new “bend”.

And perhaps China even has an advantage in this new game. As points out in his latest article on tau scaling:

There is another layer here that many Western analyses still underestimate: China’s structural advantages in systems engineering and infrastructure integration. Huawei is not merely a semiconductor design company. It is simultaneously a telecom infrastructure company, cloud company, networking company, AI systems integrator, smartphone OEM, operating system provider, and increasingly an AI platform company. Few Western firms possess that degree of vertical integration across infrastructure layers. Under a τ scaling framework, that integration becomes strategically valuable because optimization happens across the entire stack simultaneously.

Beyond tau scaling, one of the hottest trends today is CPO, or co-packaged optics, an industry inevitability that would drastically reduce communication costs within AI systems. But hey, don’t you remember that Huawei started out as a telecommunications equipment company? Also from Paul (emphasis mine):

CPO would allow Huawei to bring optical engines closer to switching and accelerator fabrics, reducing electrical trace length and improving bandwidth density. This plays to Huawei’s strengths: it has deep telecom, optical networking, silicon photonics, switching, and systems-integration capabilities that many pure-play AI chip firms lack. The strategic implication is that Huawei may use CPO not simply as a networking upgrade, but as part of a broader attempt to make Ascend clusters behave more like a unified, latency-managed AI computer, compensating for weaker process-node access through superior data movement.

It’s all but certain that what happened in Chinese EVs would happen again in AI. And just like the EV experience, it will ultimately lead to what I call the “galapagosization” of the Chinese semiconductor industry: because Huawei cannot use global-standard tools or fabrication processes, it is forced to build an entirely separate, vertically integrated ecosystem.

For example, in the tau scaling announcement, they specifically mentioned that they are developing bespoke EDA software and interconnected protocols that are completely isolated from and incompatible with the TSMC-Intel-NVIDIA supply chain. It’s also highly likely that in the quest to compensate for more inferior chip nodes with innovative folding and packaging technologies, Huawei would help usher in an entire ecosystem of material suppliers, equipment makers, and testing partners that cluster around their technology principles and standards. Their clients, Chinese AI companies, will also optimize their products around this new ecosystem, which helps it further diverge from the US tech stack.

Another thing we need to note is that while the Galápagos is a small archipelago, China is a continent. It’s the world’s largest single unified market with 1.4 billion consumers speaking the same language and tens of millions of businesses running on the same operating systems. The scale ensures that evolution will happen fast.

This galapagosization on a continental scale frightens people like Jensen Huang.

In the same episode with I quoted at the beginning of the article, Jensen said:

With respect to China… what we also want is to make sure that all the AI developers in the world are developing on the American tech stack, and making the contributions, the advancements of AI—especially when it’s open source—available to the American ecosystem. It would be extremely foolish to create two ecosystems: the open source ecosystem, and it only runs on a foreign tech stack, and a closed ecosystem that runs on the American tech stack. I think that would be a horrible outcome for the United States.

… A company developed software, developed an AI model, and it runs best on the American tech stack. I saw that as good news. You set it up as a premise that it was bad news. I’m going to give you the bad news, that AI models around the world are developed and they run best on non-American hardware. That is bad news for us.

Does technology containment work?

Jensen has been at the center of controversy lately. His petulant appearance with Dwarkesh Patel in that podcast episode is only adding to the controversy.

At the center of the controversy is the core US policy dilemma: whether semiconductor containment against China actually works.

Both sides of the debate seem to be able to find supporting evidence.

Take the example of DeepSeek V4. Proponents of export controls point to the rounds of delays in its launch, the fact that V4 hasn’t shocked and awed in the same way that DeepSeek R1 once did, and the fact that its gap with leading US models hasn’t narrowed as major evidence that the US containment strategy is working.

But looking more closely, the real significance of V4 is that the galapagosization is expediting faster than most people expected, and that export control will ultimately be a losing bet.

Here I’d like to quote from this excellent article from discussing how DeepSeek V4 represented “the first model custom-built for non-NVIDIA chips”.

Read more

Let me also use this as a small note for new subscribers about the difference between this newsletter and Baiguan. Baiguan is data-focused and is much more business and investment-focused. It is intended for anyone who has a real stake in China. It’s a product of my own company. We do data and information services for institutional investors and corporates.

This newsletter, China Translated, is my personal newsletter, where I try to explain the social, political, cultural, and historical contexts to help you make sense of what is happening in China.

Now, with my obligation to you fulfilled, today I will write about the topic I actually wanted to write about the most among the 3 topics in that poll: an interesting encounter at the (destroyed) tomb of Dai Li, often seen as Chiang Kai-shek’s Himmler/Beria/Hoover, which is a nice little showcase of how grassroots governance plays out in today’s China.

Nanjing, the old capital

Early subscribers to this newsletter know that I was born and raised in the city of Nanjing, China’s former capital before Beijing took that role.

When Chiang Kai-shek assumed power of the Kuomintang in 1927, and by extension the Republic of China (ROC) government, he decided to place Nanjing (or Nanking) as the capital. That was until 1949, when the communists took over, and Chiang fled to Taiwan.

Given this historical context, it is no surprise that Nanjing is full of important KMT-era sites. For instance, Sun Yat-sen, the founding father of the ROC, was buried in the mausoleum in the Zijin Mountain not far away from Nanjing’s city center. Chiang’s “White House”, the presidential palace, was turned into a popular museum.

This past Ching Ming festival, almost two months ago, several friends and I visited another KMT-era site we had never been to before: the Ling Gu Temple 灵谷寺. Although it is a temple dating back to the 6th century, its surroundings are closely intertwined with the history of the Kuomintang.

Take, for example, the Wu Liang Hall 无梁殿, or the Beamless Hall. The structure was built entirely from the bottom up with bricks, without a piece of wood or a single nail, and thus “beamless”. It used to be a Buddhist temple, but in 1928, it was turned into a memorial that eventually enshrined more than 30,000 lost souls in the Kuomintang army, with all their names carved on the walls.

They were the Kuomintang’s casualties during the Northern Expedition against the warlords, the 1932 Battle of Shanghai against the Japanese, and the 1933 Defense of the Great Wall, also against the Japanese. (But for the much bloodier War of Resistance Against Japan that started in 1937 and ended in 1945, claiming the lives of more than 3 million soldiers and at least 10 million civilians, the Kuomintang obviously didn’t have enough wall space, nor enough time, to carve down names.)

In the lawn outside the Beamless Hall, the remains of 316 KMT officers and soldiers, representing all the fallen, were buried.

So, this whole place, essentially, is ROC’s equivalent of Arlington National Cemetery.

Besides the Beamless Hall, there are also mausoleums of several KMT-era heavyweights that I won’t bore you with.

But to be very honest with you, the single place that got me most interested in going to Ling Gu Temple was the destroyed tomb of one of the most consequential, most controversial, and most mysterious individuals of that era: Lieutenant General Dai Li, more commonly known by his subordinates as “Boss Dai”.

Who is Boss Dai

Born as Dai Chunfeng 戴春风 with the courtesy name Yunong 雨农 in 1897 in rural Zhejiang province, he spent his early years drifting before eventually enrolling in the Whampoa Military Academy. It was there that he caught the eye of the academy’s commandant: Chiang Kai-shek. Recognizing Dai’s fierce intellect and absolute devotion, Chiang brought him into his inner circle.

Dai soon changed his name to Dai Li 戴笠, a name translating to “wearing a bamboo hat,” symbolizing an anonymous servant or an assassin in the shadows. True to this new moniker, he was tasked by Chiang with a monumental undertaking: building China’s first modern secret police system from the ground up. This organization, the Bureau of Investigation and Statistics of the Military Commission, or Juntong, would grow into a formidable, invisible empire of intelligence, espionage, and assassination.

When analyzing Dai Li, it is tempting to compare him to the other notorious secret police chiefs of his era: Nazi Germany’s Reinhard Heydrich, the Soviet Union’s Lavrentiy Beria, and America’s Hoover. (Barbara Tuchman once called Dai “China’s combination of Himmler and J. Edgar Hoover”.)

Like those figures, Dai Li was meticulous, ruthless, and highly effective. But he was a distinctly different kind of operator. He did not operate on the genocidal, industrial-killing scale of Heydrich, nor did he possess the perverse sadism of Beria. And while Hoover spent his career hoarding secrets to make himself untouchable, effectively holding US presidents hostage to his power, Dai Li did the exact opposite. He pledged his life, his agency, and his complete devotion to one man and one man only: Chiang Kai-shek. He was deadly loyal, acting as the sword and shield of the Nationalist leader without ever seeking the throne for himself.

Boss Dai’s story is one of history’s great “what ifs.” By the end of World War II, his intelligence network was vast and feared. He could have made an even higher mark in history, and Chiang himself allegedly lamented that he would not have lost mainland China and fled to Taiwan had Dai Li’s life not been cut short unexpectedly.

On March 17, 1946, Dai died (no pun intended) when his plane crashed into a hillside called Dai Mountain (different character, same pronunciation) near Nanjing during a violent thunderstorm. Given his line of work and the number of enemies he had amassed, rumors of assassination immediately swirled and persist to this day.

His funeral was a grand, state-level affair, drawing massive crowds and the highest ranks of the Nationalist government. Chiang, a man rarely known for public displays of emotion, was reportedly deeply devastated by the loss of his most trusted weapon.

Because Boss Dai had hunted and executed thousands of Communist agents, Chiang Kai-shek knew that if the Communists ever took power, the remains of his spymaster and close confidant would be a prime target for vengeance. Consequently, Chiang ordered Dai Li’s burial site, located just outside Nanjing near Ling Gu Temple, to be built like an impenetrable bunker. The coffin was encased in exceptionally thick reinforced concrete laced with coal slag, designed to make grave-robbing or desecration nearly impossible.

Chiang’s foresight was correct, though his concrete plan ultimately failed. In 1949, the Communists won the civil war, and the Nationalists fled to Taiwan. Not long after, in 1951, the new People’s Republic of China authorities ordered the destruction of Dai Li’s tomb.

The reinforced concrete proved so incredibly tough that local laborers spent days fruitlessly hacking at it with pickaxes. Eventually, they had to resort to using heavy explosives to finally crack the tomb open. Once breached, the coffin was smashed, and Dai Li’s remains, already charred from the 1946 plane crash, were dumped out, abandoned, and the grave was leveled to erase his memory from the landscape.

Locating the site

Given the rich story about this colorful figure, when some friends suggested we visit Ling Gu Temple this time, and that the tomb of Dai Li was somewhere nearby, I shouted, “Hell yeah, let’s go!”

It was not an easy place to find. There were no public signposts. Nor was it marked on the map app. Obviously, this is a place that has been intentionally obscured from public memory.

However, we did find occasional travel notes about it on Rednote (Xiaohongshu), and we relied on them to locate the site. The existence of those notes on the public internet is proof that this site, although left to fade, is not sensitive enough to warrant a total erasure.

According to the notes on Rednote, we first need to find a public cemetery for some KMT soldiers, then walk a few hundred meters to the left until we see a circular stone structure lined with trees.

Around the circular stone structure, there was a paved way. We should keep walking along this paved path for a few hundred more meters until we reach a little pond.

The tomb is by that pond.

I was worried we still couldn’t find it despite these instructions. But my worries were quickly swept away when, as soon as we saw the little pond, a guy looking like a park guard emerged.

He was the only person we saw in the last 30 minutes since we went off the beaten track. He was not a policeman, just someone with a park service uniform. He must be there for a reason, like a marker for something.

He must be guarding Boss Dai's tomb!

The guard stood at the gate of a small courtyard, lined by bushes. We walked past the guard and got into the courtyard. There, we found another man, dressed in a casual jacket, sitting in a portable chair, playing on his phone. He didn’t look like a random tourist to us.

The guard in uniform watched us intently as we entered, but didn’t say a word until one of us took out the phone.

“No photos allowed here”. The guard, who looked to be in his early 20s, spoke to us.

“But can we visit?”

“Yes, but no photos.”

“There was nothing left here,” casually commented the mysterious sitting man, eyes barely moving away from his phone when he spoke.

When we walked deeper into the courtyard, the guard followed us. About 30 meters in, we saw a stone slab lying on the ground. The marked words on it were entirely unintelligible. But it was clear this was the tombstone of Boss Dai. The words on it seemed to be intentionally scrubbed out.

“No photos”, urged the guard again, afraid that we would sneak up a photo of this unspectacular piece of rock.

There was indeed nothing else to see, and so we went away.

The second attempt

The whole experience baffled me greatly.

The presence of guard(s?) was totally unexpected. Dai Li passed away more than 80 years ago. Why is he even sensitive now? I doubt many young people would even know him. And I doubt even more that there are enough people who want to politicize him. Even Chiang Kai-shek has been somewhat depoliticized and is no longer the supervillain that Mao-era propagandists once portrayed him to be.

If even Boss Dai’s own boss was not as hated as before, why would they care about little Boss Dai? Why is this extra security even needed here? Has this arrangement always been there? It will be quite amazing if the PRC has been guarding the destroyed tomb of its former archenemy through all those years.

That evening, at my partner’s home, we told her parents the story. They suggested that the security presence we saw was likely not there all the time, but might be because that day was the Ching Ming Festival, also known as the Tomb-sweeping Festival, which is traditionally when people pay respects to the deceased.

This makes more sense, which also reminded me of another fact: Dai Li was killed by that fateful plane crash on March 17, 1946. We were almost exactly on the 80th anniversary of that event, adding an extra layer of sensitivity.

If this is the case, we reasoned, the security might be reduced later, and we should be able to take a photo.

So the next day, we visited again.

This time, since we already knew the way, we went straight to the target. Ling Gu Temple. The Beamless Hall. KMT public cemetery. Circular stone structure. Paved road. We stormed past each landmark in quick succession.

Very soon, we hit the little pond. But our hearts sank when we saw that young park guard again, this time standing on the bank of the little pond, in pensive thought.

Maybe we should have a chat with him.

“Hey, man, good job. Are you here all the time?”

“Nah, just these few days.”

“Is it because of the Ching Ming Festival, and also because of the 80th anniversary?”

“Yes, but also because, don’t you know, the chairlady of KMT is coming?”

Oh absolutely! How did we forget about that? Cheng Lin-wun, the new chairperson of KMT, was coming for a historic visit. In fact, the day of our second visit was on the 6th of April, while Cheng would arrive exactly the next day, on the 7th, and her very first stop would be paying homage to Sun Yat-sen’s Mausoleum nearby.

Obviously, in days like these, Chinese authorities don’t want anyone to use sensitive sites like Boss Dai’s destroyed tomb to make a political message. After all, Boss Dai served Chiang, who was perhaps the most consequential predecessor of Cheng as the head of the KMT. Theoretically, there is a possibility, however remote, that people with an agenda could use a photo of this place to try to derail the show.

We then notice that the guy in the casual jacket who was sitting there the previous day was not there. The young park guard was there all alone.

“That guy yesterday, he must be painclothes?” We asked.

“Yes. He came from the sub-branch. He is not coming today.”

By “sub-branch”, we assumed he meant the sub-branch of the local police office. But we didn’t ask further.

The sub-branch guy’s absence suggested this was sort of a gradient system. The most sensitive day is dealt with the heaviest arrangement, which would fade away as time goes by. Very soon, that lonely young guard would leave this post too, leaving tourists to do literally whatever they want with this site.

Overall, we were just amazed by the granularity and pragmatism of China’s grassroots management in this episode. They were not stupid enough to waste resources on putting a 24/7 guard at a supposedly sensitive site that very few people care about. But on special days, someone higher-up, sitting in an office somewhere, would still remember to send his/her subordinates a special assignment, if just to kill a one-in-a-million chance of some big trouble.

Whatever you want to say about the Chinese government, they are usually not sloppy.

The Sword

And we did take the photo!

While I chatted with the guard, my partner went in and snuck up a photo of the fallen slab.

You see, nothing intelligible. But we knew from historical records that it said “戴雨农将军之墓 The Tomb of General Dai Yunong”, as shown in this old photo I found on the internet:

And there is one last thing I want to mention here. That pond.

The pond is dirty, blanketed by algae. But despite its humble appearance, it’s also, in some sense, a historical site: After Boss Dai’s tomb was trashed, his charred remains were thrown into this very pond.

Also thrown in there was the destroyed “nine-dragon” sword that once belonged to the Qianlong Emperor of the Qing Dynasty. This is an amazing story tied to the Eastern Mausoleum of the Qing, a site I visited last year, but I never finished Part 2 of that travelogue due to what seemed like a lack of interest on your part.

You see, back in 1928, after the local warlord Sun Dianying blew up and looted several high-profile mausoleums, including those that buried the Empress Dowager Cixi and the Qianlong Emperor, Sun presented this priceless weapon to Dai Li as a bribe intended for Chiang Kai-shek. However, amid the chaos of World War II, Dai Li temporarily entrusted the relic to a subordinate, Ma Hansan, head of the Juntong system in Northern China. Treacherous and greedy, Ma secretly kept the sword for himself, eventually using it to buy his own life after being captured by Japanese forces by passing the blade to the infamous Manchu spy Kawashima Yoshiko. When the war ended, Dai Li interrogated Yoshiko and recovered that sword.

Then, he boarded that fateful plane to see Chiang, carrying the sword with him.

A young park guard, all alone, brooding over the dirty little pond that stored whatever remained of Dai Li’s body, and of Qianlong Emperor’s 9-dragon sword, as if he is the last tomb keeper of Chiang Kai-shek’s sword and shield, as if the once-feared creator of China’s first secret police system would one day be “protected” and watched over by his former enemies, is a scene that will never be erased from my memory.

Subscribe now

Related readings

To make sense of this whole saga, we need to first run down the key facts and timelines, and then I will proceed to answer some of the top questions.

Timeline of the Meta-Manus Acquisition

• 2022: The startup, originally operating as Butterfly Effect, is founded, with research and development teams based in Beijing and Wuhan.

• March 2025: Manus debuted its general-purpose AI agent, capable of executing complex, multi-step tasks such as resume screening and stock analysis. The release gained massive viral attention and is compared to “DeepSeek” as a sign of China’s rapidly closing AI capability gap.

• April 2025: Benchmark Capital led a $75 million funding round in the startup, valuing the company at $500 million.

• May 2025: The U.S. Treasury Department launched an official inquiry into Benchmark’s investment. Regulators investigated whether the funding violates outbound investment security rules, citing concerns that the capital could accelerate sensitive AI development in a “country of concern.”

• Mid-2025: To mitigate geopolitical and regulatory pressure, the company relocated its headquarters to Singapore. The founders laid off remaining staff in China, shut down local social media accounts, and rebranded to position themselves as a Singaporean enterprise, a move industry observers dubbed "Singapore-washing."

• December 2025: Meta announced its acquisition of Manus for more than $2 billion, aiming to integrate the company’s autonomous agent technology.

• January 2026: Chinese regulators launch a formal investigation into the deal. Authorities focus on potential violations of national laws regarding technology exports, data transfers, and foreign direct investment.

• March 2026: The regulatory situation intensified. Manus co-founders were summoned to a meeting in Beijing with officials from the NDRC. They were reportedly barred from leaving the country while the investigation proceeds.

• April 27, 2026: The NDRC formally blocked the acquisition.

What’s special about the Manus deal?

Under Chinese law, the acquisition of Chinese companies by foreign investors is subject to government approvals, a process similar to CFIUS in the US. If Manus were a Chinese company and were acquired by Meta (assuming CFIUS approval), the deal would need to go through the approval process in China, as with any cross-border M&A, and there wouldn’t be anything dramatic about it.

What made the Manus case unique was that it was originally a Chinese firm, but just as the product went viral, they dissolved the Chinese entity and re-established themselves as a new company in Singapore.

While they kept the same product, technology, and core team, this “Singapore-washing” move allowed them to bypass the approvals (or so they believed) that would otherwise have been required when Meta acquired them.

Fast forward to today, Beijing is basically saying that substance matters more than appearance, and that Manus cannot bypass the relevant approvals. Because they completed this deal without the necessary authorization, it must now be unwound.

It is important to note that there are no penalties or fines yet, nor is there any criminal prosecution.

So far, it is strictly business.

Why did Manus conduct “Singapore-washing” in the first place?

Manus served a mostly international market, received USD funding, and was built on US foundational models such as Claude. The US inquiry into Benchmark Capital’s investment in Manus was perhaps one of the many triggers that persuaded the team to change its identity.

They are caught between the two worlds.

How would the unwinding actually take place?

Conceptually, it’s not as hard as it sounds. If all of the 3 key parties: founders, investors, and Meta, agreed to an unwind, it can be done. The details of how to “decontaminate” Manus technology away from Meta are secondary matters.

I heard that at least some investors have already agreed to return the cash. It’s also entirely possible that Mark Zuckerberg already had buyer’s remorse since the consensus was that he was in a bit of FOMO when pulling the trigger back then. And the founders? Very likely, they were the first to fold.

Does Manus control any core technology that warrants such an unprecedented reaction from Beijing?

Manus belongs to the class of companies that are often referred to as “AI wrappers”. They do not train their own models, but through smart design and engineering decisions, they leverage model capabilities to deliver a smooth user experience.

I have been a happy Manus user for almost a year. I find that, for someone like me with no engineering background, I prefer Manus to Openclaw as my go-to AI agent for tasks like creating a simple app or conducting deep-dive research. (However, had I had access to Claude, I probably would have considered switching over.)

Overall, Manus is not DeepSeek, or Huawei, and I do not think Manus controls any core technology that Beijing would care much about.

So why is Beijing blocking the deal? Is Beijing not afraid of hurting overseas investors' sentiments and entrepreneurial spirits again?

Although Manus does not possess any core technology of national security importance, Beijing is afraid that such “washing”, if unchecked, would serve as a terrible example for others to follow, rendering China’s foreign investment regulation toothless.

A key factor that explains Beijing’s action that outside observers often missed was that Meta’s $2 billion deal was very openly touted by Manus’s founders and investors as a dizzying success story.

It was very to-the-face.

Had the Manus founders not been so overexcited about the deal and never made the $2 billion price tag public, I doubt this would ever have happened. But since it is now well known that a previously Chinese start-up achieved such huge success immediately after changing its “cover” to become a foreign company, Beijing’s hands are also tied. Allowing this to happen without any brakes will send a wrong signal to future tech entrepreneurs.

So, why were the founders and investors of Manus so giddy about the Meta deal? And also, why did the Manus founders reportedly return to China, given the possibility they would be barred from leaving?

I think the main sin of the Manus team is their naivety.

When they did Singapore-washing, they genuinely believed what they were doing was good and necessary. And when they were acquired so quickly at such a high price, they genuinely believed it was something to be celebrated. And when Beijing raised eyebrows, they genuinely believed they could explain the trouble away and chose to come back to the country and straighten things out.

Had they been conscious of the consequences, they would “Singapore-wash” or “Silicon Valley-wash” themselves from day one, as many of their peers had long ago, and they would not be where they are today.

They live in the illusion that this world is still the same globalized world when Mark Zuckerberg founded his empire.

Will this landmark episode force Chinese founders to leave and found companies overseas from Day 1?

I believe this has already happened.

My sense was that most founders who wanted to do this “Day 1 washing” had already done it. They just weren’t as outspoken as Manus about it. The remaining ones are perfectly happy with being a Chinese company. So net-net, the damage from this case to China would be limited and would be outweighed by the benefits.

The problem with Manus is that they only half-did it and thought it was not a problem. It’s the combination of this half-heartedness and the naive over-excitement I mentioned above that made Manus into a punching bag.

Most companies have already picked a side. Manus was just too late in making up their minds.

Subscribe now

So I decided to let you have the vote.

Whichever theme gets the most votes in 7 days, I will write it out by the end of next week.

• During the Ching Ming Festival two weeks ago, I visited the (destroyed) tomb of Dai Li, Chiang Kai-shek’s Himmler/Beria/Hoover in Nanjing, and had a really bizarre encounter that could showcase very well how grassroots governance played out in today’s China.

• I realize what the whole “lying flat” culture prevalent in China’s younger generation is all about.

• There used to be a clear advantage of Western governance models over China’s. But information technology and AI make that advantage less certain.

Now, please cast your vote. Again, my promise is that I will write about the topic you, the reader, choose by the end of next week.

Naturally, people start to ask: if a military invasion of Taiwan is what China has long planned for, isn’t this the perfect timing for it? Why would China waste this golden opportunity?

In this essay, I will explain to you why this won’t happen.

I understand Taiwan is an emotionally charged topic for many people. Some of you will not like my conclusions. So first, I want to clarify my role here.

I am not here to pass normative judgments about which side is right or wrong. My role is that of an analyst. I see many people, some of whom are my clients and many of whom have stakes on the table, are concerned about the situation in Taiwan right now. They deserve clarity on this issue and a level-headed analysis with some predictive value. That’s what I try to provide here.

As early as 3 weeks ago, I predicted the war in Iran was not going to end soon. And now I want to share with you this other high-conviction assessment of mine.

Recap of Part 1

If you have already read Part 1, you can skip this part. If you haven’t, here is a quick recap.

In Part 1, I laid out two fundamentally different perceptions of Beijing's core policy formula on Taiwan, namely, “We strive for peaceful reunification, but we do not promise to give up use of force.”

From an American (and, to a large extent, Western) perspective, the main focus of this formula is the “use of force”. “Reunification” ranks second, but China’s expansion won’t limit itself there. It’s just the first step of a global expansionary ambition. “Peace” is just a cover-up phrase.

From China’s perspective, though, the order of emphasis is entirely different. The first and foremost thing is “reunification”, a daily obsession of the People’s Republic of China since its founding. (It was also the daily obsession of Taiwan before Chiang Kai-shek passed away.) After it, “peace” ranks second. A peaceful reunification is much preferred to war. I concluded that the US far underestimates China’s obsession with reunification, as well as its desire for a peaceful reunification.

To set the broader context, I took some time to explain why there was this obsession. My explanation is from a pragmatic perspective. The core argument is that Taiwan is at the heart of the PRC’s founding myth, that without it the country will fall apart, and that the Party itself will lose its mandate to rule. Therefore, the whole Taiwan Question is existential for China, not just a piece of land that Beijing can bargain away for something else.

Whatever you believe about which side is right in this great conundrum, this “existentiality” of Taiwan from the perspective of Beijing sits there like a rock. Any serious analyst of the situation can’t look away from it.

Again, please go to Part 1 for a more fleshed-out argument.

Peaceful reunification: the pull factors

So now we understand that Taiwan is existential for China, shouldn’t China jump on this great opportunity to attack Taiwan and finish the deed once and for all?

No.

If there is only one key conclusion you could take away from this essay, it’s this:

China, ultimately, needs a peaceful reunification.

Peaceful reunification is both preferred and increasingly likely.

Time is on China’s side.

There are at least three reasons supporting this conclusion: there is an economic one, a cultural and political one, and a final part about Xi’s own vision for his legacy.

The economic argument is obvious. A seaborne invasion of a big, mountainous island armed like a porcupine is extremely difficult. If China succeeds quickly, international sanctions and disruptions to external trade links will cause significant pain to China’s economy. If China can’t deliver a quick win, it will also become an economic disaster that could potentially dial back all the progress we have made since the reform and opening-up era. covered this topic quite comprehensively in their essay Why China Won’t Invade Taiwan.

The cultural reason is more squishy, and it’s difficult to explain it well to people who have never had real experience dealing with the Chinese. But I will try.

You may well remember that in my rebuttal to Noah Smith, I laid out my key observation: China is not an “insecure-expansionist” power like Nazi Germany, Imperial Japan, or the Soviet Union, so war and aggression have no popularity here as solutions to big problems. I also explored why history and geography have shaped China into a generally peace-loving culture.

A few weeks ago, a TV drama called Swords into Plowshares (太平年, literally “The Peaceful Years”) became an unlikely success in China. The show covered a historical period in China that even most Chinese people knew little about: The Five Dynasties and Ten Kingdoms (907–979 AD), a chaotic 7 decades between the fall of the Tang Dynasty and the rise of the Song Dynasty.

It was a dark period marked by great political turmoil, murders, famine, and even cannibalism1, when might was right, when emperors died like flies, and when dynasties rose and fell every few years. It was also a period when rules and order broke down completely, and whoever controlled the armies claimed the throne.

In short, it was about a historical period that stands in total contrast to everything modern-day China represents.

The show began in the middle of this historical period and ended with its final chapter: a famous moment when the Kingdom of Wu Yue, which mostly occupied present-day Zhejiang, Shanghai, and southern Jiangsu, surrendered its territories to the Song Dynasty, and the country was finally reunified, peacefully.

To be sure, there is quite a big dose of pseudo-history in this show, too. The show made it seem as if surrendering to central authorities for the sake of national unity had always been the King of Wu Yue's eventual goal. The actual historical truth, however, involved far more coercion than was portrayed.

But still, the overall contours of history were respected, and it was true that no lives were lost when the Song Emperor put today’s Zhejiang and Shanghai under his fold.

The political analogy in this show is subtle but hard to escape for most Chinese viewers. One especially interesting scene was towards the end of the show, when the Song emperor discussed the strategy with his generals.

He asked General A: “How many men do you think you need to conquer Wu Yue?”

General A: “400,000 men.” (This is a huge number in that era.)

He then asked General B, who is more junior to General A: “What do you think?”

General B: “I need 300,000 men.”

Emperor: “Oh, so General B is somehow more capable than General A?”

General A: “300,000 men is possible, but it will be a war that will leave the land of Wu Yue in total ruins.”

The emperor fell silent with visible pains in his face.

You see, what’s interesting about a Chinese TV drama that you can’t find in American dramas is that this one not only has no problem of using an epic peace, not an epic battle, as a way to wrap up a big show, but also takes the pains to show you the deliberations about the lengths those characters go to in order to avoid violence.

This show is just one of the many soft reflections of China's national character. To add more hard data to this, I want to cite a very good report from the China Pulse survey by Emory University, the Carter Center, and : (Emphasis mine)

In the July-August 2025 wave of this survey, a small majority (51%) of the Chinese public agreed Taiwan should not be unified by military force under any circumstances (29% disagreed). In the October-January wave, there was no majority opinion (38% agree vs. 45% disagree).

Should Taiwan declare de jure independence from the mainland, the Chinese public supports limited military operations on offshore islands (81%) and economic sanctions (62%) as a response over a full-scale military attack (32%).

Twice as many Chinese people agree the use of military force will only make the Taiwan issue worse (50%) than disagree (24%).

These results are similar to my own reading of the situation. An unprovoked war over the Strait of Taiwan will be deeply unpopular.

Ironically, while an American president can start an unpopular war as he likes, China’s top leader needs to read the room much more carefully before launching one. This is because China’s system is delicate, even brittle, and the deep internal imbalances caused by external military adventures have led the country to implode again and again.

Peaceful reunification: the push factor

While these economic, cultural, and political factors serve as the great “pull effect”, they can still be unsatisfactory for skeptics. After all, many of you still hold the view that it’s all up to Xi Jinping himself.

If Xi Dada is determined to take down Taiwan one way or another, for a man who has sufficiently proven his resolve and willpower to the world, he will find ways to bypass all these restraints. And if that’s true, the best moment to take action is indeed right now. With just one signature from him, the People’s Liberation Army, an organization largely insulated from the rest of the population, will jump on this opportunity to sail across the narrow strait in one swoop.

Except, there is also a strong push effect here, a big incentive for him not to plunge into a war.

It’s about what he sees as his own legacy.

I think by now, very few people disagree that Xi is someone with a keen sense of mission, whether you like his mission or not. So what’s his mission? I can’t guarantee I know everything about this, because I can only gather information from public sources about his past words and deeds.

What I am very certain of is that his mission is deeply intertwined with this country's cultural impulses. In other words, it’s impossible for him to set his life’s mission for the Chinese nation as something entirely alien to what the Chinese nation considers to be right. (This statement is actually tautologically correct.)

If China is a country that has chosen to disdain violence, how would a Chinese leader want to be remembered? It will not be hard to conclude that their loftiest dream is to be remembered as someone who oversees unity, prosperity, and peace.

Here, the cultural factor is no longer just a pull factor, but also a push factor, hard-wired into the decision-making algorithms of Chinese leaders.

In such a context, a peaceful reunification carries far more prestige than a military invasion can ever achieve, and will go down in China’s 3,000-year history as one of the greatest feats of all time.

Honestly, which Chinese leader can refuse this enticing possibility?

To put it in some pseudo-quantitative way, imagine there are 3 outcomes for Xi. One is a peaceful reunification, one is a reunification through bloodshed, and the last one is no reunification.

A “no reunification” is 0. A peaceful reunification is a 100. A reunification, albeit through war, is probably only at the 20-30 mark. It’s better than “no reunification”, but it’s a huge downgrade from the other option. It’s not an easy choice to make, and will always be the painful last-resort choice

So why is PLA preparing for combat?

If the argument is that China prefers peace, how do we explain the frantic pace of its military modernization? From the rapid commissioning of naval vessels to the expansion of its nuclear arsenal and Xi’s frequent, televised orders for the PLA to be “ready to fight and win”, the optics suggest a nation on a war footing.

Analysts often point to the “Davidson Window”, the 2021 assessment by US Admiral Phil Davidson that China might attempt a military takeover of Taiwan by 2027, as a definitive invasion deadline.

No, I am only playing with you here. Davidson actually only said that China would be combat-ready by 2027, but many people already jumped to the conclusion that it’s a countdown to war. (The Wall Street Journal even had a whole report about this misconception.)

There is a massive strategic gulf between having the capability to win a war and having the intent to start one. Sun Tzu’s The Art of War famously argues that “the subjection of the enemy’s army without fighting is the pinnacle of excellence.” To achieve a peaceful reunification, Beijing must first eliminate the “military option” for its adversaries.

A deadline for combat readiness is thus a prerequisite for peace. It is only when the PLA can demonstrate a 100% certainty of victory, and a corresponding 100% certainty of defeat for any intervening force, that war can be avoided. By building a military that makes the cost of resistance unthinkable, Beijing believes it can finally bring the modern-day “Kingdom of Wu Yue” to the table without firing a single shot.

In fact, if war were the final goal, then hiding strength and biding one's time rather than showcasing daily military progress would be the better strategy for winning. After all, subterfuge is the key to military success. Eisenhower never told Hitler his forces would land in Normandy. Putin lied for many months before the war that he would not attack Ukraine.

Does Xi have to finish the job within his lifetime?

It is true that early in his tenure, Xi Jinping famously stated that the Taiwan question “cannot be passed on from generation to generation.” Many observers have latched onto this phrase as a ticking clock, interpreting it as a personal vow that he must achieve reunification before he leaves office or passes away.

However, this is a narrow and likely incorrect reading of his intent. While Xi certainly wants to move toward a solution, there is a massive difference between “not waiting forever” and “forcing a conclusion at any cost.”

If the price of reunification is the wreckage of the Chinese economy and the derailment of China’s “Great Rejuvenation”, I highly doubt Xi would bang his head against that wall to achieve it. As former top CIA China analyst Chris Johnson argued in his essay, the idea that Xi is driven by a desperate “legacy obsession” to seize Taiwan regardless of the risks lacks firm evidence. For a leader who views himself as the steward of China’s return to global preeminence, a reckless gamble that leaves the nation wide open to a high risk of self-implosion is the ultimate bitter pill I doubt he will ever want to swallow.

Interestingly, my own high-conviction assessment of this timing also closely aligns with the recent consensus within the U.S. intelligence community. In the 2026 Annual Threat Assessment, the US ODNI concluded that Beijing is not currently planning a military invasion. While I often find myself at odds with Washington’s broader narratives, on this specific point of tactical restraint, the analysts at Langley and I are reading the same room.

Xi is a man of resolve, but he is also a pragmatic calculator of risk. To him, “not passing the problem down” means building the overwhelming strength that makes reunification inevitable, not launching a desperate crusade to meet an imaginary deadline.

How and when?

To understand how this ends, we must first accept an uncomfortable, fundamental truth: the “Taiwan Question” has never been a bilateral issue between Beijing and Taipei. It is, and always has been, a function of the U.S.-China relationship.

Historically, the US has been the primary variable in this equation. In the 1950s, when the US all but gave up on Chiang Kai-shek, the Korean War broke out, leading Truman to save the KMT’s Taiwan from certain defeat. In the 1990s, U.S. carrier groups again intervened to maintain the status quo.

But the reverse logic is also simple: if it becomes clear that the United States is no longer willing or able to defend the island, the question will be solved almost instantly.

Will Washington defend Taiwan? There are both push and pull factors for the US, too.

The current disaster in Iran serves as a sobering data point for Washington’s “pull factors.” If the U.S. military is struggling to secure the Strait of Hormuz against a regional power like Iran, the prospect of defending Taiwan against a China that is a hundred times more powerful than the IRGC becomes strategically daunting.

However, there is also a significant “push factor” keeping the U.S. tethered to Taiwan: semiconductors. In a recent poll of my American subscribers, 68% cited TSMC and chip supply as their No. 1 concern regarding Taiwan, accounting for more than half of their “Taiwan mind share”. This is understandable. For NASDAQ and the broader U.S. economy, Taiwan is essentially a single point of failure.

Interestingly, this creates a profound asymmetry of needs. If I were to poll the Chinese public, I suspect chips would occupy less than 5% of their own “Taiwan mind share”. As I explained in Part 1, to Beijing, Taiwan is an existential matter of national sovereignty; to Washington, however, it is mostly a matter of supply chain security.

This asymmetry is exactly where the opening for peace lies.

In any negotiation, when one party wants “A” (sovereignty) and the other wants “B” (uninterrupted silicon), there is room for a bargain. The critical deadline for peaceful reunification, therefore, is not a military one, but an industrial one: the maturity of domestic U.S. semiconductor foundries.

Once the CHIPS Act projects reach full capacity and the U.S. feels it has achieved semiconductor independence, the strategic necessity of defending Taiwan evaporates. At that point, for the US, the liability of a potential war will outweigh the benefit of the island’s autonomy, and we will see the gears of negotiation begin to turn.

This also explains why Beijing’s reaction to Japanese Prime Minister Takaichi’s recent comments was so fierce. Some called it irrational, but Beijing was actually dead rational. By suggesting that a Taiwan contingency is an “existential crisis” for Japan, Takaichi threatens to transform a manageable U.S.-China negotiation into a much more volatile China-Japan confrontation. For Beijing, involving Japan, a country toward which the Chinese public holds deep historical grievances (unlike the attitude towards the US), raises the stakes from a cold strategic bargain to a hot emotional conflict. It puts the “30-score” scenario of a destructive war back on the table.

Make no mistake: war is not a zero-percent chance. Conflict will occur if China ever feels truly vulnerable, if Beijing concludes that all peaceful avenues are closed, and if a painful, “low-score” reunification is the only way to prevent the total collapse of its own founding myth.

But as long as the path to a high-prestige, peaceful settlement remains open, time remains on China’s side.

Subscribe now

[85% of this article is written with human hands, 15% are written by AI after I dictate the outline]

(The full version can be found here.)

The situation in Iran has dominated the news headlines for the past few weeks. Naturally, even though China is a distant party to this conflict (as Hegseth himself confirmed), many people still try to link it to China.

First, there is the claim that China stands to hurt the most from this war. I dealt with this quite easily in this tweet.

And then there is the outcry about why China has done nothing to defend Iran, supposedly an ally of China (No, they are not allies.)

At the time of writing this piece, Trump even threatened to cancel the upcoming Xi-Trump meeting if China doesn’t offer help.

China, China, China! But isn’t this war mainly about Iran, Israel, and the US?

The most egregious narrative, however, belongs to the idea that China would exploit this opportunity to take action against Taiwan.

If this is true, the likelihood of that can only grow more intensely after Iran effectively closes the Hormuz, and after it becomes increasingly likely that the US would be mired in an endless war in the Middle East again.

When the US is shipping THAAD away from Asia and sending marines from Okinawa to the Persian Gulf, and when there are now substantive questions regarding whether the US military can actually protect its own allies and sustain a long war against even a second-rate power, what better timing than now to resolve the Taiwan Question once and for all?

And yet China is doing nothing, and we are kept waiting.

At this point, I should point out that the idea that “because the US is busy elsewhere, China will take the chance to attack Taiwan” reveals the true lack of understanding about the entire matter.

I have long wanted to lay out my view about the Taiwan Question comprehensively, so I figure perhaps this is the right moment, especially given what the US has just done against Iran, a sovereign state at the opposite end of the world, is stacked against what Beijing is NOT doing against an island it has always claimed to be its own, but never moves a finger yet.

And not just the question in the title. I will also use my own mental framework set up in this essay to answer questions, including:

• Why is Beijing so angry about Takaichi’s comment about Taiwan?

• Is the Davidson window real? In other words, is there really a 2027 deadline for a military invasion?

A tale of two mental frameworks

I will start with a review of Beijing’s official stance.

Beijing’s perspective on Taiwan has been quite consistent. It considers Taiwan to be a runaway province, a remnant of the last civil war, and a remaining relic of past humiliations. It has always treated the Taiwan Question as a domestic issue and set recognition of the so-called “One China Principle” as the prerequisite for commencing any diplomatic relationship.

Concurrently, Beijing has consistently and openly treated reunification as its ultimate strategic goal. Its exact formula to achieve that can be best summarized by this oft-cited statement:

我们争取和平统一，但是不承诺放弃使用武力

We strive for peaceful reunification, but we do not promise to give up use of force.

Here is what I find interesting about this statement: there are 3 key phrases: peace, reunification, and the use of force. But at some point, I realize that these 3 phraes carry entirely different weights to different people’s ears.

This is how a typical Chinese mind will rank the 3 by level of importance:

1) Reunification > 2) Peace > 3) Use of force

Now this is how a typical American mind will rank the 3:

3) Use of force > 1) Reunification > 2) Peace

You see, they are ordered entirely differently!

For a typical American mind, it assumes that sooner or later, China is going to use force against Taiwan. And not just Taiwan. It’s only Stage 1. After Taiwan, China will enter a stage of global military expansion. Peace? It’s the least important. Might well be a cover-up phrase for Beijing’s true intentions!

This is, at least, what Washington would do if it sees something as its strategic objective. If sending some missiles and drones to kill somebody can achieve some goal, fine, they will do it.

This is not what an average Chinese mind thinks, however. For the average Chinese mind, the thinking goes as something like this: the first and foremost issue is the reunification itself. But we really want to achieve it with minimal bloodshed. We prefer peace much, much more than war. Yet, we can’t give up on the use of force as our last resort. If we give up on the “stick”, our “carrots” alone will never achieve our top priority: the reunification itself.

These two entirely different modes of thinking are at the core of the misunderstandings about the Taiwan Question. My observation is that the US, overall, far underestimates China’s obsession with reunification, underestimates China’s willingness to reunify peacefully, and overestimates China’s preference for fighting wars, and not least because the US often tries to use its own mirror to understand China’s motives.

If you have completely followed me by now, I congratulate you. You don’t have to read the rest of this essay.

If you still find this to be unconvincing, though, now is the time for a deeper dive.

Why the obsession with reunification?

The first thing we need to know is that China is absolutely obsessed with making sure Taiwan is part of China. No matter whether you agree with China on this or not, this very obsession itself is a given. No amount of strategic calculation can ignore the existence of this obsession.

But why so obsessed?

There can be many reasons. I can imagine Beijing citing the history of past humiliations, memories of the old civil war, and shared language and heritage. You might also already be tired of the phrase “自古以来 from ancient times”.

But none of them are that relevant for you, the reader, the armchair geo-strategist ruminating on the future of the world.

There is, however, a single reason for the obsession: Taiwan is already at the core of the founding myth of the People’s Republic of China. Without it, China will crumble.

What is China, anyway? Like any country in the world, it’s a shared imagination. It’s a story that people keep telling each other, and all of these rounds of storytelling make a “country” a reality, and make people of the same country willing to live under the same roof and to follow the same rules.

For the US, this story is about the Mayflower, the Boston Tea Party, the Declaration of Independence, and the Statue of Liberty. Without those stories, the US has no reason to exist as a single body. Everyone might just pledge loyalty to Texas or California.

In contemporary China, the fundamental story is about the struggle to keep the country strong and prosperous, to protect its people from bullying by foreign powers and internal warlords, and to prevent disintegration into chaotic fragments once more.

“Not an inch of land lost” is the non-subtle promise the governing party has made to the governed since the founding of the PRC, the current iteration of China, as part of the same policy package aimed at keeping the country “real” and ensuring the governing party stays in power.

That “not an inch of land” is in relation to the map the Party has drawn for its citizens from the very beginning, and it’s true that so far it has made good this Great Promise. It has successfully and peacefully reclaimed Hong Kong and Macao and settled most of its border disputes with neighboring nations.

And now, Taiwan remains the last major obstacle to achieving this Great Promise.

What would happen if Taiwan were to become permanently and irrevocably independent, and Beijing did nothing about it? Beijing would have broken this Great Promise. And with that, it would lose the mandate to rule. Soon, its rule will collapse, and very soon the country will fall apart and descend into a new cycle of chaos.

This is as serious as it will get, because, once again, what’s more to a country other than the story it keeps telling itself? If you break that story, and make people realize it’s just a story, the people will question why they are in the same country after all. They will question whether everything is just a gigantic lie. They will question what they have signed up for in the very beginning. China, in its current form, won’t survive these questions and will only lead to significant human trauma before it settles into a new equilibrium in the indefinite future.

I guess you now arrive at the gist of this section: The Taiwan Question is not only about Taiwan itself. It is, fundamentally, also an issue about China. And it’s not just any single issue for China. It’s not a trivial piece of land that Beijing can just give up or be bargained away in exchange for something else. It’s an issue of life-and-death proportions. It’s about the survival of the Communist Party of China as well as the “China” in its current configuration.

This is why, when Xi got the opportunity to spell out China’s 4 red lines, the Taiwan Question sits at the very top of this top list. Xi is literally suggesting this is the most important issue for China, outranking anything else.

This is also why I often smile at the idea that the fate of Taiwan should be left to the choice of the 23 million people living on that island. For the record, I am a big fan of democratic choice, and I believe Taiwanese people should have a voice here. But I think, to be really fair, if we were to hold a referendum on it, we should also involve the 1.4 billion people living on the mainland. It’s not fair to exclude them from voting on an issue of life-and-death for themselves.

The cynics among you may argue: why only Taiwan? Why not Mongolia? Why not Vladivostok and hundreds of square kilometers of land lost to Russia just over a hundred years ago?

To be honest, I don't have an intellectually satisfying answer to these questions other than, “Well, they drew the map this way in the beginning. Now that the map is hardened into the national consciousness, they are only prepared to stake their political survival on this map”.

For the rest, they have already chosen to forget about them. They can be flexible when they have to, but where they can’t, they won’t.

Read the rest of this essay here:

China Translated

With US trapped in Iran, why is China wasting a golden opportunity to attack Taiwan? (Full version)

The US is getting bogged down in a disaster in Iran right now, and despite Trump’s many lies, the war doesn’t seem to end anytime soon…

Read more

5 months ago · 90 likes · 29 comments · Robert Wu

He was not a famous person, but he was remembered by friends and relatives for many intellectual legends.

Born in rural Zhejiang in the 1930s, he was soon found to be a child prodigy. One of his uncles sponsored his education in the county high school. Yet, unfortunately, he had to terminate his studies because his school was bombed in the war.

Not a quitter, going back to the same rural town that our ancestors had lived in for over a thousand years, he taught himself the whole curriculum and became a teacher. What’s more amazing is that he taught math, Chinese literature, and Chinese history to students from grade 1 all the way to high school seniors.

I never understood how he managed to do that. It’s as if he were already an AI-enhanced superhuman, except that ChatGPT happened a year after he left our world.

And as one of the extremely few members of his generation in rural China who knew how to use a computer, one of his daily routines in the last decade of his life was spending hours a day playing Sudoku online. It was said that his anonymous account was often ranked highly in some global leaderboard for that game.

We often wonder what kind of achievement grandpa could have made had he been born in a different era, and his studies were not interrupted by wars, and he went on to study in Hangzhou, in Beijing, or perhaps in the United States, just like folks like Yang Chen-Ning of his era.

Grandpa was not a talkative person. When he spoke, he spoke with a thick Shaoxing-style Wu accent, so that I could hardly understand him. Furthermore, living mostly with my parents in Nanjing, I didn’t spend much quality time with him as I grew up. All of these made him this legend-type of person in my memory, and sadly made him more of a mystery even for me.

One thing I do know about him is that he dedicated his twilight years to revising a book of genealogy of our clan. I knew from my parents that it was a huge endeavor. But it was only until the last few days, during the Spring Festival holiday, when I finally opened its pages that I had an understanding of its magnitude.

The book, as revised by a committee headed by my grandpa, spanned 38 generations all the way to the year 779, in the middle of the Tang Dynasty. This revision, the 15th revision in history, was nothing short of a miracle.

I won’t bore you with the actual family tree. But in this post, I have translated 4 interesting articles that I found in this treasure trove.

The first one is the preface written by my grandpa, detailing how the original Book was destroyed by the Cultural Revolution, grandpa’s personal journey about how the revision came about, and what many challenges they faced during the endeavor.

The second one is about how a pivotal 1947 edition was re-discovered, miraculously and literally through a dream, and the third one recounts how this same edition was salvaged from certain destruction during the Cultural Revolution. It’s all but certain that if they hadn’t found this 1947 edition, my grandpa would not have been able to finish this project before he passed away.

The last one was one of the more ancient accounts written and added to the book by ancestors. They read like a history book, written in classical Chinese. This particular one that I referenced here recounted how their little town suffered during the horrendous Taiping Rebellion, the Chinese civil war in the 1850s and 60s that claimed the lives of millions.

The motivation for this translation is two-fold.

First, I want to showcase to you this side of China, about what many of our older generation treasure at heart, a glimpse into their spiritual world.

Reading about how those lively characters - my grandpa who vowed that he would dedicate his life to this project, Mr. Chuanjia who kneeled before the long-lost 1947 edition and wept with joy - reminds me that, contrary to my own perception, many Chinese people are also faithful. It’s just that the exact nature of faith may be different from other versions.

The second motivation, of course, is to remember my own late grandpa. The private, silent, reclusive old genius who taught the child me math and Chinese history lessons. The eternal stranger that I would never be able to direct my questions to.

Yet, reading his own words, written in his life’s project for both his ancestors and for posterity, left me with an understanding of him that I never had before: a man of passion who finally found his calling.

I miss you, Grandpa.

And I hope you enjoy his work.

#1 Preface to the 15th Revision of the Tangpu Wu Clan Genealogy

By Wu Shiquan, 38th Generation Descendant, Editor-in-Chief of the 15th Revision March 2010

The Seed of a Dream

In the spring of 2001, while spending the Lunar New Year at my son’s home in Nanjing, our family of five took a weekend trip to Meicun Town in Wuxi. We went to seek the ancient lands of the Wu State and to pay our respects to the enduring legacy of our ancestor, Taibo. Since I was a young child, my father had told me that the birthplace of the Wu surname was Meili (modern-day Meicun). That seed of “seeking one’s roots” was planted in my heart long ago; this journey was the fulfillment of a childhood dream.

Upon arriving at Meicun and entering the Taibo Temple, we paid homage in the Zhide Main Hall. There stands the seated statue of Taibo, the founding ancestor of the Wu clan, wearing a ceremonial crown and royal robes, his feet in red ritual shoes—a figure of dignity and kindness. We moved on to admire the statues of the twenty-four kings of the Wu State and their loyal ministers in the east and west wings before strolling through the temple grounds.

By chance, we came upon the office of the “General Genealogy of the Chinese Wu Clan” committee, then located within the temple. We were warmly received by kinsmen currently working on the records. During our conversation, one kinsman searched the catalog of the Shanghai Library’s Genealogy Department for me. He discovered they held a copy of the Tangpu1 Wu Clan Genealogy. I was overjoyed; the thought of reviving our family lineage took hold of me instantly. I immediately asked my son to visit the library the next time he was in Shanghai.

However, the result was disappointing. The Shanghai collection was a 1916 edition (the Year of Bingchen), but it consisted of only two volumes. It was insufficient for a full reconstruction. We returned empty-handed.

The Bitter Search

To revive a genealogy, one must trace back to the source and clarify the branches of the lineage; this is impossible without an “old record” to serve as a blueprint.

For someone like me, who had never been involved in such work, the old records were essential as a template to follow. In short, trying to rebuild a genealogy without the original records is like building a castle in the air—pure fantasy.

Yet, knowing the Shanghai Library held a fragment, I felt certain that a full copy must exist somewhere in Tangpu or its neighboring areas. Armed with a sincere heart and the conviction that the records could be found, I began a long and arduous search.

The journey was bitter and frustrating. Clues appeared from time to time, but they were all “flowers in a mirror or the moon in the water”—illusions that vanished as soon as I pursued them.

By 2008, my investigation revealed a tragic history: the last time the Tangpu Wu Clan Genealogy had been revised was in 1947, shortly after the victory against Japanese aggression. The work began as the smoke of war was still clearing and was hastily concluded in 1948 amidst the booming cannons of the Civil War.

That edition was ill-fated.

Almost as soon as it was finished, legal disputes over entries led to a lawsuit. The records were eventually declared void by public notice. Aside from a few private copies and some public records, the rest were sealed in the ancestral temple.

During the early years of the Cultural Revolution, the hall was demolished to build a public assembly hall. The records were moved to a private home, and a few volumes were scattered. When the “Smash the Four Olds” campaign swept the nation, the remaining books were seized. Viewed as “feudal dross,” they were left in a local office to rot. Eventually, a flood soaked the building, and the records were destroyed entirely.

As for the private copies held by branches that had moved away, they were either confiscated or quietly destroyed by families fearing political persecution.

We knew that one public copy had been entrusted to a veteran for safekeeping after the revolution. He survived the political movements, and the records remained with him until after the Cultural Revolution. Sadly, he lived a lonely life with no heirs. After his passing, his nephew cleared out his belongings; he took the furniture of value and burned everything else—including the Book.

My hopes vanished in those rising flames. After eight years of searching, I had not seen a single page of our history.

The dream seemed to have turned to ash.

A Turn of Fortune

At the end of 2008, a kinsman named Chuanjia reached out to me, enthusiastically offering his support and a willingness to search together. This reignited my passion. We searched Tangpu and its surrounding areas again and expanded our scope to the libraries and archives of Shaoxing and Shangyu, but the result was the same: nothing.

Then, just as we felt we had reached a dead end, light appeared. One evening in July 2009, Chuanjia and his son, Jianbo, came to my house beaming with joy. As soon as they entered, they shouted: “The genealogy is found!”

I was so stunned I couldn’t react. Jianbo explained: “It’s in the Zhejiang Library. I found it online.”

I asked him to pull it up on the computer. Sure enough, the website displayed: “Tangpu Wu Clan Genealogy, 1916 Xiaosi Hall Woodblock Edition, 36 volumes in 24 books, held by Zhejiang Library.”

Truly, it was a case of “searching for her a thousand times in the crowd, only to turn around and find her there in the fading lantern light.”

I was overwhelmed with emotion. Nearly a century had passed since that edition was printed. I knew that revising the records now would be unimaginably difficult, but I also knew that if we did not do it now, the difficulty would only grow with time until it became impossible. The Tangpu Wu clan would lose its history; we would become trees without roots and water without a source. I vowed then and there:

I will spend the rest of my life completing this genealogy, to honor our ancestors and to leave a legacy for our descendants.

The news spread like wildfire. Supporters gathered, a committee of founders was formed, and the 15th revision of the Tangpu Wu Clan Genealogy officially began.

Immense Challenges

This was the first time the Book of Genealogy was to be revised since the founding of the People’s Republic, and the difficulties ahead were easy to imagine. From the very outset, we were confronted by three major hurdles; had even one of these remained unresolved, the entire project likely would have been abandoned halfway.

Challenge I: The Shift in Mindset

The first great hurdle was a shift in perception. Specifically, was reviving the genealogy even necessary?

The Tangpu Wu clan has a venerable tradition of record-keeping. From the late Tang Dynasty, when we shared records with the Zhoushan and Zhuji branches, to the early Ming Dynasty, when we began our own independent registry, we have been diligent. In the 229 years between the reign of Emperor Kangxi (1718) and the Republican era (1947), the genealogy was revised eleven times, averaging once every twenty-two years. These records served as the sinew of our clan, a tangible bond of kinship that compelled each generation to continue the work.

However, after the 1949 Revolution, a “philosophy of struggle” was championed—one that claimed “there is boundless joy in struggling against others.” This ideology eroded the very foundations of family affection. This peaked during the unprecedented Cultural Revolution, which utterly subverted the traditional virtues of honoring one’s ancestors and harmonizing with one’s kin. Relatives were treated as strangers, or worse, as bitter enemies.

As we began our work, we were met with a constant barrage of skepticism and reproach: “The Cultural Revolution destroyed all the family trees; why bother now?” or “What use is a genealogy in this day and age?” Had we ignored these voices, the project would have stalled. To raise awareness, we issued a “Letter to Our Kinsmen,” widely promoting the idea that documenting our lineage in this era of prosperity is an act of remembrance and a means to reunify the family. We argued that family genealogies, alongside national history and local gazetteers, are the three pillars that support the historical edifice of the Chinese nation. They are cultural heritages highly valued by the state, and the ultra-leftist excesses of the Cultural Revolution must be corrected to set things right.

[redacted]

The second point of contention was the methodology of the revision itself. The focal point of the disagreement was whether daughters should be included in the genealogy. Historically, under the feudal patriarchal system, only males could inherit the family line—a practice that clearly discriminates against women and stands in direct contradiction to our modern Constitution and national policy.

Consequently, we explicitly proposed that only-child daughters and their children (both grandsons and granddaughters) be permitted to carry on the lineage (indicated by a “red line” in the pedigree charts). While this move comforted families with only one daughter, it drew fierce opposition from others: “Historically, only sons get the red line. If daughters can have it too, what kind of genealogy is this?” Some sons-in-law also worried: “If my children are registered in the Wu Clan records, what happens when my own family revises its genealogy?”

With only-daughter households making up nearly half of our families, the project could not proceed without dispelling these prejudices. In our “Second Letter to Kinsmen,” we argued the legitimacy of including daughters from the perspectives of law, national policy, and biological bloodlines. We clarified that a genealogy is not a government household register; a person is not restricted to appearing in only one family’s records. By presenting facts and reasoning, the controversy was finally settled. Not only did these families participate with confidence, but they also donated generously—of the nine households that donated over 10,000 RMB, three were families with only daughters.

Challenge II: Tracing the Ancestral Source

The second major difficulty was tracing ancestral origins. If a clan member cannot find their ancestor in the existing records, they cannot “link” to the lineage, making the revision impossible.

Initially, we worked from the 1916 edition, which was 94 years old. Even an infant recorded in that edition would be 95 today; effectively, not a single person currently alive appeared in those books. To find a link, one needed the names, zodiac signs, spouses, and burial details of deceased fathers, grandfathers, or even great-great-grandfathers who were born before 1916.

This forced us to add an extra step to the process: a comprehensive ancestral investigation. But, while we found matches for many, some kinsmen knew nothing of their ancestors beyond one or two names. Searching for a match in over twenty volumes was like looking for a needle in a haystack. We could neither ignore them nor manufacture a false connection, leaving us in a total deadlock.

Fortunately, a 1947 edition was discovered in the Village of Wujialou shortly after. This closed the gap by 31 years and finally solved the puzzle. (We owe an eternal debt of gratitude to Mr. Bingcan for preserving this record; his story is detailed in a separate entry in Volume Hai).

[Robert: You can find this fascinating story in Articles #2 and #3]

Challenge III: Securing Funds

The costs of such an undertaking are high. Since genealogical revision is a private, grassroots activity, no government funding is available. Given the economic circumstances of the Tangpu Wu clan, no single individual could shoulder the cost alone, and a “per-head” tax would have been insufficient and met with resistance.

We took a different path: voluntary sponsorship with symbolic rewards. We encouraged the successful to give more, allowed those in financial difficulty to give nothing, and asked everyone else to contribute what they could. This proved effective, eventually raising over 300,000 RMB.

We are deeply grateful for this generosity and have treated every cent with the utmost care. Aside from small stipends for those doing door-to-door registration and data entry, all members of the two committees served as volunteers without pay. As a result, we raised enough not only to finish the genealogy but also to allocate half the funds toward rebuilding the Wu Clan Ancestral Hall. To date, the main structure—the Hall of Ancestors—has been completed. This serves as both a comfort to the spirits of our ancestors and a token of gratitude to all who donated.

[Redacted]

#2 How the pivotal 1947 edition was found, literally by a dream

The Dream That Saved Our History

Following the opening ceremony, we worked on two parallel tracks: the door-to-door collection of modern family data and the painstaking search for our ancestral roots. While the data collection went smoothly, the search for our origins was repeatedly thwarted.

The reason was clear: we were relying on the 1916 edition of the Tangpu Wu Clan Genealogy. A century of separation is a vast chasm to bridge. Every time we hit a dead end, a collective sigh would ripple through the committee: “If only we had the 1947 edition... all these problems would vanish.” This became a mantra for everyone involved, especially for Mr.Zhigang, who was assisting me in the search.

As the saying goes, “What you think of by day, you dream of by night.”

Late on the night of October 4th, after an exhausting day of work, Zhigang drifted into a deep sleep. In his dream, he saw a mountain of genealogies in the distance. He rushed forward and grabbed one in his hands, only for it to vanish into thin air. Just as he was overcome with frustration, a mysterious voice whispered to him: “The records you need are in the village of Wujialou!” He woke with a start, realizing it was but a dream.

A Leap of Faith

The next morning, Zhigang couldn’t shake the dream. He thought to himself: Better to believe it might be true than to dismiss it. He knew there was a branch of our clan in the village of Wujialou, and a visit there was already on our schedule. However, because it was “just a dream,” he didn’t dare tell anyone for fear of being ridiculed.

After breakfast, he drove alone to Wujialou. With a mix of hope and doubt, he began visiting kinsmen and spreading news of the project. Eventually, local clansmen led him to the home of a respected elder, Bingcan.

Bingcan had long hoped for a revival of our family records. Without a moment’s hesitation, he brought out his most precious treasure: the 1947 edition of the Tangpu Wu Clan Genealogy, which he had hidden for decades and had never told anyone about. When Zhigang saw the words “Wu Clan Genealogy” inscribed on the wooden chest, he realized his dream had manifested into reality. Overcome with emotion, he dropped to his knees and bowed deeply in reverence.

This display of sincerity deeply moved Bingcan and the onlookers. When Zhigang asked to borrow the records for the committee, Bingcan immediately agreed. Beside himself with joy, Zhigang called our office to report the news. It was nearly noon, and the entire committee was waiting breathlessly for his call.

#3 How the pivotal 1947 edition was saved from destruction

The Secret Guardian of Wujialou

Exhilarating news arrived on the morning of October 5th: the 1947 edition of the Tangpu Wu Clan Genealogy—a set we had spent years searching for to no avail—had been discovered in the tiny mountain village of Wujialou.

The set consisted of twenty-six volumes. Three had been lost to history (held as court evidence in the former Zhejiang High People’s Court due to an old lawsuit), but the remaining twenty-three were largely intact, save for some minor damage from silverfish. After decades of political upheaval and natural disasters, these volumes were the “lone survivors” of a shattered legacy. The man responsible for saving these precious texts was Bingcan, a descendant of the Fusao branch.

An Unlikely Protector

Bingcan was a 36th-generation descendant of our ancestor Wenjian. Since the early days of rural credit cooperatives in 1958, he had worked at the Haoba Credit Union, eventually serving as its director from 1964 until his retirement. He was a man of integrity—down-to-earth, sincere, and widely respected. This public standing provided the “outer shield” that allowed the records to survive in his hands. But more importantly, it was his inner conviction—his deep understanding that a genealogy is the soul of a family—that drove him to act.

The records had originally been kept by a branch leader named Guijin. However, after the 1949 Revolution, Guijin was classified as an “Upper-Middle Peasant.” In that era of radical leftist politics, he realized he was powerless to protect the books and handed them to Chun-yue, a kinsman from a “Poor and Lower-Middle Peasant” background, hoping the political classification would act as a safeguard.

During the “Four Cleansings” movement in 1964, a government work team seized the records to search for political evidence. After the team left, the books were returned, but they were treated with total neglect.

Rescuing History from Playgrounds

Bingcan watched this decline with a heavy heart. He remembered a time when generations before him treated the genealogy as a sacred object, kept in multiple layers of silk and hidden away. In the old days, opening the records was a ritual: one had to wash their hands, change clothes, and burn incense as if visiting the ancestors in person. Now, seeing them cast aside was a desecration. Yet, given the social climate of the time, Bingcan didn’t dare act rashly.

The breaking point came when he saw local schoolchildren, ignorant of what the books were, using the ancient pages as scrap paper for doodling and scribbling. He knew that if he didn’t act, the history of our clan would be wiped out.

Risking everything, Bingcan leveraged his local prestige and his devotion to his ancestors to literally “rescue” the volumes from the children’s hands. Once he got them home, he commissioned a custom camphor wood chest to fit the dimensions of the books, painted it, and carefully inscribed “Wu Clan Genealogy” on the front. He then hid the chest, keeping its existence a secret from everyone.

This happened during the early, most volatile years of the Cultural Revolution. To commit such an “act of righteousness” in that suffocating atmosphere required extraordinary courage and vision.

Not only did Bingcan save the books, but he also became a self-appointed historian; from 1947 onward, he kept a small notebook where he recorded every birth, death, marriage, and burial within his branch, significantly simplifying our work today.

The Missing Link Found

When Bingcan learned that we were officially revising the genealogy, he did not hesitate to offer his cherished collection to the committee. The impact was immediate: Those who previously couldn’t find their ancestors were linked instantly; Doubts regarding “best-guess” connections were resolved with hard evidence; Rare errors where families had linked to the wrong branch were corrected in time.

The role this 1947 edition played in ensuring the quality and speed of our project cannot be overstated. The merit of Bingcan, who guarded these books through the darkest times, is indelible. His name shall be recorded in this genealogy so that he is never forgotten.

#4 An account about the Taiping Rebellion

Original title: 粤匪扰害记A Record of the Disturbance by the Canton Rebels

Since our sage and enlightened Emperors founded this dynasty, the world has long been governed with benevolence. For over two hundred years, the people shared in the blessings of peace. However, during the Daoguang and Xianfeng eras, natural disasters struck repeatedly. Solar eclipses and earthquakes occurred; rivers overflowed, and mountains crumbled. A comet appeared in the sky, trailing like a long bolt of white silk. The strange and ominous portents of that time were too many to recount.

Following these signs, bandits from Guangdong [the Taiping Army] rose up in swarms; they were known as the “Longhairs.” In the autumn of the year Xinyou (1861), they suddenly overran Shaoxing, and the people’s livelihoods were plunged into misery.

Our Wu clan of Tangpu, having previously received imperial decrees to recruit local militia, had prepared for such a crisis. When the rebel forces arrived to plunder our town, we stood our ground and fought them in several battles, both large and small. But alas, we were outnumbered, and our food supplies failed. We held out until the autumn of the year Renxu (1862), when our defenses were finally shattered. Our town was reduced to scorched earth.

In the slaughter that followed, some died as martyrs for the cause, while others chose death over surrender. Our women refused to be defiled; even as they were being beheaded, they chose to remain as ghosts of their home soil rather than suffer dishonor. Some were taken captive, while others scattered to the winds. The old and the young fled for their lives, many perishing in exile—becoming lonely souls wandering in foreign lands. Alas! This was the will of Heaven; what more can be said?

By the immense grace of the Emperor, General Jiang Yili was commanded to lead the land and naval forces. Since breaking the enemy at Quzhou, he cleared the entire province of Zhejiang in less than a year. This allowed our people to return to their ancestral lands, rebuild their homes, and reunite with kin to recount the sorrows of their separation. We have gathered the bleached bones of the fallen and laid them to rest in the snowy mountains.

Ultimately, it was the will of Heaven that restored stability to the state, but it was the exertion of men that suppressed the rebels. As subjects, we shall forever be grateful for the kindness of our Sovereign. Thus, we can say that the will of Heaven, the efforts of man, and the grace of the Emperor are all bathed in the magnificent fortune of the Son of Heaven.

We also rely on the devoted service of our generals. Now, as we undertake the revision of our family genealogy, we specially record these events to honor the fallen and clarify our history. Let future generations read this and remember.

Date: The seventh year of the Tongzhi Reign of the Great Qing (1868).

It is shocking on several levels.

First, Zhang Youxia was the most senior uniformed officer in the country. Second, his relationship with Xi Jinping went far beyond ordinary patronage. Their families’ ties dated back to the revolutionary years in the Northwest, when their fathers fought side by side. Zhang was widely seen as one of Xi’s most trusted confidants inside the military system. Third, this came after years of high-profile, rolling investigations into top generals that have already fundamentally reshaped the PLA’s command structure. With this latest development, Zhang Shengming, head of military discipline, is now the only remaining uniformed member of the Central Military Commission. Every other vice chair and member in uniform is gone.

And yet, in another sense, Zhang’s fall is not entirely surprising.

Before becoming CMC vice chair, Zhang oversaw armaments and the military space program, and essentially the core of China’s military-industrial complex. Over the past few years, the most systematic and far-reaching investigations have been concentrated precisely in this area: weapons procurement, aerospace, missiles, and their surrounding defense enterprises. Almost all of the senior military commanders and major heads of defense companies were ensnared. It was a great reckoning inside an opaque system long known for its capacity to corrupt.

Under that lens, it would have been more surprising if Zhang, as the man at the top of that system, had been personally untouched.

Yet, this is still emotionally shocking. Under the old logic of Chinese elite politics, where personal loyalty, shared history, and patronage networks often provided implicit protection, this should not have happened. The fact that it did tells us something important.

The most plausible explanation is that Xi is demanding absolute purity inside the military system, and it reinforces a broader pattern we have seen repeatedly: Xi does not care about face-saving. He is a man of a deep sense of mission, with little interest in looking benevolent or maintaining appearances. If something is rotten, he does not compromise and settle for a half-deal. If someone, no matter how close that someone used to be, must fall to serve his objective, so be it.

For ordinary observers, however, the theatrics of court politics quickly give way to two practical questions.

First, what does this mean for the People’s Liberation Army’s fighting capability? Is the PLA being weakened by the removal of so many senior officers? The second question is inevitably about Taiwan.

Read more

This briefing is reserved for paying subscribers. (Visit here to claim your Baiguan member benefits for free access)

Read more

Tactically, the operation was very impressive. It was decisive and effective, and a long-standing adversary was removed in a single stroke, without prolonged escalation or domestic political cost.

Strategically, it could well end up in a disaster. History offers plenty of reminders that tactical brilliance and strategic wisdom are not the same thing. The Iraq War, too, was executed with speed and overwhelming force, and went on to reshape the region in ways its proponents neither intended nor controlled. And the fact that Donald Trump is great at improvisation, but never at strategic thinking, only hardens the suspicion.

Inevitably, China also becomes part of the conversation. Ostensibly, only hours before the dramatic capture, a Chinese envoy had just met Maduro himself. This is bad optics showing China’s cluelessness, while also reinforcing the perception that Caracas sits firmly within Beijing’s geopolitical orbit.

For years, China has been Venezuela’s most important external lifeline: a major creditor, a long-term oil partner, and a diplomatic “friend” when the regime was isolated by the West. In the public imagination, especially outside China, Venezuela is therefore not just another failed state, but a symbol of China’s influence in the Global South and of its willingness to back embattled regimes in defiance of Washington. Seen through that lens, the Maduro episode is almost automatically interpreted as China-adjacent, whether or not Beijing had any role in the events themselves.

Two different kinds of narratives quickly emerge. One treats the Maduro operation as a sort of a green light for China, as if Beijing were watching Washington to see what precedents it might now exploit elsewhere. The other frames it as a warning, a demonstration of American resolve meant to deter China from testing U.S. red lines.

Both are quite silly.

First of all, although the Jan 3 operation is a massive gold mine for Beijing’s PR efforts, Beijing does not need a “green light” from Mar-a-Lago before making moves in its own neighborhood. China’s decisions regarding Taiwan are governed by domestic legitimacy issues and calculated risk, not by whether the U.S. follows international law. Beijing sees America’s current lack of respect for both morality and legality as a given, and a flexible tool of convenience. It is neither surprised when the U.S. breaks the rules, nor does it feel it needs “permission” to follow suit.

Also, is this a “permission” anyway? For a person as unscrupulous as Trump, the distaste for double standards doesn’t exist at all. So he could perfectly square his power to arrest another head of state at will, while preventing anyone else from doing the same. Beijing is not so misguided as to take this “moral signaling” seriously.

And to suggest that this operation can act as some kind of deterrent is beyond ridiculous. What deterrent? That U.S. special forces might parachute into Zhongnanhai and abduct Xi Jinping?

What both camps of these commentators obfuscate is that this event, fundamentally, has not much to do with China. It’s America’s problem, not China’s. What is really at stake is what kind of power the US is and what type of image the US wishes to project to the world. China is, by and large, only a bystander watching this show.

Then there is also the usual refrain that China is giving up on its “ally”. Once again: China has no allies and is not seeking alliances. There has never been a Chinese version of the Warsaw Pact, and there will not be. China does not like the idea of fighting other people’s wars. It’s just not the Chinese way of thinking about things, and I have discussed this before.

There are three substantive things, however, that China will really be pondering as a result of this episode, and I will explain in the paid section below. (Visit here to claim your Baiguan member benefits for free access)

Read more

In 2025, that assumption finally collapsed. Trump’s handling of the trade war and Ukraine didn’t just spook allies. It forced a long-delayed thought into the open: what if the system’s guarantor is no longer reliable or predictable?

People didn’t say it out loud. They just bought metal.

Gold, silver, platinum, and copper rallied from one all-time high to another, and with no end in sight. It was a clear vote of no confidence on an epic historic scale.

This is one strange thing about 2025: events of genuinely historic magnitude kept happening — and then immediately disappeared. Social-media cycles were so compressed, turbo-charged by the proliferation of AI-generated content, that our minds barely had time to register before attention moved on to the next short video.

The United States bombed Iran’s nuclear facilities. Say that sentence slowly. A few years ago, this would have been considered an extinction-level escalation scenario in many risk models. In 2025, it happened — and then it passed. It was an enormous gamble, and Trump, to his credit, won this gamble spectacularly. Iran backed down. Oil did not spike above $120 a barrel. But what exactly did he win? Peace? Or proof that peace now depends on preemptive bombing followed by social-media amnesia?

Then there was the India-Pakistan conflict, with two nuclear powers, engaging in what may have been the largest aerial dogfights since World War II. Another event that should have permanently reshaped perceptions, yet barely stayed in the global consciousness for more than a week.

The assassination of Charlie Kirk was another moment that should have lingered longer than it did. Many people saw him as one of the last public figures who genuinely believed talking to the other side was not a form of treason. He is now dead. His death marked not just political violence, but the further erosion of the idea that dialogue itself is safe. American polarisation crossed into a darker phase.

Polarisation showed up economically as well. 2025 was the year the U.S. economy visibly bifurcated. One half of it was AI. The other was everything else. Capital, talent, chips and electricity flooded into the race toward AGI, turning AI into the single pillar supporting economic growth and equity markets. At the same time, much of Main Street felt increasingly strained. Long-duration interest rates stayed stubbornly high despite Fed cuts — an ominous signal for the rest of the economy untouched by the love of AI.

But even for AI, we already start to hear nervous chatter about a bubble. Yet, this is the last engine of American growth. It will not stop. It must not stop. The cost of shutting it down would be unbearable. The music has to keep playing.

Subscribe now

Doubts about America’s future would not have felt so sharp if they had been offset by disappointment with China. But disappointing, in 2025, China clearly is not.

If 2025 was the year the myth of American exceptionalism finally cracked, it was also the year China decisively broke free from the doom and gloom narratives that had dominated the past few years.

What was striking was not just the reversal, but how fast it flipped. Barely a year ago, China was “uninvestable.” Empty investment conference seats. Sliding equities. Endless variations of the same story: China is peaking. China is collapsing. China is Japan. China is entering “an age of malaise”, poetically claimed by Evan Osnos, long-time China correspondent of The New Yorker.

Perhaps the best personification of this extreme pessimism is none other than my favorite , who pounded at the table and screamed:

China at the Peak! (July 2023)

Will China squander its moment in the sun? (Jan 2024)

(Read my rebuttal to him here)

Then suddenly the script flipped. “A great reckoning.” Dan Wang’s Breakneck became a phenomenal success and, for the first time in living memory, made China a mirror for the US intellectual elites to review itself. A flood of near-identical YouTube thumbnails and Twitter threads: I visited China and was shocked, as if written from the same template.

Admiration quickly morphed into panic. China isn’t just doing well, the story went — China is eating the world. China is too capable. China is unfair. China is making trade impossible.

The turning point had names. It started with DeepSeek and a frenzied few days in late January, causing Nvidia to lose $600 billion in value in the biggest one-day loss in human history, forever killing the previously ironclad assumption that the AI gap between the U.S. and China was vast and unbridgeable. It wasn’t.

The India-Pakistan conflict delivered another unintended demonstration effect. For the first time in history, the world saw Chinese weaponry perform in real combat conditions, and it’s not even the most advanced systems in its arsenal.

In the meantime, China’s surge in biotech was set to replicate its success in EVs and is on track to dominate the world, again. This, along with the stories of China-made diamonds, China-made wine, and even China-made caviar and foie gras, shows that this current wave of China possesses an unstoppable, all-encompassing nature.

And there is even more. The success stories of Nezha 2, Black Myth: Wukong, POP Mart, Rednote, and even luxury goods show the world that this wave of success isn’t just confined to the material world, but also manifests itself in the world of soft powers and intangibles, where China was long believed to be bad at.

At this stage, it seems only high-end chips and top talents are the US’s two remaining edges over China. But the first one appears on track to be over soon, while for the second one, Trump was busy self-sabotaging with visa restrictions and tighter anti-immigrant policies.

Then there was the trade war. Contrary to expectations, China did not panic. No theatrical escalation. No desperate concessions for Western applause. Just patience, timing and textbook game theory. Watching it unfold felt less like geopolitics and more like a Tai Chi demonstration against a furious heavyweight boxer. Every swing was met — not head-on, but sideways. The boxer thought he’d scored a knockout, then noticed he couldn’t breathe.

(I have to thank Trump and his trade war here, which made me one of the very few China-based authors writing for The New York Times, which is quite an amazing personal experience this year.)

Why did the U.S. miscalculate? At its core, it’s the hubris and self-complacency. A misguided belief in its permanent supremacy, which, in part, stems from what I dubbed “ideological monotony” last year. Just as the same belief has blinded the DC establishment and globalist elites to the woes of ordinary masses, it has also driven the Trumpists to the epic miscalculation that the US still has unmatched leverage.

And why didn’t China cave in? Because it had been preparing. A year ago, China’s emphasis on security over growth was mocked by commentators who saw it as an overreaction or ideological rigidity. Noah Smith attributed China’s peaking to China’s leadership, which wouldn’t give up “their obsession with control.” Wei Lingling claimed that “Xi’s tight control hampers stronger response to China’s slowdown.”

In 2025, the laughter stopped. It turned out that years of self-sufficiency efforts, redundancy building, and belt-tightening mattered. If not for these measures, China could have easily buckled when the real pressure arrived. What was deemed stupid suddenly looks far-sighted.

By the end of this confrontation, a de facto G2 world had taken shape. The U.S. has finally realised it misjudged the balance. In Kübler-Ross's five stages of grief terms, it has already moved past denial and anger, and now drifts somewhere between bargaining and depression, edging slowly toward acceptance. The empire is not yet collapsing, but it is clearly retreating. At home, it is consolidating and re-industrialising. Abroad, it is quietly disentangling from peripherals like Europe and Japan, while staying muscular in its own backyard, to Venezuela and Cuba.

China, however, is not free of risk. As external pressure eases, it may finally gain the strategic space to confront deep internal imbalances. If left unresolved, economic discontent will find a way of mutating into something darker, more grotesque, from both ends of the political spectrum.

Fortunately, we are witnessing clear signals in that direction. If China can pull this off again, we might have peace and prosperity for several generations at least.

With that said, I wish all of you a fruitful, exciting, and happy 2026!

2025 in pictures:

On the far left, a movie review of Youth (芳华) - a fairly unremarkable 2017 film about a military performance troupe set in the 1970s and 1980s - suddenly went viral on domestic video platform Bilibili, at least for a few days.

Youth was never meant to be a grand historical statement. It’s the kind of movie you watch once, feel a vague sense of nostalgia, and then move on. This time, however, the film was taken out of the grave and radically reinterpreted. A popular Bilibili reviewer reframed it as a hidden ode to the Great Proletarian Cultural Revolution.

Not a tragedy. Not a warning. An ode.

The claim itself is almost comical. I won’t waste your time by going into details of this movie. But what I can tell is that, if anything, the movie can only be described as something to criticize and reflect on that era. Moreover, both Feng Xiaogang, the director, and Yan Geling, the author of the original novel, are commonly known as harsh critics of the Cultural Revolution.

So this claim is about as credible as arguing that The Godfather or Zootopia is secretly praising the Cultural Revolution. In fact, under the same logic, any story reflecting some elements of class difference, injustice, or social stratification can be reverse-engineered into Maoist revolutionary nostalgia.

And yet, the video worked.

Partly because the influencer is genuinely good at storytelling, the review racked up tens of millions of views in just a few days before it was eventually taken down. On Bilibili, viewers can send so-called “bullet chats”—comments that fly across the screen as others watch the same segment. I was told this video accumulated more than 300,000 bullet chats, enough to completely drown out the actual film if you turned them on.

Many of those bullet chats repeated the same slogan: “Long live the people.”

In this worldview, the Cultural Revolution was a righteous uprising of the proletariat smashing entrenched elites. Mao was a misunderstood hero, smeared by later narratives, whose true intentions are now being rediscovered by a new generation.

This interpretation sits in direct conflict with the official Chinese position, which describes the Cultural Revolution as a catastrophic mistake—an “apocalypse” that destroyed institutions, cultural heritage, and lives. This contradiction alone made the video politically radioactive.

On the opposite end of the spectrum, another ideology, once fringe, is moving closer to the mainstream. It is often referred to as 皇汉, or Han ethnonationalism.

A focal point of this worldview is a radical re-evaluation of Qing dynasty history, summarized by what its proponents call the “1644 historical viewpoint.” The argument is stark: true Chinese civilization ended with the fall of the Ming dynasty in 1644. What followed was a Manchu conquest—an ethnic minority subjugating the Han majority. The Manchus committed atrocities, governed incompetently, and ultimately weakened China so badly that it made the Chinese nation miss the Industrial Revolution and endured more than a century of humiliation.

Had the Ming survived, the argument goes, China might have industrialized on its own and never would have had this humiliation.

Needless to say, this directly contradicts the official CPC narrative of Zhonghua minzu—a unified Chinese people composed of 56 officially recognized ethnic groups, including Han, Manchu, Mongolian, Tibetan, Uyghur, Kazakh, Miao, and others. In the official framework, Qing history is not a foreign occupation but an integral part of a continuous, multi-ethnic civilization.

The official crackdown on this was swift. One of the most popular accounts advocating the “1644 historical viewpoint” has been shut down, while the state apparatus appears to be so alarmed that it has already issued several high-profile takes attacking it. To have a sense of the severity of the official pushback, you can take a look at an article translated by , a Substack affiliated with Xinhua.

How to make sense of this

Despite standing at opposite ideological poles, these two schools of extreme thought share several striking similarities.

Both narratives mix some elements of truth with distortion. In both cases, these elements of truth are lifted out of historical proportion and reorganized into simplistic moral frameworks.

Yes, Manchus did commit atrocities during the invasion of China Proper. But the Manchu may be the most sinicized ethnic dynasty throughout Chinese history, whose leaders spoke Mandarin and wrote in Chinese, and deeply respected the Chinese way of life. By the end of the Qing, the difference between a Manchu person and any other Chinese person was almost entirely unintelligible. Also, are we really sure that, had it been the Ming, not the Qing, who had met with the Western powers at the sunset phase of its dynastic cycle, we really could have done much better?

Moreover, what Han-nationalists always conveniently forget to mention is that had it not been for the Qing, China’s territory would not have been as large as it is today. It was the Qing Dynasty that first allowed China to effectively control Tibet, the Northeast (Manchuria), and a large part of Mongolia. And if not for the Qing, a political power hailing from China Proper would not have regained Xinjiang after 1000 years of losing it following the An Lushan Rebellion. If Manchus were not “Chinese”, what about those regions? Should we split China into pieces now if we really want to be fundamentalist about it?

As for the Cultural Revolution, yes, it did involve genuine elements of class struggle and strong anti-establishment sentiment (which many people today didn’t know about; they just knew it was “bad”). But the modern-day champions of cultural revolution once more always forget to mention that it was also, at the very core, a great power struggle of Chairman Mao against his many perceived political rivals. (It’s just that, unlike typical tyrants, he used the mass movement and chaos as opposed to guns and poisons to achieve this end.) And in the ensuing chaos, the evil of people found its way out, destroying lives, both corporeal lives and mental existence. There was no revolutionary romance about it at all.

At this point, it is also imperative to mention that both narratives also thrive in the blind spots of official narratives. Some topics—such as ethnic relations—are forced down into people’s minds in rigid, slogan-like terms, without ever being put up for debate. Others—most notably the Cultural Revolution—are acknowledged and then sealed off, with little room for sustained public discourse. As a result, even educated audiences often lack a deep, nuanced understanding of these issues.

When discourse is absent, fragments of truth become dangerously persuasive. Partial facts, stripped of context, can feel revelatory. When people know nothing about the nature of the Cultural Revolution, beyond the mere fact that it is “bad”, then anyone who reveals the anti-establishment side of it and claims it to be “good” will immediately gain immense credibility.

I also remember this vulnerability personally. When I was fifteen and first learned about Tiananmen Square, you could have told me almost anything bad about the CPC, and I would have believed it—not because I was stupid, but because silence creates an intellectual vacuum.

Finally, both movements feed on anxiety, especially among young people. Economic pressure, social immobility, and a sense of lost direction make grand explanatory narratives emotionally attractive. They don’t just interpret history; they assign blame and offer clarity. This is the same psychological terrain behind the silencing of figures like Zhang Xuefeng and Hu Chenfeng that I commented on in October.

Which brings us to the real irony: The emergence of these extreme narratives is, in large part, the by-product of the censorship regime itself.

If certain histories cannot be openly discussed, if official narratives leave no room for questioning or reinterpretation, then alternative explanations will inevitably fill the gap. Crackdowns may remove the most visible expressions, but they do not address the conditions that produced them. In fact, suppression often strengthens the sense that something important is being hidden and will inevitably resurface, leaving the official machinery in perpetual cycles of whac-a-mole.

Today, these views are still not mainstream. But history rarely turns on what people believe today. It turns on what accumulates quietly over years and decades. More importantly, the socio-economic conditions that underpin their emergence can only worsen in the foreseeable future.

If extremism grows in the shadow of silence, while no organic counter-narratives are nurtured, who would have the power and credibility to contain it if one day, for whatever reason, when that whac-a-mole machine no longer functions?

Subscribe now

Starting this week, I will try to make the “Briefing” section more like the early days of this newsletter, where I write weekly reviews of events I think will matter beyond news cycles.

1. PDD’s scuffle and the media blackout A scuffle involving PDD senior employees and market-regulation officials briefly surfaced last week—and then almost completely vanished. Bloomberg ran two stories, but they failed to propagate on China’s social media at all, which was quite unusual. Caixin also published, then pulled its piece within hours.

What stands out is not the incident itself, but the level of media control it revealed. This was not routine censorship, nor a broad political taboo, but a highly targeted blackout centered on a single private company. PDD appears to possess an unusually strong ability here.

I published a piece yesterday in my personal investment newsletter about my view of PDD right now, where I also described this incident and its implications in more detail. My investment newsletter is now a paid one, but if you are already a paying subscriber of Baiguan, you can enjoy a 56% discount. You can find that coupon in the Baiguan Member Exclusive Package.

The rest of this briefing is also paywalled. But again, for Baiguan subscribers, you can claim your complimentary access, also at the Baiguan Member Exclusive Package.

Read more

The column starts with a blunt question: What can the world still sell to China? Harding’s conversations in China — with economists, technologists, business leaders — all pointed to the same awkward answer: not much beyond commodities like soybeans and iron ore, which don’t help Europeans sleep at night. Some mentioned luxury goods, some mentioned higher education, and several even suggested that Europe should just let Chinese companies build factories there. But after running through all these options, Harding lands on a stark conclusion: “nothing.” Nothing that China truly wants to import, nothing it believes it cannot make better and cheaper on its own.

He does acknowledge, to his credit, that China still buys semiconductors, aircraft, advanced software, and high-end machinery — but then immediately frames China as the “resident doctor,” merely buying these goods until it can produce its own and eventually export them. And he concedes China’s insecurity caused by US export controls. All of this is reasonable analysis. Yet it leads him to a sweeping final diagnosis: “The only good solutions lie with Beijing,” because only China can fix its deflation, boost consumption, strengthen the renminbi, or curb industrial subsidies. The rest of the world, he argues, can only respond with either politically painful reforms that reduce welfare or protectionism.

This is where I think the framing needs a reset — not because Harding misses the right issues, but because he stops at exactly the moment the harder questions begin. He correctly identifies the pressures. But he underestimates the number of levers the West still has, and overestimates China’s ability (or willingness) to close every door forever. Most importantly, he jumps past the most basic principle of trade: if you want to sell, produce something worth buying. If you don’t, the customer goes elsewhere. Trade is reciprocal: you sell when your value is competitive; you buy when it isn’t. It’s astonishing how much geopolitical commentary often loses sight of this simple truth. Instead of complaining that the client no longer buys from you, maybe start by asking whether the product still holds up.

So let me offer three specific areas where the FT narrative — and much Western policy thinking — unnecessarily narrows the field.

First, high-end technology. Harding writes that China still buys the sophisticated machinery it cannot yet produce. Yet the column treats this like a temporary, fading opportunity. What it doesn’t ask is why these sales dropped so sharply: not because China lacks demand, but because Western governments blocked the supply. China did not lose interest in EUV lithography machines; it lost access. Nvidia did not decide China was no longer a customer; Washington did. If those controls are relaxed tomorrow, demand would be immediate. The FT frames China’s tech trajectory as inevitable; what’s equally inevitable is that the world’s top producers will always have products China cannot yet replicate - if they stay competitive. The opportunity remains — unless the West refuses it. Without a choice, China will be forced to develop its own competitiveness in these areas.

Second, services. The FT barely touches this, except for mentioning higher education. But services are precisely where Western economies are strongest. Tourism, creative industries, finance, legal services, management consulting, higher education, and design — these are not sectors where China has announced a national drive for airtight self-sufficiency. Yet instead of expanding service exports to China, many Western governments have spent the last five years discouraging them: higher visa barriers and suspicion toward Chinese students, restrictions on consultants working with Chinese institutions. China isn’t rejecting Western services; Western governments and institutions have been choking off their own exports.

Third, inbound investment. Harding recounts that several economists told him directly: Let Chinese companies build factories in Europe. And then — oddly — he drops the thread. Inbound investment is one of the cleanest, most direct ways to rebalance an economic relationship. If China buys fewer goods, let it buy more assets. Let it create jobs. Let it finance new industries. But recent cases suggest the opposite instinct. The Nexperia saga is a perfect illustration: a Chinese-owned semiconductor firm was forced to be taken over without a clear justification. I wrote about that earlier — the real message to Chinese investors was that even lawful, productive capital is politically unwelcome. Under these conditions, it’s hardly surprising that economic engagement shrinks. You can’t slam the door on someone’s investment and then ask why they aren’t contributing to your economic future.

By this point, I can’t help but sense that the undertone of FT’s article is that of condescension: you can’t be allowed to sell more than I do, nor can you be allowed to invest in my territory, because c’mon, how can you be better than us!

Harding does gesture toward the need for competitiveness: Europe needs reform, less regulation, more dynamism. That part is right. But he frames competitiveness as something Europe must pursue because China stopped buying, rather than as a way for any economy to survive in global competition. Competitiveness is not a reaction; it is a precondition. If China is manufacturing faster, cheaper, and higher-quality goods, that’s not an argument that trade is impossible — it’s an argument that Europe needs to sharpen its own game.

And this brings us back to the FT’s concluding line: that “the only good solutions lie with Beijing.” But the reality is far more mixed. Some solutions lie with Beijing. Others lie with Washington, Brussels, Tokyo — and with the companies in those economies. Relaxed export controls, expanded service openness, clearer investment rules, and improved competitiveness are all choices available to the West.

China is not making trade impossible. China is making trade harder for those who stopped trying.

Subscribe now