China's "High-Speed Industrialisation System" and the Real Problem of Global Competition
I spent over a decade going back and forth to China. Within the China–Hong Kong–Taiwan professional circles, I frequently encountered a very common cognitive tendency: after China's rapid rise in AI, EVs, drones, supply chains, and other fields, many people would jump straight to the conclusion that "China has comprehensively surpassed everyone, the US is declining, and Taiwan is beyond saving."
Recently I saw a similar argument from a veteran Taiwanese E-bike enthusiast on Facebook. But I believe reality is far more complex than this kind of linear extrapolation.
China's True Strength Lies in the System
What makes China truly formidable has never been just "cheap." It is that China has formed an extraordinarily efficient industrialisation system — a hyper-scale domestic market, high-density supply chains, a massive pool of engineering talent, centralised capital allocation, policy-driven local governments, platform integration mechanisms, and extremely fast product iteration cycles, all stacked on top of one another.
Today, many Chinese companies' advantages do not come from fundamentally new theoretical breakthroughs, but from the ability to engineer, mass-produce, and ecosystemise technology at extreme speed. From EVs to solar, from drones to AI applications, this model has been validated again and again.
"Powerful" Does Not Equal "Healthy"
But many people simultaneously overlook the other side.
Over the past two-plus decades, globalisation's structural division of labour has concentrated high-pollution, long-hours, low-margin manufacturing heavily in China. And the Chinese government actively attracted and absorbed these industries through policy incentives and infrastructure investment. This was a two-way process — not one-sidedly "forced into it," but a co-conspiracy between global capital seeking low-cost bases and China's development strategy.
China then hard-pushed the acceleration of its entire industrial civilisation through land-based fiscal policy, local-government debt expansion, the demographic dividend, 996 work culture, and extreme involution among engineers.
The result: industrial capability is immense, but societal pressure is equally immense.
Real-estate bubbles, local government debt risks, soaring youth unemployment, severely compressed margins in certain industries, extreme price wars — these problems have all surfaced. The profit structures in solar and low-end EVs are especially alarming. While leading companies like BYD and CATL can still maintain relatively healthy margins, the situation for companies at the tail end is an entirely different story.
The Real Question: What Happens If This Model Becomes the Global Mainstream?
So what the world truly needs to think about is not just "will China win," but:
If this high-efficiency, high-involution, high-compression industrial model ultimately becomes the dominant paradigm for global industrial competition, what kind of society will humanity become?
High efficiency does not equal high happiness.
The Structural Reality of AI
The situation in AI is much the same.
People often say "there are many Chinese people in foundational AI research," but the more accurate statement is: among the top AI laboratories in the United States, the proportion of ethnically Chinese researchers is quite significant. And most of these researchers were trained and developed within the American university, laboratory, and tech-company ecosystem. Their research output belongs to the American institutional ecology, not to a simplistic classification by nationality.
Therefore, the truly core question is not "who is smarter," but: who controls the compute, capital, markets, ecosystem, academic networks, and the entire system-level coordination capability.
And China's most noteworthy current trend is precisely its transformation from "the world's factory" to a "system-level industrial platform." In the open-source model space, DeepSeek, Alibaba's Qwen, and others have been making frequent moves recently, reshaping the open-source ecosystem landscape that was previously dominated by the US — which is itself a concrete manifestation of "system-level coordination capability."
Taiwan's Problem: Not Just Technology
On the other hand, the challenges Taiwan faces are not solely about technology itself.
It is worth noting that the "supplier positioning" itself is not the problem. TSMC's success was built precisely on the strategic choice of "pure foundry" — not developing its own brand, not competing with its customers, and winning the deep trust of Apple, Nvidia, AMD, and others through neutrality. This is not a lack of ambition, but an extremely precise form of strategic self-discipline, and the foundation of Taiwan's irreplaceable position in the global semiconductor ecosystem.
However, TSMC's model success should not obscure structural shortcomings in other industries. At the software, platform, end-user brand, and AI application layers, many Taiwanese companies still lack the awareness and capability to transform from a pure supply-chain role to that of an ecosystem integrator. The difference is: TSMC deliberately chose not to be a platform; other industries often want to but cannot, or have not even realised they need to.
The problems of Taiwan's traditional industries are in fact far more complex, and I will write a separate article to discuss them later.
Conclusion
The competition of the future may have long since moved beyond single-product-level rivalry, becoming instead a comprehensive contest of platforms, ecosystems, standards, data, and system-level coordination capability.
AI is amplifying this to an infinite degree.