*There is an old Chinese saying: a clever housewife cannot cook without rice. Beijing is only now grasping how literally it applies to semiconductors.*

—–

For the past three decades, China built one of the most formidable technology industries in the world — not by inventing new tools, but by deploying existing ones with extraordinary skill. It took the internet, and made WeChat. It took e-commerce, and made Alibaba. It took short-form video, and made TikTok. The model was consistent: absorb, adapt, scale, dominate.

In culinary terms, China became a nation of brilliant chefs. It could take any ingredient and produce something more refined, more addictive, and better suited to local taste than the original. The rest of the world watched, alternatively impressed and unsettled.

What it did not do — or did not need to do — was grow the ingredients itself.

The Umami Trap

Chefs understand that the deepest flavours come from combination. Glutamates from vegetables, inosinate from meat, guanylate from mushrooms: stack them correctly, and the result is more than the sum of its parts. Chinese technology companies understood this instinctively. Fuse payments with social media with logistics; layer algorithmic recommendation with bottomless content supply; compress the distance between desire and delivery to near zero. The products were genuinely extraordinary.

But beneath this sophistication lay a structural vulnerability that prosperity had made easy to ignore. The processors powering those servers came from Taiwan and America. The operating systems came from California. The design software, the industrial tools, the foundational layers of the entire stack — imported, licensed, assumed to be permanent.

The assumption held, until it didn’t.

When the Supply Chain Becomes a Weapon

In 2019, the United States placed Huawei on its Entity List, cutting the company off from American chips and software. What followed was a masterclass in geopolitical leverage: a company with $100bn in annual revenue, one of the world’s most sophisticated telecom equipment makers, suddenly unable to source the components its products depended upon.

The lesson was not lost on Beijing. Nor on anyone paying attention.

The most talented chef in the world cannot serve dinner if the market is closed. And if the market is controlled by a rival who has decided that your success is a threat to their interests, the closure can be permanent.

China’s technology sector had, in the language of supply chains, concentrated risk it had not priced. The growth years had been spent mastering the kitchen, not building the farm.

The Reluctant Farmer

What followed has been less a strategic pivot than a forced reckoning. Huawei spent years developing its own chip architecture, its own operating system, its own ecosystem — a process the company’s founder, Ren Zhengfei, described as “building a plane while flying it.” The costs were staggering. The gaps remain visible.

More recently, DeepSeek’s emergence as a credible large-language model built under export-control constraints surprised Western observers. It suggested that restriction, applied with enough pressure, can accelerate domestic capability — though whether this represents a genuine breakthrough or an efficient workaround is a debate still unresolved.

The pattern is consistent: China’s foundational technology push is reactive before it is proactive. The roadmap was not drawn in ambition so much as written by adversity.

The Problem With Learning to Farm Late

Growing food takes time. So does building semiconductor fabs, training chip designers, and developing the deep industrial knowledge that underpins frontier hardware. TSMC’s moat is not merely capital — it is decades of accumulated process knowledge, skilled engineers, and iterative refinement that cannot be replicated by a government mandate and a large cheque.

Beijing knows this. Its current five-year plan pours resources into basic research, materials science, and advanced manufacturing. Progress is real but uneven, and the timeline is long.

The uncomfortable arithmetic is this: China needs roughly ten years to close the most critical gaps. It is not obvious that ten years are available. Export controls are tightening. The network of allied restrictions around advanced chips, chip-making equipment, and AI infrastructure is expanding. Each year the frontier moves; the distance to catch up does not shrink as fast as the investments suggest it should.

A Bet on Time

China’s technology strategy is, at its core, a wager on time. The application layer — the part that touches consumers, drives revenue, and produces the data that trains models — remains world-class. The engineering culture, the speed of iteration, the sheer scale of the domestic market as a proving ground: these are genuine advantages that do not disappear because Washington tightens export rules.

The question is whether those advantages can be sustained long enough, and monetised deeply enough, to fund the slower, less glamorous work of building the foundational stack from scratch.

Brilliant chefs have, before now, become serious farmers. The transition is not impossible. It is merely expensive, slow, and — for a technology sector accustomed to compounding returns — deeply unfamiliar terrain.

The harvest, if it comes, is still years away.