The Chinese “cannot be allowed to export their way back to prosperity”, argues US Treasury secretary  Scott Bessent, who claims that China’s economy is the “most unbalanced in history”. Such remarks reflect the growing fear in Washington that China’s overcapacity, subsidies and dumping are distorting global trade.

The more pressing concern, however, is not what China exports, but how. Global cost structures are indeed being reshaped, but by a quieter and more complex force: relentless productivity improvements. China is not merely moving more goods; it is exporting a new production model powered by automation, artificial intelligence ( AI ), and state-guided industrial optimization. This shift is disruptive, deflationary and still largely misunderstood.

China’s rise as the world’s factory in the late 20th century was driven by labour and scale. But now, China aims to achieve a new form of dominance through intelligent infrastructure. No longer confined to apps or chatbots, AI has been embedded across the physical economy – guiding everything from robotic arms and warehouse fleets to autonomous production lines. For example, Xiaomi’s “lights-out” factory in Beijing can assemble ten million smartphones annually with minimal human intervention. AI conducts a symphony of sensors, machines and analytics that form a tightly woven industrial loop, driving efficiencies that traditional manufacturers can approach only incrementally.

Nor is this technology-driven ecosystem confined to a single factory. DeepSeek’s 671-billion-parameter open-source large language model is already being deployed not just for coding but also to optimize logistics and manufacturing. JD.com is revamping its supply networks through automation. Unitree is exporting bipedal warehouse robots. And Foxconn ( Apple’s primary manufacturing partner ) is developing modular, AI-led microfactories to reduce its dependence on static production lines.

These examples may not represent “prestige innovation”, but they do attest to a broad culture of industrial optimization. Under the banner of “new quality productive forces”, the Chinese government is rolling out AI pilot zones and subsidizing factory retrofits; and cities like Hefei and Chengdu are offering local grants that rival the scale of national initiatives elsewhere.

The strategy echoes the one pursued by Japanese industry in the 1980s, when automation, lean production and industrial consolidation helped firms outcompete global rivals. But the Chinese approach goes further, blending AI with economies of scale, feedback loops and a unique cultural dynamic known as involution ( nei juan ): a self-perpetuating race to optimize and outcompete, often at the expense of profit margins. BYD, among the most vertically integrated automakers globally, recently cut prices across dozens of models, triggering a US$20 billion stock selloff.

In sectors from e-commerce to electric vehicles ( EVs ), this practice has driven such relentless cost compression that the state has occasionally seen fit to intervene. In April 2025, the People’s Daily newspaper warned that extreme involution was distorting market stability, citing a destructive price war in food delivery between JD.com, Meituan and Ele.me. And the problem is even more acute in the EV industry. While more than 100 Chinese EV brands currently compete, more than 400 have gone out of business since 2018.

The arena of global competitiveness is unforgiving. Those who survive emerge leaner, more adaptive and better positioned than their legacy counterparts. That is how successful Chinese EV makers have managed to edge into Europe, offering models at price points that local firms struggle to match. Viewed from afar, the process looks chaotic. In practice, though, it resembles natural selection. China is deliberately promoting industrial evolution: the state fosters a wide field of contenders and then lets the market winnow the field.

This approach is rippling across industries. In solar panels, Chinese manufacturers now account for over 80% of global production capacity, driving prices down more than 70% over the past decade. And a similar trend is emerging in EV batteries, where Chinese firms dominate the cost-per-kilowatt curve. But make no mistake: this deflation does not stem from oversupply or dumping. It reflects redesigned cost structures, which are the result of AI, intense competition and relentless iteration.

Thus, Chinese industry has made efficiency a tradable asset – one that is reshaping global pricing dynamics. Once this shift really takes hold, businesses around the world will find themselves adjusting their own pricing strategies, labour deployment and supply-chain configurations.

But this development presents new challenges for many economies. Consider the role of central banks, whose mission is to ensure price stability. What can they do if inflation is subdued not by weak demand, but by superior supply-side efficiency coming from abroad? Most likely, monetary policy will lose traction in such a scenario. The march of software advances will not slow just because interest rates rise or fall. Instead, industrial policy will have to come to the fore – not as protectionism, but as an adaptive necessity. The core divide will no longer be between capitalism and state planning, but between static and dynamic systems.

The US Inflation Reduction Act and Chips and Science Act, as well as the EU Green Deal Industrial Plan, did represent early Western efforts to challenge China’s lead; but these packages were largely reactive, siloed or focused on upstream nodes like chips. While the US and its allies deploy tariffs, subsidies and export controls, the real competition is over integration of AI into the real economy: not who builds the smartest chatbot, but who builds the smartest factory, and whose model can be sustainably replicated at scale.

Of course, the Chinese model has trade-offs. Labor conditions may worsen under relentless cost-cutting; oversupply remains a systemic risk; regulatory overreach can derail progress; and not all efficiency gains translate into shared prosperity. Consumers may benefit, but workers and smaller firms will bear the brunt of the adjustment.

But even if the Chinese model is not universally replicable, it raises important questions for policymakers everywhere. How will others compete with systems that produce more, faster, and cheaper – not through wage suppression, but ingenuity?

To dismiss China’s approach as merely distortive misses the point. The Chinese government is not just playing the old trade game harder; it is changing the rules, and it is doing so not through tariffs, but through an industrial transformation. If the last wave of globalization chased cheaper labor, the next one will chase smarter systems. Intelligence will no longer live only in the cloud – but in machines, warehouses, and 24/7 assembly lines.

China’s most important export today is not a product, but a process. And it will redefine the nature of global competition.

Jeffrey Wu is the director at MindWorks Capital.

Copyright: Project Syndicate