Involution, neijuan (内卷) in Mandarin, literally means “inward curling.” It became everyday Chinese speech around 2020, describing what happens when too many people chase too few opportunities. And in the AI boom, that familiar pressure is turning into something new: not just more competition, but a different kind of entrepreneurship, where workers use AI to squeeze value out of constraints.

The core idea is simple, and the stakes are not. Backed by the state, Chinese workers are using AI to deal with limited resources and drive innovation to compete with Silicon Valley. That is the promise of the boom, and also the lens for how it operates. Instead of treating AI as a glamorous frontier project, many teams treat it as a productivity engine and a shortcut for experimentation. When you do not have endless compute, time, or slack, you build differently. You iterate faster. You automate more. You test more ideas with fewer resources.

To understand why this matters, it helps to decode the cultural and economic context. The story of neijuan borrows its framing from American anthropologist Clifford Geertz, who described a pattern of behaviors in everyday life that becomes self-reinforcing. In China, the term captures how intense internal competition can produce incremental improvements, but it also hints at a darker side: people “curl inward,” competing against each other rather than expanding the pie.

Now overlay AI. AI changes the “pie” calculus. It can reduce bottlenecks, such as access to expensive expertise or specialized tooling, by letting fewer people produce more output. In practical terms, AI can help workers handle tasks that would normally require more resources, and it can enable teams to build, test, and refine offerings without waiting for perfect conditions. That creates an entrepreneurship model that is less about one moonshot and more about continuous application, where workers become the agents of experimentation.

This is where state backing enters the picture. The source emphasizes that the push is backed by the state, which matters because it changes the incentive structure. When national priorities align with industrial capability building, AI becomes not just a product roadmap, but also a coordination mechanism. State support can encourage rapid adoption, reduce some early-stage risk, and help mobilize talent and resources toward shared objectives. That can compress timelines and amplify learning loops.

At the same time, regulatory background is always part of the Chinese AI story, even when the details are not front and center. The reason is that AI systems touch sensitive areas: data, employment, information integrity, and more. When regulation tightens or shifts, it shapes where companies and workers can experiment. So a strategy that looks like “entrepreneurship from the constraints” is also entrepreneurship within boundaries. Teams that succeed are the ones that learn what they can do quickly, how to do it efficiently, and how to keep operations within whatever guardrails exist.

Second-order implications follow quickly. If AI experimentation is distributed among state-supported workers, innovation may look less like Silicon Valley’s venture-backed, founder-led rocket and more like an ecosystem of practical builders. That can affect everything from speed of iteration to the types of problems that get solved first. It can also change where competitive pressure lands. Rather than just facing a handful of well-funded model labs, peers may face a swarm of teams applying AI to real workflows, compressing the time it takes for new capabilities to reach customers.

For executives and board members, the strategic question is what this shift means for competitive planning. If Chinese AI development is producing innovations by using AI to manage limited resources, then the speed advantage is not only about better models. It is also about better execution under pressure. That is the kind of capability that can be hard to imitate, because it relies on incentives, organizational learning, and talent deployment. Silicon Valley and other AI hubs may still lead on certain foundational research, but the market can still move quickly when applied innovation scales.

The bottom line: involution describes the pressure, and the AI boom turns that pressure into a production system. Backed by the state, Chinese workers are using AI to deal with limited resources and drive innovation to compete with Silicon Valley. Decision-makers should treat this as a signal that AI entrepreneurship is evolving, and it may be evolving in ways that reward operational agility as much as technical brilliance.