China’s AI Boom: The Rise of One-Person Businesses and a New Entrepreneurial Model
In Shenzhen, a former product manager, laid off from a major platform company, now runs what she describes as a “business of one.” Utilizing generative AI, she creates ad copy, designs storefronts, and produces short-form video dramas—all from a repurposed industrial park with rent subsidized by the local government. Meanwhile, in Silicon Valley, the cost of a single frontier AI training run has soared to unprecedented heights. Both examples, presented without irony, are hailed as the future of entrepreneurship.
This stark contrast reveals much about where AI innovation is being built today and by whom. The dominant Western narrative about Chinese AI has long been one of catch-up—cheap imitations, intellectual property theft, and state subsidies masking an inability to compete at the frontier. However, this story is increasingly untenable. Instead, a new narrative is emerging, less about chasing trillion-dollar valuations and more about a slow, distributed reinvention of entrepreneurship itself.
Neijuan: The Concept Explaining the Shift
To understand this new wave of Chinese entrepreneurs, it helps to start with a term that gained popularity in everyday Mandarin around 2020: neijuan, literally “inward curling,” commonly translated as involution. The term was borrowed from American anthropologist Clifford Geertz, who used it in the 1960s to describe farming systems in colonial Java that became increasingly labor-intensive without productivity gains. Chinese internet users found the concept uncannily reflective of their own experiences.
Neijuan captures a specific kind of exhaustion. Everyone works harder and competes more fiercely, yet no one truly advances because the baseline continuously rises. For example, studying longer to score higher on the gaokao exam only pushes the cut-off scores higher; working extended hours in tech firms merely sets new expectations. Chinese policymakers have taken notice. According to the South China Morning Post’s coverage of the central economic work conference, combating neijuan-style competition has become an explicit priority.
More than a linguistic curiosity, neijuan drives a fundamental change: when traditional pathways to success falter—graduate degrees no longer guarantee stable employment, housing becomes unattainable, and platform economies fail to offer upward mobility—many ambitious individuals begin to look sideways instead of upwards. They start building smaller ventures, and in 2026, AI is the primary tool they turn to.
Entrepreneurial Workers: A New Class Emerges
The clearest description of this emergent group comes from a recent essay in Rest of World, which coins the term “entrepreneurial workers.” These digital laborers occupy a space between traditional employees and capital owners. They leverage generative AI to handle tasks like copywriting, e-commerce, video production, customer service, and managing creative businesses across podcasts, blogs, and independent shops.
This contrasts with the previous generation of Chinese entrepreneurs, who flourished during the mid-2010s “mass innovation, mass entrepreneurship” campaign. That era was defined by high-growth aspirations—IPOs, unicorn valuations, and the dream of becoming the next Jack Ma. Founders raised capital, scaled rapidly, and exited. Their pitch decks resembled those from Silicon Valley.
Today’s wave is different. Instead of aiming for unicorns, they seek sustainable income sufficient for basic living. This is less a growth story and more a survival narrative cloaked in entrepreneurial language, with AI as the enabler.
Thanks to AI, a single individual can perform the work of a small agency—one person, aided by AI agents and a laptop. Costs, speed, and margins compress. Yet, margins compress because everyone operates similarly. Neijuan follows entrepreneurs even into their own businesses.
DeepSeek: Efficiency in Constraint
The entrepreneurial worker model also shapes how Chinese foundational AI models are developed. When DeepSeek launched its R1 model in early 2025, Silicon Valley responded with skepticism. Despite operating under US chip export controls and with a fraction of the compute budgets of OpenAI or Anthropic, DeepSeek’s model performed competitively on reasoning benchmarks.
Western critics suggested the lab must be misrepresenting costs, distilling US proprietary models, or receiving hidden subsidies. While some critique holds weight, a more compelling explanation is straightforward: constrained by limited resources, Chinese labs prioritize efficiency. They innovate through model compression, architectural optimization, engineering rigor, and leveraging open-source ecosystems.
This reflects a new kind of frontier innovation—not the absence of ambition but ambition channeled through resource scarcity. Research by Silicon Canals has highlighted similar patterns in smaller, sovereignty-focused AI models built with far less capital. The idea that transformative AI innovation demands massive capital investment is a historical artifact, not a universal truth.
Frugal Innovation Goes Mainstream
This approach aligns with the concept of jugaad—a term from India describing resourceful, frugal innovation that creates more value with less. Previously dismissed in Western business circles as a coping mechanism for poorer countries, frugal innovation is now proving to be a structural advantage.
DeepSeek’s rise underscores that efficiency gains from Chinese labs are substantial and structurally embedded. Open-sourcing these advancements has accelerated their diffusion globally. As detailed in Silicon Canals’ report on Global South AI development, capable AI models can now run on hardware costing as little as $50. The technical stack for operating such models on modest infrastructure is accessible to anyone with the engineering skills.
What was once considered a bottom-of-the-pyramid strategy is now a mainstream innovation model, and China is the world’s largest laboratory for this experiment at scale.
The State’s Strategic Role
The romantic image of solo micro-entrepreneurs thriving solely through ingenuity overlooks a critical factor: the Chinese state’s deep involvement. The nature of this engagement is evolving.
Local governments offer support programs for solo entrepreneurs, including computing vouchers, subsidized office space in repurposed industrial parks, access to AI models and datasets, and sometimes direct financial aid. The explicit aim is to reabsorb laid-off workers from major tech platforms into new employment forms aligned with state-led AI strategies.
This contrasts sharply with Silicon Valley’s founder mythos, which idolizes solitary heroes detached from familial or governmental influence, changing the world through personal will and venture capital. Chinese entrepreneurs rarely work in isolation; they operate within policy frameworks, capital allocation systems, and governance logics, while remaining closely tethered to family obligations.
The Fragility of the One-Person Model
Despite its accessibility, the one-person AI business model is inherently precarious. Platforms can abruptly cut off traffic; algorithms shift unpredictably. AI itself compresses the value of skills these workers monetize. For example, copywriters using ChatGPT compete with clients who also use the tool, causing clients to question the need for human intermediaries.
Policy support mitigates some risks but introduces dependencies. Businesses reliant on computing vouchers or subsidized spaces are vulnerable to shifts in government priorities, fiscal tightening, or political changes. Thousands of one-person enterprises could lose crucial scaffolding if the policy environment changes.
Moreover, neijuan’s logic extends beyond individual businesses. With millions running similar AI-powered e-commerce stores competing for the same customers on the same platforms, profit margins tend to erode toward zero. The very AI tools designed to empower workers become mechanisms intensifying competitive pressure.
Rethinking Founder Profiles
Who thrives in this environment? It is not the young, charismatic founders so often funded by Western venture capital. The new entrepreneurial worker seeks stability, autonomy, and just enough margin to sustain operations.
This aligns with emerging research on founder success. A Kellogg School study of 2.7 million founders found that the average founder of the fastest-growing tech companies is 45 years old, and a 50-year-old is nearly twice as likely as a 30-year-old to build a runaway success. The myth of the 22-year-old college dropout founder is largely a cultural export, not a predictor of real-world outcomes.
Further studies suggest psychological consistency—the ability to persist through uncertainty and make steady decisions—is the strongest predictor of success, rather than visionary charisma. This precisely describes the Chinese entrepreneurial worker: steady, adaptive, and focused on sustainable operation rather than swinging for the fences.
Global Implications and the Future of Work
The critical question is not whether Chinese AI will catch up with American AI in technological terms, but whether the entrepreneurial model incubated in Shenzhen’s subsidized industrial parks proves more portable globally than Silicon Valley anticipates.
The necessary ingredients—affordable, capable AI models; a generation of educated workers displaced from traditional sectors; and tools enabling one person to perform agency-level work—are emerging worldwide. These conditions exist in southern Europe’s graduate-heavy job markets, gig economies of South Asia, and Latin America’s sprawling urban centers. China’s experiment is simply the largest and earliest.
If this model succeeds, the one-person AI-powered business will not remain a Chinese anomaly—it will become a global preview. This raises uncomfortable questions: Is someone tethered to platforms, policy cycles, and family resources truly an entrepreneur, or merely a worker bearing all the risk alone?
Silicon Valley is misreading China’s AI story by focusing on the search for another Jack Ma. The real development is quieter and more unsettling: the realization that you can run a business of one powered by AI and still lack true freedom.
Source: Here
