According to industry reports, new registrations for one-person companies (OPCs) in China will reach 12 million in 2026, a year-on-year increase of 47%. AI isn't destroying jobs; it's smashing the cost of "doing" to the floor and pushing the value of "knowing what to do" to the ceiling.
What this is
A One Person Company (OPC) is not a solo proprietorship. The difference lies in the ceiling: a solo proprietor's ceiling is their daily time and physical energy, while an OPC's ceiling is how many digital employees it can mobilize. Anthropic CEO Dario Amodei has publicly predicted that the first one-person company with a $1 billion valuation will emerge in 2026.
An office building in Nanshan, Shenzhen, packs in 200 companies, none of which have a second employee. In Hangzhou, someone uses 12 AI agents (AI programs capable of autonomous task execution) to run market research, copywriting, ad placement, and customer service, making decisions themselves and earning 2 million RMB a month. A Dutch developer used Cursor (an AI coding tool) to build a mini-game in 3 hours, achieving an annualized revenue of $1 million in 17 days.
The source article outlines eight directions: the time-lag play in AI video, precision services via vertical niche accounts, full-process automation in AI e-commerce, information gap monetization in enterprise services, traffic funnels in education and training, dimensional strikes in going global, the OPC model, and the endgame form of OPC plus Agents plus globalization. There is only one underlying logic—in an era where technical barriers approach zero, "knowing" is more valuable than "doing".
Industry view
We note that these cases have indeed proven viable. A Shenzhen team used Jimeng AI + JianYing to produce a 23-episode comic drama for 12,000 RMB, garnering over 100 million views and over 400,000 RMB in ad revenue. "Sister Li" in Zhengzhou used an Xiaohongshu account with zero followers to generate beauty notes via AI, profiting 400,000 RMB in 6 days. A 3-person team in Hangzhou provides AI automation solutions for cross-border e-commerce, with annual revenue exceeding 5 million RMB—they aren't selling technology; they are selling "helping enterprises eliminate 3 operations positions."
However, we are concerned about the counter-voices. The time-lag play implies an extremely short window—early birds profit, while followers most likely lose money; plummeting viewership in the AI video track once supply surges is already the norm. A $1 billion valuation for a one-person company is an extreme case; a large number of OPCs earn less than 10,000 RMB per month. More crucially, a zero barrier to AI tools means a zero moat—if you can use DeepSeek to write scripts, so can others. Liu Run mentioned that software is shifting from expensive durables to cheap consumables, but the profit margin of consumables itself is razor-thin.
Impact on regular people
For Enterprise IT: SME owners still think a data dashboard costs 50,000 RMB and a corporate website costs 15,000 RMB; in reality, AI has crushed these costs to under 1,000 RMB. This information gap is both an opportunity and a warning—if your business is essentially an information middleman, your survival window is closing.
For Personal Careers: The value of execution-oriented roles continues to decline; judgment and information-gathering capabilities have become core competitive strengths. As the cost of "doing" is driven to the bottom, "knowing what to do" appreciates instead.
For the Consumer Market: AI-generated content will flood the market, and users' sensitivity to quality will conversely rise. Lowered content barriers ≠ lowered monetization barriers; the logic of consumers paying for scarcity and trust remains unchanged.