「一人公司」这四个字，第一次出现在了国家级政策文件里。
工信部印发的《人工智能中小企业创业支持计划（2026—2028年）》里有这么一句话：鼓励各地对依托智能工具开展敏捷创业的「一人公司」、超级个体等微型主体给予包容支持。
看起来只是一句普通表述，但它的分量不轻。在此之前，「一人公司」更多是创业者圈子里的说法——一个人加一套AI工具，从零开始做产品、做内容、做服务，营收可能不小，但公司只有一个人。官方层面从来没有给过一个明确的名分。
现在有了。
政策终于追上了现实
一人公司不是什么新概念。但AI让这件事的量级和爆发力完全不同了。
以前你一个人创业，能做的事情很有限——开发做不了设计、做内容做不了运营、做销售做不了客服。每缺一个职能，你要么自己硬扛（效率低），要么雇人（成本高），要么外包（质量没保障）。一个人能做的业务天花板很低。
AI把这个天花板捅破了。一个开发者可以用AI做设计、做文案、做客服、做数据分析；一个内容创作者可以用AI做剪辑、做封面、做选题、做分发。原本需要三五个人的团队，现在一个人加上一组AI工具就能搞定。
这件事已经在真实发生了。武汉江岸区的「众生云OPC生态社区」，入驻的一人公司创业者靠AI工具做短视频，单条作品平均浏览量突破4万。Indie Hackers上，一个人的SaaS产品月入过万美元的案例越来越多。不是个别现象，是越来越普遍的趋势。
工信部这份文件的意义，不是给了多少钱（虽然普惠算力也是真金白银），而是给了一种确定性：一人公司是被认可的、是被支持的、是有政策配套的。对很多想创业又犹豫的人来说，这份确定性比补贴更重要。
四个方面，最值得关注的是什么
整份计划有4方面15项重点任务，普通创业者最值得关注的有三点。
第一是普惠算力。深入实施普惠算力赋能中小企业发展专项行动——这意味着初创团队和个人开发者，未来可能以更低的成本甚至免费拿到算力资源。算力是AI创业最大的固定成本之一，如果这块能降下来，创业门槛会再低一截。
第二是智能原生企业培育。加快培育「智能原生」企业——从诞生第一天起就深度依赖AI的公司，而不是传统企业再加一层AI。这意味着政策资源会向真正的AI原生创业项目倾斜，而不是撒胡椒面。
第三是配套政策探索。探索根据企业自主评价结果落实落户安居、医疗保障等配套政策。这句话的潜台词是：未来一个人开公司，也可能享受到和大公司员工一样的公共服务保障。这对创业者来说是最实在的安全感。
当然，政策只是外部条件。创业能不能成，最终还是靠产品、靠市场、靠执行力。但不可否认的是：当AI把创业的边际成本压到趋近于零，当政策给了一人公司明确的名分，想试试的人会越来越多。
未来三年，我们可能会看到一波前所未有的AI创业潮——不是那种拿融资烧钱的创业，而是一个人、一台电脑、一组AI工具，从一个小切口开始慢慢长大的创业。它们可能永远不会成为独角兽，但它们会是经济生态里最有活力的那部分。
明天见。
The term "one-person company" has appeared in a national policy document for the first time.
Buried in MIIT's "AI SME Startup Support Plan (2026–2028)" is this line: Encourage local governments to give inclusive support to "one-person companies," super individuals, and other micro-entities that rely on intelligent tools for agile entrepreneurship.
It might look like an ordinary sentence, but its weight shouldn't be underestimated. Until now, "one-person company" was mostly a phrase used in founder circles — one person plus a set of AI tools, building products, creating content, delivering services from scratch, maybe generating significant revenue while employing just one person. There was never an official designation for it at the national level.
Now there is.
Policy Has Finally Caught Up to Reality
One-person companies aren't a new concept. But AI has completely changed the scale and explosive potential of this model.
Before, when you started a business alone, what you could do was very limited. A developer couldn't do design, a content creator couldn't do operations, a salesperson couldn't do customer support. For every missing function, you either struggled through it yourself (low efficiency), hired someone (high cost), or outsourced it (unreliable quality). The revenue ceiling for a solo operation was very low.
AI has smashed through that ceiling. A developer can use AI for design, copywriting, customer support, data analysis. A content creator can use AI for editing, thumbnails, topic selection, distribution. What used to take a team of three to five people can now be handled by one person with a stack of AI tools.
This is already happening for real. In Wuhan's Jiang'an district, "ZhongSheng Cloud OPC Ecological Community" hosts one-person company founders who use AI tools to make short videos, with average views per piece exceeding 40,000. On Indie Hackers, there are more and more cases of single-person SaaS products hitting five-figure monthly revenue. It's not an anomaly — it's a growing trend.
The significance of this MIIT document isn't about how much money it allocates (though inclusive compute is real money). It's about giving certainty: one-person companies are recognized, supported, and have policy backing. For many people hesitating to start something, that certainty matters more than subsidies.
Four Areas — What Matters Most for Regular Founders
The plan covers four areas with 15 key tasks. For ordinary founders, three points matter most.
First: inclusive compute. Deeply implementing the inclusive compute empowerment action for SMEs — this means early-stage teams and solo developers might get access to compute resources at lower cost, or even for free. Compute is one of the biggest fixed costs in AI startups. If that comes down, the barrier to entry drops another notch.
Second: AI-native enterprise cultivation. Accelerating the cultivation of "intelligence-native" enterprises — companies born from day one with deep AI dependency, not traditional businesses with an AI layer on top. This means policy resources will tilt toward truly AI-native startups, rather than being spread thinly everywhere.
Third: supporting policy exploration. Exploring the implementation of supporting policies like residency/housing and medical insurance based on enterprises' self-evaluation results. The subtext here: in the future, even a one-person company might get access to the same public services and benefits that big-company employees enjoy. That's the most tangible sense of security for founders.
Of course, policy is just an external condition. Whether a startup succeeds ultimately depends on product, market, and execution. But there's no denying it: when AI pushes the marginal cost of starting a business toward zero, and when policy gives one-person companies official recognition, more people will be willing to try.
Over the next three years, we might see an AI startup wave unlike anything before — not the VC-funded, burn-money kind, but a wave of solo founders starting from a small niche with one computer and a stack of AI tools, growing steadily. They might never become unicorns. But they'll be the most vibrant part of the economic ecosystem.
See you tomorrow.
当AI把创业的边际成本压到趋近于零，当政策给了一人公司明确的名分，想试试的人会越来越多。
—— Dawn Vision编辑部
When AI pushes the marginal cost of starting a business toward zero, and when policy gives one-person companies official recognition, more people will be willing to try.
— The Dawn Vision Editorial Desk
MIIT · one-person company · AI startup · super individual · AI-native enterprise · inclusive compute · SME support plan · 2026-2028
Sources · 信源 Sources
本文基于 Dawn Vision 认知引擎处理的 8 个源信号生成，经编辑部人工审核。素材来源：中国政府网、中国工信新闻网、上观新闻、武汉发布。
This article was generated by the Dawn Vision cognitive engine processing 8 source signals, with human editorial review. Sources: Gov.cn, CII News, Shanghai Observer, Wuhan Release.
