In the past, if you wanted to start a tech company,
the first question usually wasn’t:
“Is the idea good?”
But rather:
Where are your people?
Who writes the code?
Who handles the design?
Who writes the marketing copy?
Who manages customer support?
If the answer to all four was “none,”
many products wouldn’t even make it to the first version.
But now, a new type of company is emerging in Hangzhou, China.
At its core, the company may have only one person.
All other tasks
are first assisted by AI.
Young People Struggling to Find Ideal Jobs Choose to Start Their Own Companies
The Wall Street Journal reported on September 20th that some young people in China are turning to AI-powered solo entrepreneurship.
The reality is quite practical.
Some can’t find satisfying jobs.
Some are fed up with long hours at big companies.
Others don’t aim to build the next tech giant—they just want to sustain themselves with their own product.
Generative AI has lowered one of the biggest hurdles:
You can start without a full team.
These entrepreneurs use AI to help code, create marketing content, and handle customer communication, while they focus their time on product, customers, and business decisions.
This phenomenon is also rooted in a tough job market.
Official statistics from China show that in August 2026, the urban youth unemployment rate for 16 to 24-year-olds excluding students reached 18.9%, the highest in about a year; it was 17.9% in July.
For some young people,
“finding a company willing to hire me”
and
“just start my own company”
are now two viable options at the same time.
Over 2,000 AI-Related One-Person Companies Already in Hangzhou
Hangzhou is one of the cities where this trend is most obvious.
The Hangzhou government’s English portal cites data from People's Daily, stating that there are already over 2,000 AI + OPC companies locally, along with more than 50 related communities.
OPC stands for One Person Company.
But first, a clarification:
“One-person companies” don’t legally always have just one employee.
Later, Zhejiang province released AI OPC group standards defining them as "controlled by one core individual," usually with no more than 10 employees.
The key idea is closer to:
One main founder + AI + a streamlined team.
So it’s not “only one person working in the entire world.”
The real change is:
Tasks that previously required five people to start,
can now often be prototyped by one person with AI.
Some Entrepreneurs Let AI Handle 80% of the Coding
In the Honghub one-person company community in Hangzhou, there’s an entrepreneur named Li Yunfan.
He used to lead software teams and later switched to solo entrepreneurship.
He developed an AI-based essay correction product.
Students take pictures of their essays,
and the system offers automated feedback.
Public data shows this product has over 15,000 users.
More interestingly,
Li Yunfan says:
About 80% of the coding work is now done by AI.
His own role has evolved to:
Strategy.
Product management.
Operations.
In other words,
he hasn’t stopped working entirely.
Instead, he shifted time away from “writing every line of code himself”
toward:
Deciding what to build, who will use it, and what the next steps are.
Without Engineers Before, the Product Might Never Have Launched
This is the real significance of one-person companies.
Imagine you were a teacher ten years ago.
You noticed a problem with students’ essays
and wanted to make an app.
The first obstacle might have been:
“I don’t know how to code.”
So you look for engineers.
Next:
“The interface looks bad.”
You find designers.
The product is finally made:
“Nobody knows about it.”
You need marketers.
Users start signing up:
“Who handles questions?”
You hire customer service.
A tiny idea,
not yet profitable,
already needed a team.
AI is changing this order.
Now it may go like this:
First, let AI help you create a prototype.
Then have AI assist with interface tweaks,
organize product descriptions,
draft community content,
and categorize common customer questions.
Only when real users arrive,
decide where it’s worth spending to hire real people.
This doesn’t mean “AI is eliminating all jobs.”
Instead,
where you once needed a team to validate an idea, now you can validate first and then build a team if needed.
Non-Programmers Are Joining In Too
Another noteworthy aspect is that entrepreneurs themselves are changing.
Honghub’s 2026 data shows about 75% of OPC founders come from non-technical backgrounds.
One founder, Zeng Xiaofeng, was an IT consultant and solutions architect without coding experience.
Using AI tools,
he created a first app prototype within two weeks,
then learned HarmonyOS and continuously refined the product with AI until it was fully functional.
This marks a big shift.
Previously, one of the biggest questions in tech entrepreneurship was:
“Can you actually code?”
Now it’s gradually becoming:
“Do you know what should be coded?”
Knowing where customer problems lie,
what users truly need,
what features shouldn’t be built,
and even when AI makes mistakes.
These judgments become more important.
You Can Prototype in Just Two Hours
This trend continues to grow.
On September 17th, APEC Young Entrepreneurs Conference held a solo entrepreneurship workshop in Hangzhou.
Participants teamed up and used AI tools to tackle real business challenges like costs, management, and customer service.
The core was a two-hour hackathon:
Generate ideas.
Build prototypes.
Debug.
Finally, present live.
Two hours isn’t enough to build a mature company.
But it proves something:
The gap between “I have an idea” and “I have something to show” is shrinking.
For entrepreneurship,
this matters more than how well AI can perform on tests.
AI Doesn’t Mean One Person Can Do Everything Alone
Reading this, it’s easy to jump to an extreme:
“Does that mean companies won’t need to hire people anymore?”
Not really.
The Wall Street Journal article also warns that many solo entrepreneurs still have limited income and frequently pivot their products.
AI can lower the cost of making a first version,
but it doesn’t automatically solve:
Will anyone buy it?
What price should be set?
Why do customers leave?
Who takes responsibility when issues arise?
How are contracts handled?
Is the data usable?
Is there enough cash flow?
None of these disappear because of chatbots.
In fact, it may be even harder,
since decisions once shared by five people
now rest on just one person.
What One-Person Companies Actually Save Is the Pre-Startup Resources
More precisely,
AI reduces the need for:
Resources you had to have before starting.
No full-time copywriters?
You can still create the first version of marketing content.
No complete design team?
You can start with a prototype.
No dedicated engineers?
You can get the product up and running.
No customer service department?
You can organize FAQs and handle key issues yourself.
When customers and revenue arrive,
when you know where hiring is truly needed,
you bring people in.
This is a very different story from “using AI to eliminate all employees.”
One approach:
From 100 employees down to 20 using AI.
The other:
Starting with just one person who, thanks to AI, finally has a chance to get something done.
Even the Number of “One-Person Companies” Needs Careful Interpretation
This concept is trending,
and various statistics use different criteria.
For example, some Hangzhou public data claims there are over 2,000 AI-related OPCs.
But Zhejiang province’s August OPC map, using official group standards, listed 2,536 OPCs across nine cities, with 1,128 in Hangzhou.
Thus, these figures can’t be directly compared or conflated.
Whatever the metric,
the direction is clear:
AI is no longer just a tool inside companies—it is reshaping the minimum viable size of a company.
Why Does This Matter to Regular People?
Maybe you have no plans to become an entrepreneur.
But this story still affects you.
Because we used to think of work as defined by “roles”:
Engineer.
Designer.
Marketer.
Customer service.
Now AI is breaking work down into capabilities:
Build a first version of a webpage.
Organize data.
Draft text.
Answer common questions.
Create simple visuals.
Analyze user feedback.
When these skills can be combined by one person plus AI,
the basic unit of work may change too.
In the future, what really matters may be not “What role am I?”
But rather:
Can I define problems myself, then combine AI, tools, and humans to get things done?
Can One Person Really Become an Entire Company?
The answer now might be shifting from:
“Impossible, you lack people.”
to:
“Just try.”
This might be the real significance of one-person companies.
AI doesn’t suddenly transform one person into four workers—engineer, designer, marketer, and support.
It just lets a single person temporarily access parts of those capabilities at lower cost.
Whether it eventually becomes a sustainable company
still depends on the true test:
Will anyone pay for what you made?
Recommended Reading
Understand AI in the simplest way every day.
SasaDaily.com
