@10xmylife: https://x.com/10xmylife/status/2078690248645910965
Summary
Independent developer Haiming shared his experience and thoughts on running a one-person company (OPC) in the AI era, including using AI tools such as Cursor and ChatGPT to reduce production costs and improve efficiency, while emphasizing that market risks still exist, and providing an execution methodology.
View Cached Full Text
Cached at: 07/20/26, 07:36 PM
OPC: Challenges and Opportunities in the AI Era | Offline Sharing
Invited by Zhihu, I attended an offline sharing event in Shanghai on the 18th, with the theme “Challenges and Opportunities for One‑Person Companies in the AI Era.”
At the event, I exchanged thoughts on AI, solo entrepreneurship, and the changing nature of work.
Today I’m organizing and publishing the slides and speech content from that sharing, hoping it can offer some inspiration and help to those exploring one‑person companies, independent development, or paying attention to changes in the AI era.
Hi everyone, I’m Haiming, an independent developer. I’ve built many apps, ranked on the App Store charts, and even experienced earning a thousand dollars a day. This time, invited by Zhihu, I want to share my experience with OPC (One Person Company).
The theme of this article is: The challenges and opportunities OPC faces in the AI era. I’ll cover three aspects:
The fundamental difference between OPC and employment, how to improve one‑person company workflow efficiency in the AI era, and the advantages and disadvantages of OPC.
1. OPC vs. Employment: The Fundamental Difference Between Specialization and Versatility
The biggest difference between employment and OPC can be summed up in one sentence: Employment requires you to be more specialized; OPC requires you to be more versatile.
When I worked as a programmer at ByteDance and Tesla, I found that in a workplace environment, I faced clear job logic: I needed to do my own job well, and my work value would be amplified by the organization. The organization typically had its own unique evaluation system. Workplace demands were about digging deep into a specialty, and risks were borne by the organization—rarely falling on me personally. This was a protected, specialized way of survival.
But after starting an OPC, everything changed. A one‑person company is essentially a startup—you have to face market competition directly. The value of your labor must be validated by the market: whether the product or service you create is valuable depends on whether users are willing to pay and how much they’ll pay.
If no user pays you, and you haven’t found any other monetization method, then what you make has no value for your company—it can’t help you survive.
So after going solo, I realized that in the past I only had to handle development work. Now I have to get involved in operations, growth, marketing, product, and more. Any weak link directly impacts monetization. At the same time, all risks are on your shoulders—there’s no organization to back you up.
2. The Shape of a One‑Person Company in the AI Era
Last year, one of my tweets went viral, essentially saying: Starting a business in 2025 only costs $20 a month.
The company structure is: CEO – me; Technical partner – Cursor; Interns – ChatGPT, DeepSeek, etc.; Promotion channels – X, Xiaohongshu, Reddit, Douyin; Company advisors – Google and GitHub; Documents stored in Notion. Stacking all these tools, the monthly cost is just Cursor’s $20 subscription. The office can be any garage.
This is the simplest form of a one‑person company: a computer with internet access plus AI is enough.
A major change in the AI era is that technical development roles are compressed into callable tools like Cursor. They handle execution, while my job shifts to decision‑making.
By 2026, the biggest change is the emergence of Agents. AI is no longer just a tool for answering questions—it can take on continuous tasks on my behalf, has memory, can use tools, and handle more complex workflows. My feeling is that these Agents are increasingly like employees in a company—finance, advisor, tech, assistant, etc. Most of my time now is spent interacting with Agents because they have stronger connectivity and operational capabilities. I barely need to operate software or web pages myself, and I’m consciously reducing my involvement in their work.
For example, I recently integrated Hermes into Feishu to manage my finances. I tell it all my income and expenses. Behind it, I use DeepSeek’s API, costing less than 10 RMB a month. It works brilliantly.
With Agents, most of my time goes to interacting with them, letting them do the concrete execution, while I handle review and then deliver to the market for validation.
3. OPC Execution Methodology
With AI, how should a one‑person company actually execute? Here are a few takeaways from my experience.
First, find a real problem. Real problems reflect real needs. Many people already know this, but it remains the starting point for everything.
Second, once you find the need, don’t execute it yourself—hand it to AI. Whether it’s building an app or creating content, let AI do it. If I were to build another app today, I wouldn’t write the code myself. I’d use an Agent to run a simple version and take it directly to market for validation. If someone pays for it, I assess if it can sustain. If no one pays, no problem—move on to the next. The cost of delivering with AI is low enough to absorb the loss from a single failure.
Third, always share the experience after practice. As long as you practice, you’ll gain insights and lessons. These experiences are very valuable. I believe that running an OPC must go hand in hand with building a personal brand on social media—it’s the cheapest way to drive traffic. My products have gained nearly ten million views on Xiaohongshu, simply because I keep sharing.
Fourth, in this AI era, whether you end up making money or nothing from building products, tools, or services, put the process and insights on social media. People will be interested. For example, making a product with AI in 2026 and not sharing it is like taking lots of short videos in 2016 and not uploading them to Douyin—a huge waste. You must share frequently.
4. Advantages of OPC: One Person Can Also Have Leverage
Best productivity leverage: AI; and product leverage that generates income
First, OPC is low‑cost entrepreneurship. No need to build a team or invest in large assets. A computer with internet access is enough to quickly validate a need, keeping overhead very low. You can also obtain AI—the best productivity leverage—at an extremely low cost. Whether $20/month or $200/month, it’s still much cheaper than hiring someone.
Second, you can invest a lot of time in building replicable assets. As mentioned in The Almanack of Naval Ravikant, code, books, courses, and social media content are all replicable assets. You invest in creation once, and the cost of subsequent distribution is near zero. You don’t need to spend a lot of time again, yet they can generate passive income continuously.
Finally, passive income brings you closer to freedom. You no longer rely on a salary or sell your time to survive. Then you can continuously invest in your products, letting them keep generating value, freeing you to create more content—a clear positive cycle.
5. Disadvantages of OPC: AI Lowers Production Costs but Not Market Risk
I need to be honest: AI can lower production costs, but it does not lower market risk. You may still, like any other entrepreneur, invest a lot of time with zero income, create a product nobody needs, or receive praise without sales. That’s completely normal.
In the AI era, OPC offers huge opportunities, but the cost is also real.
How to avoid this risk? My advice: try not to burden yourself too much—don’t quit your job rashly, don’t go into debt, and learn to reduce costs.
Conclusion: Freedom Is a Beginning, Not an End
Finally, I want to leave you with this: Freedom is a beginning, not an end.
Many people want to be freelancers or start an OPC because they dream of time freedom—no more punching a clock. But in reality, you need to spend more energy thinking: How will you use that time? What will you do?
When you take back the power of choice, the real challenge begins. You have to face market pressure head‑on. That challenge is difficult—something I never encountered while working a regular job—but it’s also incredibly interesting.
Honestly, when you earn your first dollar, ten dollars, a hundred, a thousand, or ten thousand dollars from your own product, service, or content, that feeling is incomparable to receiving a paycheck. I’ve always been grateful that I tried OPC.
If you want to start an OPC, I hope this sharing has inspired you. I’m Haiming. Thank you.
Similar Articles
@ba_niu80557: Let's talk some hardcore practical knowledge while I have time this morning. What actually happens between signing a contract for an AI project and it finally running in production? I'll lay out the entire playbook. People in this field can copy it directly, and those not in it can still understand why 95% of enterprise AI pilots end up dead. First, let me say something counterintuitive to the point you might not believe...
This article discusses common reasons for the failure of enterprise AI projects from proof-of-concept to production deployment, highlighting key practices such as MLOps, early inspection of real data, and clear human-machine boundaries. It argues that project failures are often not due to model issues but due to neglect of the engineering implementation phase.
@Dragonhau66: A young Chinese entrepreneur with no programming background leveraged AI tools to build a one-person company (OPC), achieving an annual revenue of 10 million RMB! He used AI tools like ChatGPT to design paid app utilities. His workflow is as follows: 1. Insight into needs 2. Refining requirements 3. Summarizing requirement documents 4. Creating the product…
A young Chinese entrepreneur with no programming background successfully developed a paid app and achieved an annual income of 10 million RMB by leveraging AI tools like ChatGPT within a one-person company model. The article summarizes his workflow for using AI to assist in requirement insight, prototype design, and MVP validation.
@PandaTalk8: https://x.com/PandaTalk8/status/2065311206840353163
The article analyzes the wave of layoffs caused by AI, pointing out that one-person companies (OPC) are a trend but most people can't make money; the key to success lies in serving high-value clients, doing dirty work that cannot be standardized, and maintaining long-term patience.
@xiaohu: https://x.com/xiaohu/status/2071796715162857477
The Every team publicly shares its compound engineering methodology and open-sources supporting plugins. Through a four-step cycle and multiple AI agents working in parallel, a single-person team manages 5 products, spending 80% of the time not writing code.
@CMGS1988: https://x.com/CMGS1988/status/2074488576356876585
The author shares personal insights on the relationship between AI and infrastructure, arguing that while AI amplifies execution capabilities, it cannot replace human judgment on complex trade-offs and organizational issues. The author illustrates this with their own experience developing the cocoon/sandbox project using models like Fable 5.