Is it really possible for a one-person company to rewrite the future? See the new era of AI entrepreneurship from Polsia
See the new era of AI entrepreneurship from Polsia
What would happen if one day, starting a business no longer required a whole team, no need to raise large sums of money first, and no need to learn complex technology first?
This is exactly what Polsia wants to prove.
In this interview, Kevin Rose talks to Polsia founder Ben Cera. The most shocking thing about this company is not just that it broke out in just a few weeksMulti-million dollar revenue scale, but its entrepreneurial team has only one person. Ben is not only the CEO, but also a product manager, customer service, marketer, and even the operator of the entire system at the same time. And to be more precise, he’s not entirely alone, but instead works alongside a swarm of AI agents.
What Ben wants to build is not another chatbot, nor another set of tools to help you write copy or do market research for you. What he really wants to do is a platform that can help people start from an idea and go all the way to the prototype of a company.
In his words, “Push a button and get a company.”
From using AI to assist work, to making AI a team
Ben said that the birth of Polsia was not the result of a flash of inspiration one day, but the answer that emerged naturally after he started his own business and worked with AI for a long time.
In the past year, he spent almost ten hours every day working with AI: polishing products, discussing strategies, planning marketing, and dismantling problems. Slowly, he began to realize that AI was no longer just a tool, but a kind of team prototype.
At first, even he himself didn’t quite dare to believe this idea. After all, everyone is still used to thinking of AI as an assistant: helping you write a copy, reply to a letter, and organize information. But as the models became more powerful, the reasoning capabilities became more mature, and various agent SDKs, MCPs, skills, and workflow infrastructure gradually took shape, he began to see another possibility: AI was not just helping, but could actually play a role in a team with clear roles.
This cognitive change is very critical. Because the most painful part of starting a business is often not making something, but the consumption day after day.
What exactly should the next feature do? Do you need to find customers for verification first, or add functionality first? Do you want to do marketing now? How to contact customer service? Who will fix the bug? Will a trivial matter like password reset have to be redone for the next project?
Ben believes that these tedious and repetitive tasks are actually the parts that AI agents are best suited to take over. Entrepreneurs should not waste their energy on launching again and again, but should focus more on direction, judgment and taste. If you resonate with this point of view, I also recommend reading Stop Chasing Tools! Build your “unbeatable system” in the AI era.
Polsia does not help you register a company, but helps you create a company.
When Kevin asked: “What does Polsia mean by ‘building a company’?” Ben’s answer was interesting.
He said that the first step in starting a business is never to register a company or establish a legal person. The real first step should be an idea, a gut feeling, something you think someone in the world might want.
Therefore, in his design, what Polsia does is to help you complete the most critical, but most likely to get stuck, steps in the early stages of starting a business, including: assisting in clarifying ideas, doing market research, generating mission statements, creating sales pages, starting external communications, and even arranging unfamiliar development and collecting initial feedback.
In other words, Polsia is not a company generator in the legal sense, but a system for advancing an idea into a working business prototype.
Moreover, it does not end with generating a bunch of things for you at once, but it will continue to iterate. Ben specifically mentioned that Polsia is designed as a daily cycle autonomous system. Every night, it will decide what to do next based on the current situation? If the user has explicit instructions, it will refer to it; if the user does not indicate it, it will push forward by itself.
That step might be product development, or it might be marketing, customer service, research, QA, or other operations work.
Why every night? Because starting a business requires rhythm, not just speed
A special design of Polsia is that it does not run endlessly 24 hours a day, but uses a nightly cycle. On this point, Kevin also raised a special question: Since AI never sleeps, why not let it continue working immediately?
Ben’s answer seemed pragmatic to me.
**First, entrepreneurship is not a one-way output, but requires waiting for feedback. **
For example, if you just designed the sales page of your product today and just contacted a few potential users, then the most reasonable time for the next round should be the next day, not five minutes later. Otherwise, AI can easily make products more and more complex prematurely without any real market response.
**Second, cost is also a reality. **
Ben said that although AI is cheaper than human labor, it is not really cost-free. Behind every task, model resources, server costs and overall infrastructure costs are consumed. If you want the system to have 30 days of autonomy and keep the price affordable for the average person, you must strike a balance between product design and business model.
Therefore, what he finally chose was not unlimited speed, but rhythmic autonomy.
This kind of design is actually closer to the real entrepreneurial scene: instead of rushing forward all the time, you do one round, see the reaction, and then move on to the next round.
This is not a toy for engineers, but an entrepreneurial entry for ordinary people.
Ben made it very clear that Polsia’s main target customer group is not technical experts, but people without technical background but with ideas.
Because people who really understand development have many tools available. They will think Polsia is cool, but at the same time they will soon start to want more control: their own keys, customized deployment environments, their own accounts, and their own proxy processes. If all these requirements are added, the product will quickly become complex and lose the “Oh my god, I can actually start a company” experience that ordinary people have when they first come in.
Ben used a good analogy: some tools, like Linux, are for people who love tinkering and customization; and what Polsia wants to be is more like Apple—not being able to control everything, but making the most important things simple enough for more people to dare to start.
What he is really worried about is not whether the technical community is using these AI agent tools, but that the vast majority of ordinary people do not know that they exist. Many people think they have used AI, but in fact it only means they have used ChatGPT. However, most people have not even heard of capabilities such as Claude Code, Codex, various agent SDKs, and MCPs that will truly reshape work and entrepreneurial models.
Ben believes that if only a few people take the lead in understanding and mastering these tools, then future wealth, knowledge and opportunities will be further concentrated in the hands of a few people.
What he wants to do, to a certain extent, is a product that wakes everyone up. This reminds me of what I wrote before The Cruel Divide in the AI Era: Why has the income of 95% of freelance workers been cut in half, while the income of 5% has doubled? , talks about the same warning.
AI will not only help you with market research, but also with advertising.
Another amazing feature of Polsia is that it can even run ads for you.
Ben said that in the early days, the platform’s marketing capabilities were only X (formerly Twitter) and E-mail. But this is obviously not enough. If the product really promises to help you push your business forward, then sooner or later the ability to acquire paid customers will come in. So he went through the advertising agency process.
Users only need to click “Run Ads” and set a daily budget. Polsia can automatically generate a video ad, write the ad copy, add subtitles, and execute the ad through Meta’s delivery structure. It will also add a tracking mechanism to the website to determine whether the advertisement actually leads to registration, purchase or other targeted actions.
What’s more, this entire process is deliberately kept very simple.
Ben believes that a lot of people think they want to manage advertising when in fact they don’t. What you really want is, can you help me run first? Therefore, the first version of the product did not design too many adjustable parameters, but made it a system that was tested first and the results were seen first.
This idea is actually in line with his overall product philosophy: instead of letting users control everything from the beginning, let them experience it first: it turns out that these things can really be taken over by AI.
When the platform became popular and the system went wrong, he did not hire anyone first, but asked AI to put out the fire first.
A very exciting part of the interview was when Kevin asked him: You have already reached this scale, why don’t you just hire a few engineers and a few customer service personnel?
Ben’s answer is very representative of his thoughts.
He said that when the platform suddenly experienced a sudden surge in traffic this week, the database exploded, advertising errors occurred, and users began to complain, his first reaction was not that I should recruit people, but: “If entrepreneurs on the platform encounter the same problem in the future, what will they do?”
So, what he did was to establish a customer service agent, a refund agent, a compensation mechanism agent, and a monitoring agent, and let these AIs check the user status for him one by one, determine whether to refund, give credits, or upgrade.
This is not to show off, but he deliberately let himself experience those pain points first, because he believes that only by experiencing it once can he know how to make the product and build these capabilities into the platform. In other words, he is not saving manpower, but is doing an extremely thorough dogfooding (eating one’s own dog food, which means a strategic thinking in which company employees personally use self-produced software, products or services to experience product quality, discover problems and optimize them):
Since Polsia wants to help the solopreneur survive, he must first pass these levels using the solopreneur method.
Readers who are not familiar with how AI Agent is implemented in the workplace are recommended to refer to How can professionals make good use of AI Agent? The critical shift from asking AI to letting AI do it for you.
He believes: The next company to survive must be 80% autonomous
Ben made a very strong judgment in the interview: if new startups in the future do not achieve 80% autonomy, they are likely to be eliminated.
What he means is not that humans will disappear completely, but that there are a lot of repetitive tasks such as engineering, customer service, product optimization, marketing execution and feedback collection. If the company still relies entirely on manual labor, the company will lose to the AI native team in terms of speed and cost structure.
Especially in today’s era where products are becoming cheaper and cheaper and copied faster and faster, you will not be the only manufacturer in the market, but five or ten competitors will soon appear.
Once your competitors have turned AI into a coordinated orchestra of agents, and you are still operating in traditional ways, sooner or later you will find that you are losing not on creativity, but on rhythm and execution density.
This is actually not a technical problem, but a problem of organizational imagination. If you want to understand how AI can rewrite the rules of business games, it is recommended to read Command Economy is Coming! AI is rewriting the rules of the business game in the next decade.
What is really important is not the emergence of the first one-person unicorn, but that more ordinary people wake up
Kevin asked a question that everyone wants to ask: Do you believe that a single founder will actually build a billion-dollar company in the future?
Ben’s answer was very layered.
That milestone itself isn’t the most important, he said. The important thing is, once that happens, it becomes a huge awakening event. It will force the world to admit: These times have really changed.
But compared to the birth of a single unicorn, he actually cares about another thing: if more and more people can rely on AI to turn an idea in their heads into a real business with a monthly income of hundreds or thousands of dollars, then this matter may have greater significance to the entire society.
He believes that there are too many creators and experts today. They clearly have their own opinions, their own communities, and their own sensitivities, but their monetization tools are still very limited. Either it’s professional distribution, it’s subscription-based content, or it’s just selling some peripherals.
But in the era of AI agents, they can theoretically extend their ideas into a service, a product, or a truly operational micro-business.
In the future, it is not necessarily only the super large companies that will become stronger. There may also be a whole new economic layer composed of small and precise individual businesses.
This reminds me of the article Low-risk-slash-for-freedom: From the blood-and-tear lessons of naked entrepreneurship to the practical wisdom of smart trial and error. The way to start is completely different from the past.
What Polsia wants to do is actually not just a product, but a new economic imagination.
From the entire interview, the most fascinating thing about Polsia is not just what functions it performs, but the fact that there is a very distinct worldview behind it.
Ben didn’t just want to prove how much he could do on his own, or simply show off his skills. What he really wants to push is a question: If AI is already here, will it only make a few people stronger, or will it also become a lever for more ordinary people?
This question will determine what entrepreneurship will look like in the next few years, and will also determine many people’s emotions towards AI - is it a threat or an opportunity?
Polsia 当然还在很早期的阶段，当然还有很多不稳定、很多问题以及很多尚未完成的部分。 Ben himself admits that the platform is still only part of his ideal. But because of this, this case deserves more attention. Because it is not a perfect answer, but an experiment that is happening in the field.
The most noteworthy thing about this experiment is that it allows us to truly see for the first time that the so-called “one-person company” may no longer just be synonymous with freelance workers or small businesses, but may be a new type of entrepreneurial unit.
In the future, the size of a company may no longer be determined by the number of people, but by how much AI capabilities you can mobilize, whether you can make correct judgments, and whether you have a clear enough sense of mission.
In such a world, being able to write programs is of course still important, but perhaps more importantly, do you know what you want to create?
Extended reading:
- How can professionals make good use of AI Agent? The key shift from asking AI to letting AI do it for you
- Stop chasing tools! Build your own “unbeatable system” in the AI era
- A cruel watershed in the AI era: Why has the income of 95% of freelance workers been cut in half, while the income of 5% has doubled?
External Resources:
