Over the past two years, one of the questions I have been asked most often is: Will AI replace people? Over time, I have become less willing to answer it. Not because it is unimportant, but because it may be drawing many people’s attention in the wrong direction. AI may replace some tasks, and it may also create new kinds of work. But there is something else that makes me more cautious: If one day everyone has roughly the same tools, what will still be yours?
We are going through a rather unusual leveling of capabilities. In the past, if you did not know photography, editing, design, English, or code, many things were hard to accomplish. Often, simply knowing how to do something was itself a competitive advantage. AI is rapidly reducing the value of simply knowing how. If you cannot draw, you can generate; if you cannot edit, AI can handle much of the basic work; if you do not know English, you can translate; if you cannot write copy, you can get ten versions in seconds. For many people once kept outside the door by specialized skills, this is the first time they have had the chance to turn an idea into something tangible. But when everyone can do it, knowing how may matter less on its own.
We once worried about not knowing how to do something. Increasingly, the concern may be having nothing worth doing. Many techniques that required specialized learning only six months ago can now be reduced to a button. So I spend less and less of my energy on how to use a particular tool. I care more about this: once the tools are the same, what will you do with them? The difference may increasingly lie not in the tools, but in the person.
AI may raise the baseline very high. Images, videos, articles, music, and proposals that are “pretty good” could become abundant and inexpensive. Then the real question becomes: why you?
I am increasingly wary of a certain kind of person: someone who knows a little about everything, yet has nothing of their own. If all of those capabilities are bestowed by tools, others may soon have them too. So I ask less often, “Which tools do you know?” and more often: what remains after those tools are gone?
What genuinely belongs to a person? The first thing is experience. What you have truly done, where you have failed, what you have learned the hard way, whom you have met, and which problems you have handled. I have worked in the content industry for nearly a decade. Much of my judgment about things does not come from having read a book; it comes from having actually encountered them. Not all of these experiences were successes. Some were not especially flattering. But over time, they gradually become judgment.
The second thing is judgment. AI can keep adding answers, but it does not relieve you of deciding which one is worth trusting, which one should be set aside, what should be done now, or what should not be done even if it can make money. AI can help you analyze, but someone still needs to take responsibility for the final words: “I choose to do it this way.”
The third thing is taste. Taste here is not about whether something looks good; it is about what you are willing to keep. What you believe is worth bringing into being, what kinds of people you are drawn to, what kinds of stories you are willing to tell, and what you would not make even if the data looked good. Repeat these choices a hundred or a thousand times, and they gradually form a person. Real style is the one option you keep choosing, over time, from among a hundred possibilities.
The fourth thing is relationships and credibility. AI can generate an excellent business email, but it cannot give you a decade of working relationships. AI can write a highly professional article, but it cannot automatically make others trust you. When anyone can create a very polished presentation in ten minutes, we may still end up asking: who is this person, really?
The fifth thing is the assets you truly own. Do not merely have traffic; have characters. Do not merely have a video; have a system that can keep producing. Do not merely have a follower count; build understanding among users. Do not only produce content for platforms; retain your own IP, copyright, data, methods, relationships, and brand. I increasingly feel that one of the greatest risks for an ordinary person in the future may not be unemployment. It may be realizing, after ten busy years, that nothing was left behind.
AI may be forcing us to understand “work” anew. If machines can perform a large share of execution-based work, what should people really spend their time doing? I am increasingly drawn to one answer: do the things that can remain. Build judgment, work, relationships, credibility, IP, your own data and methods, and the kind of recognition that makes people naturally think of you when something comes to mind.
Do not compete with AI over who can work faster. If someone’s competitive approach is “I can execute standardized tasks more diligently than a machine,” they have chosen the wrong contest from the start. The more interesting question is: what will you accumulate with the time AI saves for you?
What is truly “super” about the super individual is not one person doing the work of ten. If it is only one person doing the work of ten, that is still essentially an improvement in labor efficiency. What interests me is whether one person can use AI to build assets that once only an institution could build. Their own content system, users, IP, products, global distribution capability, digital identity, and perhaps even their own AI agent.
Of course, not everyone needs to become an IP. Assets are not necessarily followers. An engineer can possess technical judgment; a designer, a body of work; a salesperson, long-term client trust; and a manager, an organizational method that has been tested. What truly matters is whether some part of what you do today can enter tomorrow.
I have not fully worked out how far AI will ultimately develop either. But not knowing does not mean doing nothing. Do not wait for the future to become certain before you begin accumulating what is your own. We cannot control how quickly models change, how platforms change, or where industry tailwinds go. What may be somewhat within our control is whether something remains after we finish something today.
If one day everyone uses roughly the same models, everyone has an AI assistant, everyone can write, draw, make videos, and code, and everyone may even have a digital counterpart of their own, how will we distinguish ourselves from one another? Perhaps it will still come down to very old things: What you have experienced. What you believe. What you refuse. The choices you have made. Why others trust you. With whom you have built real relationships. What you have created. And, in the end, what exists because you were here. So when someone now asks me, “Will AI replace ordinary people?” I would rather ask in return: If AI truly keeps getting stronger, what are you preparing to leave for yourself? Perhaps the real risk was never that AI would replace you. It may be looking back one day and realizing: For the past ten years, you have been busy. But nothing truly belonged to you.
Copy this essay for someone thinking about AIGC content, IP, organisation or digital identity.
