AI won't make everyone a creator. But when an ordinary person's small idea can quickly become a prototype, their curiosity, judgment, and desire to learn may also grow again through collaboration.

AI won't turn everyone into a creator.

For the first time, it simply means an ordinary person's small idea need not die at "I can't code."

Technology can lower the barrier to creating, but it cannot produce the desire to say, "I want to make something," on someone else's behalf.

What's interesting, though, is that when a vague idea really becomes a game, tool, or piece of work for the first time, curiosity, judgment, and the desire to learn may begin growing from that point too.

This may be one of AI's most interesting and most easily overlooked powers:

It changes more than the efficiency with which we finish tasks.

It also changes how a person becomes themselves.

PART 01

The barrier of ability is lower. The barrier of desire remains.

In the past, someone who hadn't systematically studied programming would usually meet a series of hurdles if they wanted to make a small game.

They needed to learn a programming language, understand the development environment, deal with errors, and know how to get the program running.

Many ideas died at "I don't know how" before they had really begun.

Now, if someone can describe what they want, they may be able to use AI coding tools to make a simple prototype of a game, web page, or little tool.

It may not be stable enough, and it may not be ready to launch immediately.

But an idea that once could exist only in their head can first become something that runs, that they can experience, and that they can keep changing.

The importance of this change isn't just how much time it saves.

It begins to change a person's role.

Before, if you wanted to play a game, you generally had to find, buy, and consume a product someone else had created.

Now you can say, "I want a game like this."

You begin deciding what its rules are, how the characters move, what it should look like, which parts are fun, and which aren't.

AI can generate code, but it can't judge for you what is worth keeping.

AI can offer ten options, but it can't decide for you which one is closest to what you really want.

Through repeated choices, rejections, and adjustments, you are no longer only a consumer.

You begin to propose the needs, judge the results, and help create the work.

Image generated by AI

PART 02

Where does the change from consumer to creator happen?

AI changes something else too: the order in which learning happens.

The traditional path often required you to master enough knowledge before you were qualified to begin doing something.

You had to learn syntax, frameworks, and tools first, and wait for the day when you could finally finish a work independently.

But the path of human collaboration with AI may be exactly the reverse.

You can first make a rough prototype, then discover what you need to learn as you revise it.

When the program throws an error, you begin wondering where the error came from.

When AI repeatedly misunderstands what you need, you begin learning to express it more precisely.

When attempt after attempt consumes your usage quota, you begin exploring how to organize context, divide tasks, and reduce unproductive communication.

This knowledge is no longer a textbook unrelated to your own life.

It starts connecting to a specific problem, a work you want to finish, and an actual failure.

A new cycle appears:

A little curiosity leads to a prototype;

A prototype brings real feedback;

Feedback forces you to make a judgment;

Judgment reveals gaps in your knowledge;

To keep going, you begin learning of your own accord;

New abilities, in turn, lead to greater curiosity.

In this cycle, learning is no longer merely preparation for creating.

Learning itself is a byproduct of creation.

Image generated by AI

Of course, this doesn't mean everyone will become a creator as a result.

Some people only want to use AI to look things up, edit copy, or plan a trip. There's nothing wrong with that.

Consuming is not inferior to creating, and a tool has no obligation to remake everyone into an entrepreneur or programmer.

The real question is whether someone still has a little curiosity about the world, and whether there's something they want to try, change, or make with their own hands.

AI cannot manufacture that desire out of nothing.

But it can lower the cost of getting an idea its first piece of feedback.

And inner motivation doesn't necessarily exist only before action.

Sometimes it appears afterward.

Someone may not make their first prototype because they have an intense desire to create.

It may be that the first prototype actually runs, and only then do they realize for the first time:

I can do this too.

PART 03

Can AI go beyond Kant?

I once discussed a question with friends in a writing group: can AI go beyond Kant?

If we state the question more rigorously, the answer may be no.

At the heart of Kant's argument is that human knowledge is constrained by the structures of our own sensibility and cognition. We can know phenomena within experience, rather than directly access the "thing in itself."

AI has not demonstrated that it has crossed that boundary.

It seems more like an extension, within the world of experience, of the scale of information, density of relationships, and speed of exploration that humans can handle.

So, as a metaphor, AI might be thought of as a new cognitive organ humans have grown.

The telescope let people see distances invisible to the naked eye; the microscope brought them to scales their senses had previously been unable to reach.

AI lets people discover relationships in vast amounts of information, generate hypotheses, simulate paths, and quickly turn a vague idea into something that can be tested.

But an organ does not choose a direction for the person who uses it.

AI can extend human abilities. It cannot answer for us why we should create, what is worth creating, or what responsibility we should bear for the consequences.

Desire, judgment, and responsibility still belong to people.

The stronger AI becomes, the more important these things become.

PART 04

An AI ecosystem that truly grows ultimately leads back to people.

The capacity for growth in an AI ecosystem, as I understand it, therefore involves at least three transformations.

The first · Turning technology into a network

A closed system may have powerful models, abundant data, and many features without necessarily forming a true ecosystem.

An ecosystem means that different people, tools, and organizations can connect with one another, and that those connections produce new combinations.

MCP can be seen as an early example of this kind of open connection. It attempts to use an open standard to connect different AI applications to external data, tools, and workflows. When capabilities are no longer enclosed within a single product, technology has a better chance of forming a collaborative network that can keep expanding.

The second · Turning the network into value

Connection in itself is not value.

A network begins to produce value only when its connections start solving real problems: turning an idea into a prototype, bringing an ability into work and daily life, and enabling people to exchange, collaborate, and create together.

The third · Bringing value back to people

A 2025 study by the International Labour Organization found that roughly one quarter of workers worldwide are in occupations that may be affected by generative AI to some degree.

But "affected" doesn't mean those jobs are about to disappear. Because most jobs still require human involvement, tasks and roles are more likely to be reshaped than people directly replaced.

This means the value of an AI ecosystem cannot be measured solely by model parameters, the number of connections, or the proportion of work automated. We also need to ask whether it gives people new capacity to act, enables people who were previously silent to participate in creation, and helps people develop more judgment, learning ability, and agency through using the technology.

An AI ecosystem that can truly grow is not the system with the most resources.

It is a community that can turn technology into a network, the network into value, and then return that value to people.

What it ultimately leaves behind should not be only a more powerful AI.

It should also be a person whose abilities have expanded, whose desire to create has been reawakened, and who is more willing to take responsibility for their choices.

Image generated by AI

So perhaps we don't have to begin by asking whether ordinary people can become creators.

We can start with a smaller question:

Is there something you once wanted to make but gave up on because "I don't know how"?

It might be a little game you want to play, a small tool to solve a problem in daily life, a map, a way of keeping records, or a strange idea that's hard to explain to anyone else.

You don't have to learn everything first, or guarantee that it will succeed.

Let it grow into the roughest possible first version.

Because often, creation doesn't begin once we are ready.

A new version of ourselves slowly begins to grow after the creating has started.

References

International Labour Organization: Generative AI and Jobs: The 2025 Update

Official introduction to the Model Context Protocol

Stanford Encyclopedia of Philosophy: Kant