There are probably millions of things people have imagined but never made because they lacked the skills to bring them to life.

A game, an interactive story, a strange digital world, a website that feels like nowhere else on the internet. Ideas that never made it past the imagination because creating them required skills their creators didn’t have.

AI is changing that.

I’ve been experiencing this myself, using AI to build things I might never have built otherwise. And it got me thinking that something bigger than software development might be changing: software is becoming a medium through which more people can express themselves.

Not everything needs to be useful

There’s a difference between building software to solve a problem and creating something because you want it to exist.

The first is about utility. You have a need, and software is a way to satisfy it.

The second is harder to explain. You might have an idea for a game, a strange interactive experience, a personal website, or some little digital thing that doesn’t have an obvious purpose. You make it because you’re curious about it, because it reflects something you’re interested in, or simply because you want to see what happens.

It doesn’t need to become a business. It doesn’t need to attract users. It doesn’t even need to be useful to anyone else.

Sometimes you just want to make something.

We’ve always done this in other forms. People write stories nobody reads, take photographs they never publish, play music in their bedrooms, or make things for their homes that nobody else will ever see. We don’t need to justify every creative act through its practical value.

Software has traditionally been different. Making even a small interactive experience required a fairly specific set of technical skills. If you wanted to build a game, you needed to learn programming, game engines, and probably a bunch of other things. Having an idea wasn’t enough. You also needed to know how to implement it.

AI is changing that equation.

The gap between an idea and its expression

There have always been people with ideas they couldn’t bring to life because they lacked the skills to execute them.

Someone might imagine a beautiful painting but not know how to paint. Someone might hear a melody in their head but not know how to play an instrument. Someone might imagine an interactive world but have no idea how to build it.

We tend to accept these limitations as part of the creative process. Different mediums require different skills, and most people can’t master all of them.

But what if the skills required to use a medium become less of a barrier?

AI doesn’t eliminate the need for creativity, judgment, or taste. You still need to decide what you want to make, guide the process, evaluate the result, and figure out what’s missing. But you no longer have to personally master every technical discipline involved in bringing an idea to life.

You can focus more on the thing you want to express and less on the mechanics required to express it.

I think this is particularly interesting with software because it’s such a flexible medium. You can create images, sounds, stories, games, simulations, tools, and interactive worlds. You can build things that respond to people, change over time, or behave in ways that would be impossible in a static work.

And you don’t need to fit into any established category. You can make something that doesn’t have a name yet.

You don’t have to be an artist

I don’t think the interesting outcome is that AI will turn everyone into an artist.

Some people will deliberately use software to create art. Others will simply make things that express who they are, what they care about, or how they see the world, without thinking of themselves as artists at all.

That’s already how people interact with other creative mediums. Some people paint because they want to be painters. Others draw, decorate, write, or make music simply because they enjoy it.

The same impulse could find its way into software.

Someone might build a tiny game inspired by a childhood memory. Someone else might create an interactive visualization of an idea they’re obsessed with. Another person might make a personal website that feels more like a place than a collection of pages.

None of these things needs to qualify as art. They just need to be things someone wanted to create.

And maybe some of them will be art, whether or not their creators intended them to be.

A new kind of expression online

The internet made publishing accessible to almost everyone. You no longer needed a publisher to distribute your writing or a television network to share a video. You could create something and put it online yourself.

But publishing and creating are different things.

What if AI makes it similarly accessible to create interactive experiences? Instead of sharing only text, images, and videos, people could share little worlds, games, simulations, and experiments that others can actually interact with.

Imagine browsing the internet and finding something that someone made in an afternoon because they had an idea they wanted to explore. Not a startup, not a polished commercial product, not content optimized for engagement. Just a small, personal creation.

Maybe you play with it for five minutes, send it to a friend, or make something of your own in response.

I don’t know what this kind of internet would look like, or whether it will become a meaningful part of how people express themselves online. But I like the possibility.

We’ve spent years making it easier to consume and distribute things. Perhaps we’re entering a period where it’s also becoming easier to create entirely new kinds of things.

What do we make when we can make almost anything?

Of course, making things easier doesn’t automatically make them more interesting. AI can generate an enormous amount of generic output, and lowering the cost of creation doesn’t guarantee that the result will have much meaning.

There’s another thing I find interesting about this. Even if AI does most of the actual work, the result can still be deeply personal.

You can have a very specific idea of what you want to make, care about every little detail, reject a hundred versions until something feels right, and keep iterating until the result matches what you had in your head. AI might write every line of code, generate every asset, and handle most of the implementation, but it doesn’t mean the work is no longer yours.

The creative process doesn’t have to be about personally producing every piece. It can be about knowing what you want, recognizing when something is wrong, and guiding the work toward a vision that only you had in the first place.

Maybe that’s another thing AI changes: not just who gets to create, but what it means to be the person creating something.

I’m more interested in what people might create when the barrier between imagining something and bringing it into existence becomes smaller.

How many ideas never became anything simply because the person who had them couldn’t execute them? How many people have never considered themselves creative because they didn’t have the skills to use the mediums available to them? What kinds of things might they make if those constraints changed?

I don’t know the answers. I’m experiencing this shift myself, and it’s one of the reasons I’ve been enjoying building things outside the usual boundaries of software development.

Maybe the most interesting consequence of AI isn’t that we’ll build the same software with fewer people, or build it faster.

Maybe it’s that we’ll start making things that wouldn’t have existed otherwise, for no particular reason other than that someone wanted to make them.

And I think that’s worth being excited about.