AI as Creative Partner

Last week I wrote a text. One that didn’t work. I deleted it and started over. Three times. In the end, the text was half as long as planned, said something different from what I’d intended, and was better than anything I’d set out to write.

That’s creativity. Not the result. The path to it. The deleting. The not-knowing. The feeling that something is off without being able to say what. And then the moment it tips. When a sentence is there and you know: That’s it.

AI is increasingly called a creative partner. It recognizes patterns in customer data, generates personalized content, tests variants, optimizes. That gets called creativity.

It’s not creativity. It’s combination.

Creativity begins where combination ends. Combination takes what exists and rearranges it. That’s useful. Sometimes even impressive. But it is something fundamentally different from the moment when someone thinks something that wasn’t there before. When a musician plays a note that is wrong and still right. When a sentence emerges that captures an experience for which there was no word before.

AI can generate a thousand variations of an ad. Each one optimized for click rate, time on page, conversion. Each one tested, measured, improved. The result is output that works. But working is not creativity. Working is the opposite of it.

The interesting thing about creative work is that most of the time it doesn’t work. Most ideas are bad. Most drafts end up in the trash. The process is inefficient, frustrating, slow. That’s exactly what makes it valuable. Because in that failure, something happens that no algorithm can simulate: You understand the problem differently than before. You reach a point where you no longer know what you’re looking for. And then you find something you weren’t looking for.

An algorithm doesn’t search. It computes. It has no sense that something is off. It has no frustration, no doubt, no moment where it deletes everything and starts over because it feels wrong. It has no “feels.”

When people talk about a “creative partner,” they usually mean efficiency. Less effort for more output. Faster variants. More tests. Better numbers. That’s a valid promise. But then call it what it is. Efficiency. Not creativity.

The word gets used because it sounds better. Efficiency sounds like a factory. Creativity sounds like art, like something human. It transfers a quality that isn’t there. And it simultaneously devalues what creativity actually means.

If everything is creativity, nothing is creativity. If an algorithm that recombines text fragments is creative, then what is the person who deletes three drafts because none of them are right?

I’m afraid of the answer that’s taking shape. Not that machines become creative. But that we water down the word until we forget what it once meant. That we mistake optimized output for creativity and regard the inefficient, painful, human process as unnecessary.

AI can do a lot. It can combine, vary, accelerate. It can take work off my hands that bores me and give me time for the work that challenges me. That’s valuable.

But it’s not a creative partner. It’s a tool. A good one. The difference is not academic. It determines what we expect from people in the future and what we don’t.

What do we mean when we say creativity? Who benefits when we stretch the word to cover everything that produces output?