AI makes your ideas more original

The argument

AI can make your next idea better and make everyone's ideas more alike.

Those two outcomes can happen at the same time.

In a controlled writing experiment published in Science Advances, people given GPT-generated story ideas produced work that independent raters judged more creative, better written and more enjoyable, with the largest benefit among participants who began with lower creativity scores. But the AI-assisted stories were also more similar to one another.

A 2026 meta-analysis across 19 studies found a small but statistically significant homogenization effect in human-AI creative work. A much larger observational study in Nature found a related pattern in science: AI-augmented researchers gained individual output and impact while the collective range of topics studied narrowed.

AI can improve your idea relative to your first draft while moving your idea closer to everyone else's.

Why people believe it

Brainstorming with AI feels expansive. In seconds you get ten directions instead of one.

But the model is generating from shared statistical patterns. If thousands of people ask similar systems similar questions, their expanded option sets can overlap.

What the evidence says

The evidence does not say AI destroys creativity. It says creativity has at least two levels: the quality of an individual's output and the diversity of outputs across people.

AI can improve the first while weakening the second.

That distinction matters in writing, naming, advertising, product ideas, presentations and any situation where being competent is not enough—you also want to be distinct.

The better question

Do not ask only, "Did AI give me better ideas?"

Ask: Did it take me somewhere I would not otherwise have gone, or did it pull me toward a very polished average?

Use AI to widen the search, then deliberately introduce constraints, lived experience and sources the model could not have inferred from the generic prompt.