Is AI killing creativity?

You can have an idea for a short film without knowing how to animate. You can write a song without knowing how to create the visuals that go with it. Having the idea and having the skills to make it happen are two different things.

That's the use of AI I'm most interested in.

It gives people a way to attempt work they might otherwise leave alone. Someone can explore an idea without first learning every piece of software or mastering every discipline involved.

I'm more cautious about asking AI to come up with the idea, choose the best option and then produce it. At that point, I'd want to ask what decisions I'm actually making.

Creativity includes making things

Cambridge defines creativity as "the ability to produce original, imaginative ideas or to make new things."

We spend a lot of time arguing about whether AI can generate an original idea. The second part of that definition deserves attention too.

Someone without formal training can still have something interesting to say or make. What they lack might be the technical knowledge to express it in the form they imagine.

AI can help with that. It might help a writer build an interactive version of a story, or help someone turn a visual concept into a rough animation.

That doesn't give them the knowledge of an experienced developer or animator. But it could give them a way to explore the idea, see what works and decide whether to take it further.

For me, that is a meaningful expansion of what a person can create.

My concern is handing over the direction

AI regularly helps me find approaches that would have taken longer to discover myself. I don't want to pretend that isn't useful.

But its suggestions influence what I consider next. If I keep asking for adjustments to the same idea, I can spend a long time working within a direction I haven't properly questioned.

There's research behind the concern about similarity.

A 2024 study asked 293 people to write short stories. Some could request AI-generated starting ideas, while others worked without them. Their stories were assessed by 600 independent readers.

Writers offered up to five ideas received average ratings 8.1% higher for novelty and 9% higher for usefulness. However, their stories were also more similar to one another. Individual pieces received better ratings, while the collection became less varied. Read the study.

That doesn't establish that asking AI for ideas is always bad practice. Another experiment found that using deliberately varied AI personas to generate starting ideas could preserve diversity in the resulting stories. Read the research.

But it does give me a reason to be cautious about routinely accepting whatever an AI suggests, particularly if I'm also relying on it to choose which suggestion is best.

An expert in one thing can be a beginner in another

This is where I think AI's potential gets more interesting.

An experienced writer might already be very good at writing. They could still struggle to build a website that presents their work in the way they imagine.

A musician might know exactly what they want a performance to feel like, but have little experience creating animation or interactive visuals.

AI could help with those surrounding tasks. The person brings their existing knowledge and creative intention, while getting assistance with something outside their specialism.

That is different from assuming AI will improve the work they already know how to do.

In a study involving 120 professional designers, introducing AI image generation only during implementation increased completion time without a significant additional improvement in creativity ratings. The designers spent more time making revisions. Read the study.

So adding AI to a skilled person's established process isn't automatically helpful. The study also didn't test whether those designers could benefit from using it for a different task, such as coding an interactive presentation of their work.

The useful question is where the person needs help, and whether the tool actually provides it. See our AI content service for how we combine AI tools with human review.

Making more of your own ideas

I'd approach this by starting with what I want to make, then working out what's stopping me.

The obstacle might be technical knowledge. It might be the time needed to build an early version. It might be that I don't know which tools are suitable.

Those are specific problems I can ask AI to help with.

Explain what matters before asking it to produce anything

Write down the idea in your own words. Include the details you care about and what you want someone to experience.

For a story, that might be a particular feeling or relationship. For a visual piece, it might be the movement, composition or contrast that first interested you.

This is a practical habit I'd recommend. It gives you something to compare the output against, especially when the result looks polished but doesn't feel like what you intended.

Ask for help with the part you can't do

Make the request specific:

"I've written this story and want readers to explore it through an interactive webpage. Explain a few ways I could build that, what each would involve and where I'd need additional help."

You're asking for a route into making something. You can assess the options against an intention you already have.

That still leaves room for suggestions. A technical limitation might lead you towards a better version of the idea. You don't need to treat your first thought as untouchable.

Build a small version and judge it yourself

Before committing to a finished piece, try making enough to see whether the idea works.

Does the movement feel right? Does the interaction add anything? Has something important disappeared during production?

I know from my own use of AI that it's easy to trust an approach when I'm unfamiliar with the field. Getting something concrete in front of me gives me a better basis for questioning it.

There is also evidence that active reflection matters. A workplace study found greater creativity benefits among people who more actively monitored and adjusted their thinking. It didn't test the exact steps above, but it supports taking that involvement seriously. Read the study.

More opportunities to create

I don't think using AI makes someone's work less creative by default. I care about what they've chosen to make and how they've shaped it.

My preference is to use AI to help act on an idea I care about. That could mean learning enough to attempt something unfamiliar, producing an early version or working beyond my existing skills.

There's no guarantee that the result will be original. Starting with a human idea doesn't settle that either.

But giving more people a chance to explore their ideas is something I see value in. I want to keep that possibility while staying involved enough to recognise when the work is becoming something I actually intended to make. For more practical marketing perspectives, read our Go Further newsletter.

Want to explore an idea for your business? Talk to Lunr Launch about your content.