AI Won't Replace Creatives.
It'll Replace Mediocre Ones.

Every few months a new tool arrives that can generate an image, a video, a voice, a script, and the same headline runs again: creatives are finished. I want to make the opposite argument. AI isn’t coming for creatives. It’s coming for the gap between an idea and its execution — and that gap was never where the value lived.

For most of the history of this industry, having an idea and seeing it exist were separated by weeks. You imagined a film; then you scheduled a shoot, cast it, lit it, edited it. That distance is collapsing. What used to take a fortnight can now take an afternoon. That is genuinely new. But notice what it changes and what it doesn’t.

"AI compresses the distance between idea and execution. It does not have the idea for you."

Taste is the thing that doesn’t automate

A model can generate a thousand images in the time it takes to describe one. What it cannot do is know which of the thousand is the right one. It cannot feel why a particular frame carries the brand and the other nine hundred and ninety-nine don’t. That judgement — taste — is the entire job. It always was. The execution was just the part that was visible.

This is why the people who are thriving with these tools are the ones who already had strong instincts. AI gives a good creative leverage. It gives a mediocre one a faster way to produce forgettable work. The tool amplifies whatever taste you bring to it, and if you bring none, it amplifies that too.

The compression is the opportunity

When execution is cheap and fast, you can try more ideas. You can prototype a campaign before you commit a budget to it. You can show a client three directions instead of describing one. For a studio built on ideas, this is a gift — it means more of our time goes into the thinking and less into the mechanical work of production.

But it also raises the bar. When everyone can produce a polished-looking asset in minutes, polish stops being a differentiator. Craft alone isn’t enough when craft is a commodity. What separates work now is the idea underneath it — the thing the model can’t generate because it doesn’t know your brand, your audience, or what would make a specific human being care.

• Use AI to try more, not to think less. The win is more shots on goal, not fewer decisions.
• Protect the idea. Spend the time you save on making the concept sharper, not on making more of the same.
• Keep a human in the taste seat. Someone has to choose. That someone is the value.
• Learn the tools deeply. The creative who understands the model out-produces the one who fears it — every time.

The future of this work looks a lot like the past, only faster: an idea, chosen well, made real. The machine handles ‘made real.’ The rest — the part that was always the point — is still ours.