It became fashionable, briefly, to argue that AI would commoditise creativity. The opposite has happened. AI has commoditised production, which has made the part of the work that production used to disguise — the actual creative thinking — more valuable, not less.
What AI commoditised, and what it didn’t
AI is now excellent at execution: making a moodboard, generating variations, producing a clean mockup. These were the tasks that, for decades, were the entry point to creative careers and the bulk of the work on any project. They are now near-free.
What AI has not touched: deciding which direction is the right one, knowing why a design feels wrong even when it follows every rule, and inventing something that has no precedent in training data. Those are creative judgement — and they have become the only things worth paying for.
Three things humans still do better
Taste — the ability to look at fifty options and confidently choose the one with the longest half-life. Restraint — the discipline to remove four-fifths of the work to make the rest sing. Originality — generating an idea that genuinely had not been made before, not a remix of existing patterns.
AI averages the past. Humans imagine the future. Those are different jobs.
Why this matters commercially
When every competitor has access to the same AI tools, output converges. Brands start to look alike, websites feel interchangeable, products blur into each other. The only durable differentiator becomes the human judgement applied on top of the tools.
That is why studios that lean entirely on AI ship work that is fast and forgettable, and studios that pair AI with strong human direction ship work that is fast and distinct. The same tools, two very different outcomes.
How to invest in human creativity now
- Hire for taste, not technique — technique is automatable, taste is not.
- Protect quiet time — original thinking does not happen in Slack threads.
- Pay for senior judgement — the cost of bad direction has gone up, not down.
- Encourage strong opinions — averageness is the failure mode of the AI era.
- Treat reviews as creative reviews, not approval gates.
The misread of the moment
The temptation right now is to assume AI has lowered the value of creative work. It hasn’t. It has separated the part of creative work that scales (execution) from the part that decides (judgement) — and bid up the price of the second one. The agencies and in-house teams who understand this will be the ones doing the most interesting work of the next decade.
