000%
TUNING INTUITION TO VELOCITY
← All insights

Why human intuition still decides what is worth making

2 min read · November 2026

Ask an AI model for a logo and it will give you twenty. Ask it which one is right, and it will give you an opinion dressed up as a fact. That's the gap we build around: AI is extraordinary at generating options and terrible at knowing which one matters. Deciding what's worth making is still, stubbornly, a human job.

Taste is a judgement, not an output

A good creative decision isn't the statistically most popular one — it's the one that's true to the brief, the audience and the moment. That kind of judgement comes from having made things before, got them wrong, and learned what 'wrong' actually feels like. A model trained on everything that's already been made can't tell you what hasn't been made yet, or why it should be. That's still ours.

Where the machine actually helps

None of this makes AI useless on a project — it makes it useful in a specific place. We use it to widen the research before a strategy call, to generate variations once the direction is set, and to handle the repetitive production work that used to eat the week: resizing, retouch passes, first-draft copy. It buys back time. It doesn't spend the time for us.

What this looks like on a real brief

On a branding project, that might mean: AI pulls competitor and category references overnight; a strategist reads them, throws out the noise, and writes the one line of positioning that actually holds. On a web build, AI drafts the component variations; a designer picks the one that respects the hierarchy the brief asked for, not the one that looks cleverest in isolation. The output is faster. The decision-making isn't delegated.

The brief it still can't write

Every project starts with a brief, and a brief is an argument for what matters and what doesn't. That argument has to come from understanding a client's business, their audience and the risk they're actually taking by hiring us — not from pattern-matching against a training set. So the brief gets written by a person, the options get widened by AI, and the final call comes back to a person again. That loop is the whole method, and it's why intuition still runs the studio, not the other way round.