Main idea

The user described in a framing workshop and the one who works on the ground are rarely the same: the second has constraints, habits, shortcuts, workarounds, a mental load and resistances the first does not have, because the first was built by the people who design.

Producing the screen faster does not bring these two figures closer. A bad answer to a real problem stays a bad answer when it is generated in a few minutes; what speed changes is only the moment you discover it was bad — and the number of wrong variants produced in the meantime.

What separates two teams is therefore not the speed of making but the quality of what they know: how people actually work, what they tolerate, what they reject, what they do not dare say, and what changes with the role, the sector, the digital maturity or the size of the company.

Layer added by "PM, Developers, and AI: Roles Are Blurring, Responsibilities Remain" (2026-08-01). The same observation holds for the prototype, and sets its condition of usefulness: a prototype produced in a few hours is worth something only if it tests a real customer signal. Otherwise it merely accelerates a wrong direction — and the speed at which it was produced then serves as an argument for committing further to it.

Why it matters

This prevents reading fast interface generation as a reduction of product risk. It reduces the cost of the attempt, which is useful; it does not reduce the gap between the assumed user and the real user, which is the risk in question.

It also names a counterpart to fund when making speeds up: if you multiply proposals without multiplying opportunities to observe the field, you increase the production of unverified answers.

Nuances and limits

Producing fast can serve understanding when the mock-up is used as an instrument of observation — shown early, at the workstation, to provoke a reaction. The gain then comes from the contact, not from the generation.

And some failures of understanding show only in prolonged use: no rate of iteration reveals them within the week.

Open questions

• What frequency of field observation has to be maintained so that faster making does not widen the gap with the real user?