The Fidelity Trap

In design, a good rule of thumb is to match the fidelity of an artifact to the certainty of the designer. That's why early ideation should happen on the figurative back of the napkin, tentative prototypes in a wireframe tool and only the final design in a high-fidelity tool. This prevents spending too much time on early ideas that are likely to be reworked or discarded. More importantly, it prevents the psychological lock-in that happens once too much time was sunk into it.

I was reminded of this today during a roadmapping session with a client who had used AI to build an interactive prototype. In general, yes, a picture says more than a thousand words, and if the picture is interactive, it communicates even more so. But it also came baked in with a lot of assumptions that we invalidated over the course of our session, and I had to work extra hard to overcome an internal bias for assuming that polished-looking features were definitely must-haves.

At one point, I even asked "Okay and so then what should this particular button here do in the functional app" and the response was: "Oh I think Claude hallucinated that bit of functionality; please disregard." There's no way that would happen if you'd use pen and paper, or a low-fidelity tool (just toss things around in Canva, Miro, Balsamiq and the likes).

High fidelity used to take effort. Now it doesn't, but the lock-in it creates can trap you.