If design’s value isn’t execution—and AI is making that argument harder to resist—then what is it? Dan Ramsden offers a framework I find useful.

He breaks thinking into three types: deduction (drawing conclusions from data), induction (building predictions from patterns), and abduction—generating something new. Design’s unique contribution is abductive thinking:

When we use deduction, we discover users dropping off during a registration flow. Induction might tell us why. Abduction would help us imagine new flows to fix it.

Product managers excel at sense-making (aka “Why?”). Engineers build the thing. Design makes the difference—moving from “what is” to “what could be.”

On AI and the temptation to retreat to “creativity” or “taste” as design’s moat, Ramsden is skeptical:

Some might argue that it comes down to “taste”. I don’t think that’s quite right — taste without a rationale is just an opinion. I think designers are describers.

I appreciate that distinction. Taste without rationale is just preference. Design’s value is translating ideas through increasing levels of fidelity—from sketch to prototype to tested solution—validating along the way.

His definition of design in a product context:

Design is a set of structured processes to translate intent into experiments.

That’s a working definition I can use. It positions design not as the source of ideas (those can come from anywhere, including AI), but as the discipline that manages ideas through validation. The value isn’t in generating the concept—it’s in making it real while managing risk.