The author retells the (acknowledged-as-possibly-apocryphal) Henry Ford line: “If I had asked people what they wanted, they would have said faster horses.” His framing of why it matters: “Analogy says: people use horses for transportation. Give them better horses. First principles says: what problem is actually being solved? Getting from A to B. What’s the best way to solve that? Maybe not horses at all.” The user isn’t lying or wrong — they’re applying analogy mode honestly and returning the only thing analogy can return: a better version of what already exists.
The deeper point: “The person asking for a faster horse isn’t wrong exactly. They’re just reasoning from what exists. They can’t imagine what doesn’t exist yet.” This is the categorical limit of stated preferences, distinct from the noise/dishonesty failures that Revealed preferences trump stated preferences — track what users do, not what they say addresses. Even an honest, articulate user can only describe within the constraint set they know — so the design question to ask them is not “what do you want?” but “what are you trying to do?” (the underlying problem) and “what stops you?” (the constraint). The shape of the answer determines what mode of reasoning a builder can apply: stated solutions live inside Reasoning by analogy has a ceiling — you can never get beyond what already exists by copying what already exists, stated problems open the door past it.