Ideas as hypotheses
A startup idea is a bundle of problem, customer, channel, and pricing guesses. Generated prompts break symmetry so you can articulate what you would test first on a landing page or concierge MVP.
Problem-first framing
Strong ventures nail a painful job-to-be-done. If the generator gives you a clever gadget without suffering, rewrite until someone would pay to make the pain stop.
Moat reality check
Network effects, regulatory licenses, and proprietary data beat feature lists. Ask what still works if an incumbent copies UI in a quarter.
Idea journals
Save outputs with date stamps. Revisit quarterly—timing often kills good concepts that were early, not wrong.