I've recently been watching a lot of the Y-Combinator YouTube videos. One idea that comes up a lot is that you should listen to the problems that users have but

I've recently been watching a lot of the Y-Combinator YouTube videos. One idea that comes up a lot is that you should listen to the problems that users have but not their solutions. It's an idea along the line of "If I'd asked what they wanted, they would've said 'A faster horse'." In other words, you would take away "Faster transportation" and discard "Faster horse". There's a subtle but important distinction. When you're trying to figure out what to build, and you ask your users, they'll tell you they want a faster horse, but really they want faster transportation. When you're trying to figure out how to build faster transportation, now you've got a very different dilemma: is your daring new idea daring and new because nobody has tried it before due to technological changes, market structures, etc.? Or is it just daring because it's been tried before, and you're about to learn things that experts already knew? There's no easy formula to answering this question.