At least 70% of Y Combinator's top 100 companies were built on ideas the founders never deliberately brainstormed. That number comes directly from YC partner Jared Friedman, who analyzed the portfolio and found that organic ideas — ones that grew from founders' own lives — dramatically outperformed ideas generated by sitting down and trying to think of a startup. Paul Graham's conclusion is blunt: explicitly trying to think of startup ideas is one of the most reliable ways to come up with bad ones. While determination is crucial, identifying the right opportunity is equally important. Many stumble at this stage, unable to pinpoint a worthy target. Graham's framework flips the standard advice entirely. The verb he uses is 'notice,' not 'think up.' Good startup ideas are not invented at a whiteboard; they are recognized in daily life by people whose experience has prepared them to see a gap others walk past. Graham emphasizes 'living in the future' as a key strategy for recognizing opportunities. Find an area where you are genuinely ahead — technically, professionally, or experientially — and look for things that seem missing or broken. The founders of the earliest personal computers weren't guessing at demand; they were power users who felt the absence of accessible machines before the mass market knew to want them. That prior experience is the filter. It's what separates a real problem from what Graham calls a 'sitcom idea' — a concept that sounds plausible in a pitch but has no roots in actual frustration. A common barrier to recognizing opportunities is 'Schlep Blindness,' where founders overlook ideas due to their tedious nature. Stripe is his canonical example. For years, developers complained about how painful it was to accept payments online. The problem was obvious, the demand was massive, but the solution required grinding through banking relationships and regulatory complexity. Most founders looked away. The Collison brothers didn't. Schlep Blindness is why the best ideas are often hiding in plain sight: they're unpleasant to think about, so almost no one does. To effectively notice opportunities, begin with your team's unique strengths and identify problems that align with those skills. Record every frustration you encounter for 21 days — not to manufacture ideas, but to surface patterns in genuine pain. Study existing markets where solutions are weak or absent. Work on ambitious side projects and toys; ideas built from curiosity tend to reveal missing infrastructure naturally. Afterpay, for instance, began as a locally adapted version of a proven overseas model — a noticed gap, not a brainstormed concept. The best ideas, Shiyu, are ones you want yourself, can actually build, and that few others have recognized as worth pursuing. Here is what ties all of this together, and it matters for you specifically. The reason brainstorming fails is that it asks your brain to generate demand from nothing. Organic discovery works because it asks your brain to report what it has already observed. Graham's three-part test for a great startup idea is clean: you want it, you can build it, and most people don't yet see it. Two out of three is common. All three together is rare — and that rarity is exactly what creates the opportunity. The key takeaway is that successful startup ideas are noticed, not invented. Living in the future — being genuinely ahead in some domain — is what positions you to see the obvious thing that everyone else is missing. Stop asking 'what should I build?' Start asking 'what do I wish existed?' That shift in question is not semantic. It is the difference between a made-up problem and a real one, and in the startup world, that difference is everything.