Someone fed ChatGPT Paul Graham’s essays and asked how Paul Graham would answer this.
His real answer is to look at the frontiers of knowledge, and to look for anomalies, gaps that might have been missed. Knowledge grows irregularly. When you look from far, you can’t see it, but you get close, those gaps are obvious. Why wasn’t this or that tried.
His answer is based on: “How do you get new ideas that are on the frontier of knowledge?” This assumes that all knowledge growth happens in the newly developing fields. This must be based on his experience in the developing startup tech. But there must be opportunities for new ideas all around us. Dyson redesigned the hoover.
My answer would be: to practice. To spend time ideating, and brainstorming. To spend time learning new things. To mix and match different fields. New ideas often come from mixing two things together. To mix new technologies with old one. To travel.