The way to have ideas is to understand the field really well. That sounds kind of banal, but I haven't seen anyone approach startups this way explicitly. So, I

The way to have ideas is to understand the field really well. That sounds kind of banal, but I haven't seen anyone approach startups this way explicitly. So, I would study history of technology and of money, and especially recent history of computing and the Internet. Then I'd try to simulate it all in my mind, and try to see patterns. and think of all the alternative ways in which it might have happened. What were the good ideas of any given time? Which of them were inevitable? Which of them relied on timing? By what chain of reasoning could they have been foreseen? I don't really know if this would help. It's a big undertaking, and I have other stuff to do now, so I haven't tried it.