This is exactly why it's a hard question to answer. Innovation is not always "new problems to solve", but more often "new clever ways to solve a problem", whic

This is exactly why it's a hard question to answer. Innovation is not always "new problems to solve", but more often "new clever ways to solve a problem", which rules out number 2 IMHO. For example, Kubernetes was an innovation in the cluster management field where many had already solved that problem with other means, and many still do today. When you're trying to innovate, it's easy to fall in number 1 (being blind to the competition: current means to solve the problem).