In many domains mapping the problem to a tree search already gives you a superhuman solver or at least a passable solver. Problem mapping is what most of modern

In many domains mapping the problem to a tree search already gives you a superhuman solver or at least a passable solver. Problem mapping is what most of modern AI research is about. That's how the field was redefined in recent years. Just like Vladimir Vapnik says[1], it's becoming more engineering than science. (And sometimes more software alchemy than engineering.) [1] https://www.youtube.com/watch?v=5mvfpSdWsOo https://www.youtube.com/watch?v=5mvfpSdWsOo "Brute Force and Intelligent Paradigms of Learning"