#computing #ai See [[Philosophy/Rationality/Models/Exploration vs exploitation]]. Basically the same, how much [[Artificial intelligence]] agent should use it's current learned model, how much it should try new states?