No, the unsolved problem of AI is continuous learning. We never stop learning, we don't have a "training phase". You are always updating your world model even when you sleep. Also more quality training data does lead to greater learning efficiency as you have more priors to work with.