While it seems to come close to a general AI, AlphaGo is sadly not there yet. While a logical approach (i.e. as used in chess) doesn't work for Go, the deep learning algorithm's purpose is still focused around Go.
Deep learning was also used to beat a professional DOTA2 player, except that was also applied in a very limited setting. It's a slow development but an exciting one!
You are viewing a single comment's thread from: