Comment by Philpax

Well, they're being trained with reinforcement learning now, so I, for one, am excited to see what kinds of unique extremely performant insanity they'll produce in the future =D