Check out MuZero. It learns an embedding of the game's state space potentially allowing AlphaGo like dominance in more ways. https://www.youtube.com/watch?v=We

Check out MuZero. It learns an embedding of the game's state space potentially allowing AlphaGo like dominance in more ways. https://www.youtube.com/watch?v=We20YSAJZSE https://www.youtube.com/watch?v=We20YSAJZSE https://arxiv.org/abs/1911.08265 https://arxiv.org/abs/1911.08265 (Actually, ML is still not very good at causal reasoning so we have sometime. I'm more excited and worried about crispr at this point; what happens when we can make people genetically super human)