Hacker News new | ask | show | jobs
by axelfreeman 3748 days ago
I don't get the mystery of this. This algorithm is complex. SURE! But deep learning is very fast training / repeatition of a game (or some other goal) while saving the good or bad results. Predict user moves. Find good positions/patterns. Or did i miss some here?

https://web.archive.org/web/20160128151110/https://storage.g...

1 comments

Yeah...AlphaGo has played more games in the last few months than a human will in their life time.

I'd be interested in how strong it would be if given the same constraints as human learning (playing thousands of games, rather than millions).