AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Challenges of Machine Learning in Go and Starcraft
In go, there's only 400 actions, but one action is what people would call the god action. So i think the problem there is, in terms of action space size is way harder. For humans, maybe the playing stark ofic seems more intuitive because it looks real aminon like the graphics and everything moves smoothly. Whereas i don't know how to min go is a game that i wouldn't really need to study. But for machines, kind of, maybe it's the reverse, yes,. which shows you the gap, actually, between deep learning or machine learning. And however, the hekor brains work a so you developed a lot of really interesting ideas.