[Link] AlphaGo: Mastering the ancient game of Go with Machine Learning

ESRogs

DeepMind's go AI, called AlphaGo, has beaten the European champion with a score of 5-0. A match against top ranked human, Lee Se-dol, is scheduled for March.

Games are a great testing ground for developing smarter, more flexible algorithms that have the ability to tackle problems in ways similar to humans. Creating programs that are able to play games better than the best humans has a long history

[...]

But one game has thwarted A.I. research thus far: the ancient game of Go.

http://googleresearch.blogspot.com/2016/01/alphago-mastering-ancient-game-of-go.html

DeepMind's go AI, called AlphaGo, has beaten the European champion with a score of 5-0. A match against top ranked human, Lee Se-dol, is scheduled for March.

Games are a great testing ground for developing smarter, more flexible algorithms that have the ability to tackle problems in ways similar to humans. Creating programs that are able to play games better than the best humans has a long history

[...]

But one game has thwarted A.I. research thus far: the ancient game of Go.

http://googleresearch.blogspot.com/2016/01/alphago-mastering-ancient-game-of-go.html

How big a deal is this? What, if anything, does it signal about when we get smarter than human AI?

This is a big deal, and it is another sign that AGI is near.

Intelligence boils down to inference. Go is an interesting case because good play for both humans and bots like AlphaGo requires two specialized types of inference operating over very different timescales:

rapid combinatoric inference over move sequences during a game(planning). AlphaGo uses MCT search for this, whereas the human brain uses a complex network of modules involving the basal ganglia, hippocampus, and PFC.
slow deep inference over a huge amount of experience to develop strong patte

... (read more)

2bogus10y

AI researcher Yoshua Bengio says machines won't become dangerously smart anytime soon. Choice quote: [...]

20bogus10y

It shows that Monte-Carlo tree search meshes remarkably well with neural-network-driven evaluation ("value networks") and decision pruning/policy selection ("policy networks"). This means that if you have a planning task to which MCTS can be usefully applied, and sufficient data to train networks for state-evaluation and policy selection, and substantial computation power (a distributed cluster, in AlphaGo's case), you can significantly improve performance on your task (from "strong amateur" to "human champion" level). It's not an AGI-complete result however, any more than Deep-Blue or TD-gammon were AGI-complete. The "training data" factor is a biggie; we lack this kind of data entirely for things like automated theorem proving, which would otherwise be quite amenable to this 'planning search + complex learned heuristics' approach. In particular, writing provably-correct computer code is a minor variation on automated theorem proving. (Neural networks can already write incorrect code, but this is not good enough if you want a provably Friendly AGI.)

24

[Link] AlphaGo: Mastering the ancient game of Go with Machine Learning

24

24

24

[Link] AlphaGo: Mastering the ancient game of Go with Machine Learning

24

24