AlphaGo Summer Experiments
Posted: Thu Mar 10, 2016 1:04 pm
The video of Demis Hassabis presenting at Oxford University some two weeks ago is most interesting.
http://livestream.com/oxuni/StracheyLec ... isHassabis
In the Q&A he was asked "AlphaGo now has an evaluation function for Go, is there any way of giving that back to Go players?"
The reply was "We'd love to do that" "Actually, after the match we have a whole series of expirements planned" "What representations
has our system built and do they match the intuition of top Go players?"
The timing of this question is at 1:02:40
Roll on summer!
Perhaps the AlphaGo Robot will walk into our club, buy us all a drink and then offer teaching games with feedback targeted at the appropriate level.
http://livestream.com/oxuni/StracheyLec ... isHassabis
In the Q&A he was asked "AlphaGo now has an evaluation function for Go, is there any way of giving that back to Go players?"
The reply was "We'd love to do that" "Actually, after the match we have a whole series of expirements planned" "What representations
has our system built and do they match the intuition of top Go players?"
The timing of this question is at 1:02:40
Roll on summer!
Perhaps the AlphaGo Robot will walk into our club, buy us all a drink and then offer teaching games with feedback targeted at the appropriate level.