Re:Google achieves AI 'breakthrough' by beating Go champion
by Inufu on Wednesday January 27, 2016 @01:44PM
(#10982008)
The version of AlphaGo used in the paper beats CrazyStone about 99% of the time. Even if we give CrazyStone 4 stones handicap, AlphaGo still beats it 80% of the time.
[ Parent
| on HN ]
Re:Google achieves AI 'breakthrough' by beating Go champion
by cobaltblue on Wednesday January 27, 2016 @01:47PM
(#10982034)
Very nice. How come you guys didn't enter the UEC Cup? (Edit: rereading this it kind of sounds snarky, not meant that way. Really impressive if you can kick CrazyStone's teeth in...)
Re:Google achieves AI 'breakthrough' by beating Go champion
by thomasahle on Thursday January 28, 2016 @02:18AM
(#10986853)
Is that with 1200CPU's?
Re:Google achieves AI 'breakthrough' by beating Go champion
by mlindner on Thursday January 28, 2016 @04:09AM
(#10987204)
From the paper, no that's with "single machine AlphaGo" which is 48 CPUs and 8 GPUs. Distributed AlphaGo beats Single Machine AlphaGo 77% of the time. (Distributed AlphaGo being 1202 CPUs and 176 GPUs.)
Re:Google achieves AI 'breakthrough' by beating Go champion
by thomasahle on Tuesday February 02, 2016 @05:00AM
(#11018549)
It beats single machine alpha go 77% of the time with four stones handicap. Right?