It's important to remember that this is an accomplishment of humanity, not a defeat. By constructing this AI, we are simply creating another tool for advancing our state of being.
(or something like that)
AlphaGo beats Lee Sedol 3-0 [video]
31–40 of 428 posts
Re: AlphaGo beats Lee Sedol 3-0 [video]
#32I don't think Ke Jie would win against Alpha Go either.
Re: AlphaGo beats Lee Sedol 3-0 [video]
#33What happens if you give the human player a handicap? I wonder if the games are really as close as the commentators say, or if it's just a quirk of the MCST algorithm.
Re: AlphaGo beats Lee Sedol 3-0 [video]
#34Earlier quoted context omitted.
> as he played a bunch of his moves with only 1 or 2 seconds left on the clock. That's on purpose to make full use of time and think about the next move
As a spectator, I was on the edge of my seat :) As a 3d amateur, I'm really curious about when he resigned. It really seemed like he was playing the position out to go for the win (or perhaps to see how AlphaGo would fare in ko). It didn't look like he was searching for a place to resign.
The AGA commentator (Cho Hyeyeon 9p, at https://www.youtube.com/c/usgoweb/) had already been calling the game completely hopeless for Lee Sedol for nearly 2 hours at that point, estimating Lee down by at least 20–30 points before komi, unless by some miracle he could win the ko fight.
Re: AlphaGo beats Lee Sedol 3-0 [video]
#35Super interesting to watch this unfold. So what game should AI tackle next? I've heard imperfect information games are harder for AI...would the AlphaGo approach not work well for these?
Hassabis discusses some StarCraft here ... http://www.theverge.com/2016/3/10/11192774/demis-hassabis-in... ... good interview.
Re: AlphaGo beats Lee Sedol 3-0 [video]
#36I really like the moments when Alpha-Go would play a move and the commentators would look stunned and go silent for a 1-2 seconds. "That was an unexpected move", they would say.
In game 2 there was a point where Michael Redmond seemed to do a triple take and couldn't believe the move AlphaGo played.
Re: AlphaGo beats Lee Sedol 3-0 [video]
#37I am really curious about the reviews from An Youngil 8p and Myungwan Kim 9p. The commentary by Redmond always tend to leave something to be desired.
Myungwan Kim 9p and Cho Hyeyeon 9p's commentary on the American Go Association YouTube channel:
Game 1: https://www.youtube.com/watch?v=6ZugVil2v4w
Re: AlphaGo beats Lee Sedol 3-0 [video]
#38I really want to see how a team of humans would do against alpha-go with a 3 or 4 hour time limit.
Yeah, if AlphaGO wins all 5, I'd next like to see some Blitz Go (if that's a thing).
Re: AlphaGo beats Lee Sedol 3-0 [video]
#39Super interesting to watch this unfold. So what game should AI tackle next? I've heard imperfect information games are harder for AI...would the AlphaGo approach not work well for these?
One perfect-information game that's at least as hard: constructive mathematics. (Proof assistants even give it a sort of videogame-style UI.) I've been wondering about some kind of neural net for ranking the 'moves' coupled with the usual proof search.
Re: AlphaGo beats Lee Sedol 3-0 [video]
#40Earlier quoted context omitted.
As a spectator, I was on the edge of my seat :) As a 3d amateur, I'm really curious about when he resigned. It really seemed like he was playing the position out to go for the win (or perhaps to see how AlphaGo would fare in ko). It didn't look like he was searching for a place to resign.
The AGA commentator (Cho Hyeyeon 9p, at https://www.youtube.com/c/usgoweb/ ) had already been calling the game completely hopeless for Lee Sedol for nearly 2 hours at that point, estimating Lee down by at least 20–30 points before komi, unless by some miracle he could win the ko fight.
Right, all about one or two wrong moves by AlphaGo. He still had a chance around the time Lee started to play the bottom of the board. But AlphaGo did not fall for the trick. I believed the English commentator said Lee could build a ladder but wasn't sure if he meant to say whether that technique would defeat AlphaGo or not.