This seemingly insignificant event — a computer program besting a human at a board game — could well turn out to represent a historical breakthrough for AI (Artificial Intelligence) and "neural networks":
A go playing program — AlphaGo
, developed by a Go(ogle)
team of programmers at MindGame
, London — has beaten the European go champion, China-born Fan Hui
, with a shut-out 5-0 result! In even game
matches (no handicap stones)!
A-stone-ishi-ng (go fans may get the triple pun)!
Having played go since 1972 (thanks, +Bert Rubaszkin
), I have always thought this could NEVER be done. As in Not Ever
(well, I've been wrong before). For those of you familiar with the ancient game of Ba Duk (Wei Chi, Igo, Go), what do you think? Is this mind-boggling news, or what?
One of the best go players in the world, Korean 이세돌 (Yi Sae Tol), 9-Dan pro, has accepted a challenge to play a series of five games, slated for March this year. I'll be there! And I'm willing to wager that AlphaGo will
be beaten. The European Champion is a strong player, but 이세돌 선생님 (Master Yi Sae Tol) is definitely playing on a different level altogether! Here are (some of) his very
impressive achievements, to date:http://senseis.xmp.net/?YiSeTolhttps://en.wikipedia.org/wiki/Lee_Se-dol
And here's a brilliant game record (Yi Se Tol
9P vs. Lee Chang Ho
9P). In Korean, no subtitles, but the moves speak loudly for those of you that know go (in Korea,the name of the game is "Ba Duk,"
which roughly means "hand talk"