It beat him 5 games to zip. Computer Go is (or was) a bigger problem than computer chess.
Okay maybe a might premature to say that, but not by much. An AI developed by Google has beat the European champion 5-0!
Apparently, this has occurred about ten years sooner than anyone had predicted.
To me this seems like a real holy-shit moment. What are some tasks this technology can be turned towards next?
Hopefully it will lead to AI for a driving game where the opponents aren’t complete idiots.
That’s actually pretty interesting - I remember trying Go a looong time ago but never realized it was so intricate and such a complex problem for AI.
That’s amazing! I once had a go game on my computer at a time when the best chess programs could beat anyone below the level of master and, not only did I win with ease, playing on a 10 x 10 board I often wiped the computer out. And I am really what might be called a stonethrower (by analogy with wood pusher), 6 Qu.
Or FPS games where your companions don’t run RIGHT IN FRONT OF YOU during a firefight. (Fallout 4, anyone?)
A year or two ago i remember seeing a go program that could give an amateur a decent game being seen as state of the art. Pretty impressive to now be able to beat a pro. I do note it was the European champion though. I suspect the real masters are in Asia.
Thread in the Game Room on this topic (reported for possible merge).
Ha. I posted this thread about half an hour earlier than that one. I didn’t see it as something for the game room because I was looking for a discussion about what it says about the current state of AI research, rather than what it means for the game of Go. The fact that the AI plays a game is incidental.
Thank goodness, Im tired of hearing how a glorified checkers game is better than Chess. THE WEST WINS!!!
The West still loses - their game got knocked out first!
Guess Randall Munroe needs to update this.
The AI is called AlphaGo? Seriously?
They couldn’t have bought the right to call it GoBot?
Go Go Gadget?
Go Forth and Die?
The article mentions going after “Go’s greatest player ever,” Lee Sedol.
The other article from the Game Room even correctly realized that his surname is Lee.
FYI, match 1 of the challenge match with Lee Sedol begins now. Link.
And the first game between AlphaGo (white) and Lee Sedol (black) is over.
Lee Sedol resigned. AlphaGo has won the first game in the 5-game match.
This is history in the making, and we are being witnesses to it.
Just saw this thread, after replying to another one in MPSIMS on the same subject. I will see if I can get a moderator to merge the two threads.
To sum up: AlphaGo is playing a 5-game match against the world #1 player, Lee Sedol, in Seoul. Today was the first game. Lee played black (moving first) and AlphaGo played with white.
Lee Sedol resigned in the end, handing AlphaGo its first win. A historical event.