AlphaGo vs Lee Sedol

[Moliere]

Google DeepMind challenge match

Our first match was a historic moment. AlphaGo beat Lee Sedol, the first computer program to ever beat a 9 dan professional player.

Anyone following the match or play the game?

[Jeff V]

What game?

[GreenGoo]

What game?

Go.

[Jeff V]

Where?

[LordMortis]

I used to play and I used to suck but I enjoyed.

I generally played against a guy who gave me a 9 stone advantage and still won and he often played against a guy who was ranked in the US and that guy gave my friend a 9 stone advantage and still won. So that ought to give you a clue as to how bad I was.

As I mentioned on FB in regard to this, I never even got to the point where I could see a ladder coming until I was already a victim and I never could set one up. I imagine that the first lesson you really have to learn and I simply couldn't see it until, "Oh, there it is, now that I'm guaranteed to lose three pieces."

[Moliere]

Tutorial

[Zaxxon]

Anyone following the match or play the game?

Have played, but not seriously. Following the match and fascinated by the implications for AI. Go is much, much more complex from an AI perspective than Chess. It's not computationally possible to brute force the game, nor will it ever be. And yet Deep Mind won the second game today. This is much earlier than folks had originally anticipated AI overtaking the best humans. (Granted, the set of matches is not yet over and Lee Sedol could still come back.)

For some context: there are more possible positions in Go than there are atoms in the universe.

Why Google's Go win is such a big deal, To test a powerful computer, play an ancient game, from ye olde wayback machine.

[Anonymous Bosch]

I used to play and I used to suck but I enjoyed.

I generally played against a guy who gave me a 9 stone advantage and still won and he often played against a guy who was ranked in the US and that guy gave my friend a 9 stone advantage and still won. So that ought to give you a clue as to how bad I was.

As I mentioned on FB in regard to this, I never even got to the point where I could see a ladder coming until I was already a victim and I never could set one up. I imagine that the first lesson you really have to learn and I simply couldn't see it until, "Oh, there it is, now that I'm guaranteed to lose three pieces."

FWIW, when I first learned how to play, I found The Interactive Way To Go to be a helpful tutorial and introduction to the basics of the game (including learning the ladder). Another excellent resource is Sensei’s Library – Pages for Beginners, which contains an abundance of useful info for new players.

[Daehawk]

Played it as a child.

[tjg_marantz]

I used to play and I used to suck but I enjoyed.

I generally played against a guy who gave me a 9 stone advantage and still won and he often played against a guy who was ranked in the US and that guy gave my friend a 9 stone advantage and still won. So that ought to give you a clue as to how bad I was.

As I mentioned on FB in regard to this, I never even got to the point where I could see a ladder coming until I was already a victim and I never could set one up. I imagine that the first lesson you really have to learn and I simply couldn't see it until, "Oh, there it is, now that I'm guaranteed to lose three pieces."

FWIW, when I first learned how to play, I found The Interactive Way To Go to be a helpful tutorial and introduction to the basics of the game (including learning the ladder). Another excellent resource is Sensei’s Library – Pages for Beginners, which contains an abundance of useful info for new players.

That was a great tutorial! Thanks

[Moliere]

Match score: AlphaGo 3 - Lee Sedol 1.

[tjg_marantz]

4-1

[wonderpug]

AlphaGo sounds really good at strategic decisions. We should upgrade all our stealth bombers with AlphaGo and make them fully unmanned. I bet they'd fly with perfect operational records.

[Moliere]

The Genius of Go Grandmaster Lee Se-dol Made AlphaGo's Performance That Much Greater

AlphaGo is more like an artificial intelligence created to win the game of Go, rather than a program that runs and plays Go. AlphaGo contains algorithmic knowledge that allows it to beat virtually anyone, no matter how skillful they are, and so long as its computing capacity is maximized. If AlphaGo faces one of the best players, it undergoes an intensive computing process to find the best move possible, and it will eventually find it. In other words, AlphaGo plays better against a stronger opponent.

[Pyperkub]

How are you going to run the universe if you can't answer a few unsolvable problems, huh? Come on, big fella, let's see what you got.

[Moliere]

Hovertext: The really scary thing will be when the computer can beat us at designing a computer that can beat us at designing a computer that can beat us at designing a computer to play chess.

[Daehawk]

Think it was called Reversi when I played.

[Isgrimnur]

Think it was called Reversi when I played.

Reversi is trademarked as Othello by Mattel. Different game.

[Moliere]

Another loss for the humans.

Google's AI star, AlphaGo, wins again. It bested Ke Jie, the world's best Go player, by just half a point -- the closest margin possible. After the match, Google's DeepMind CEO Demis Hassabis explained that this was how AlphaGo was programmed: to maximise its winning chances, rather than the winning margin. This latest iteration of the AI player, nicknamed Master, apparently uses 10 times less computational power than its predecessor that beat Lee Sedol, working from a single PC connected to Google's cloud server.

[DD*]

Where?

To hell in a handbasket, most likely.