Google DeepMind's AlphaGo takes on Go champion Lee Sedol in AI milestone in Seou

[horace08122]

and AlphaGo won the 1st match 

will it be another 5-0?

[horace08122]

AlphaGo got the second match....maybe really 5-0

btw...no one cares?

[Gleb Gamow]

https://en.wikipedia.org/wiki/Go_(game)

Upon reading, I see how that would be a difficult game to program unlike chess, checkers, etc.

[Batmanvsduecare]

Asian people do have many board games hmmm.

[Fortify]

https://en.wikipedia.org/wiki/Go_(game)

Upon reading, I see how that would be a difficult game to program unlike chess, checkers, etc.

Yeah it's kind of interesting that there are more possible moves in the game of Go than there are molecules on earth. A lot of the champions are just taking educated guesses and using gut feeling - unlikely to work against a virtual supercomputer mind.

[nihilnegativum]

5:0 incoming

http://www.wired.com/2016/03/googles-ai-wins-pivotal-game-two-match-go-grandmaster/

[nihilnegativum]

Its 3:0 already
http://www.theverge.com/2016/3/12/11210650/alphago-deepmind-go-match-3-result

[nihilnegativum]

 The comeback is real 3-1


http://www.theverge.com/2016/3/13/11184328/alphago-deepmind-go-match-4-result?utm_campaign=theverge&utm_content=chorus&utm_medium=social&utm_source=twitter

[swogerino]

The comeback is real 3-1


http://www.theverge.com/2016/3/13/11184328/alphago-deepmind-go-match-4-result?utm_campaign=theverge&utm_content=chorus&utm_medium=social&utm_source=twitter

But even though they're playing these extra matches, it's really impressive how Google's AI was able to win all three first games! It just goes to show how quickly our technology is growing.

[nihilnegativum]

The comeback is real 3-1


http://www.theverge.com/2016/3/13/11184328/alphago-deepmind-go-match-4-result?utm_campaign=theverge&utm_content=chorus&utm_medium=social&utm_source=twitter

But even though they're playing these extra matches, it's really impressive how Google's AI was able to win all three first games! It just goes to show how quickly our technology is growing.

Yes, it shows that neural networks enchanced by the ability to learn from humans, can achieve specific goals of complex pattern recognition really fast. Its really bizzare how it uses human knowledge and applies its own type of artificial abstract reasoning in a way very different from the human abstraction. In the fourth game, Alpha made a mistake, the interesting question is whether its mistake was a result of its own type of abstraction or a result of the data it was fed and how it will learn from it.

[fuathan]

It was the third...

[nihilnegativum]

Fourth, I edited,
there is a lot on the wiki about the match : https://en.wikipedia.org/wiki/AlphaGo_versus_Lee_Sedol

[horace08122]

last game today, will it be 3-2

[Gleb Gamow]

last game today, will it be 3-2

Why the frown? Actually, two wins at this stage would be a great accomplishment, setting the stage for a rematch.

[Indrawan77]

I think its amazing that machine can win againts human, its human that created machine and yet its winning againts human

[Gleb Gamow]

I think its amazing that machine can win againts human, its human that created machine and yet its winning againts human

"Hey, Chuck, after seeing AlphaGo beat Lee, I think it's amazing that a machine can win against a human, given that machines are made by humans. What do you think?"

[nihilnegativum]

If they can beat us at go, imagine what neural-network enchanced wood-chippers could do...

[Gleb Gamow]

Breaking news! The match may be nullified due to an after match medical examine uncovering that AlphaGo was tested positive for electroids. AlphaGo tweeted the following upon given word:

00100000 01010000 01110010 01101111 01101001 01101110 00100000 01101110 01110101 01101100 01101100 01100001 00100000 01101100 01100101 01100011 01110100 01110101 01110011 00100000 01100101 01110101 00100000 01110000 01101111 01110010 01110100 01100001 00100000 01110101 01110100 00101110

[onlinedragon]

AlphaGo got an self learning system so would it be impossible to beat it in the future.

[mikehersh2]

I had done some research on this myself after my programming teacher in school had showed us what it had done. They say for each move, it went through 5,000 scenarios on what to do next! And as the match went on, the robot learned the play style of its human opponent, and would get progressively better. This is impressive, yet scary at the same time.