Googles AI beats European Go Champion 5-0

https://googleblog.blogspot.de/2016/01/alphago-machine-learning-game-go.html

This graphic now needs to be fixed.

probably sorian

tfw you can program a thing to be smarter than you

I recommend this article - http://www.sciencemag.org/news/2016...mimics-human-brain-beats-professional-game-go

this needs to become a meme xD

Well ****. First Arimaa, then Go.

At least we'll always have Calvinball.

So the real question is:

Did it google the best strategy?

>computers may never outplay humans
>seven minutes in heaven

Clearly this graph was made before the advent of VR ( ͡° ͜ʖ ͡°)

I can't tell if it's a song, game, or just a euphemism for sex.

It's a party game played by teenagers, where a guy and girl are thrown into a closet and are locked in for seven minutes, said teenagers then do whatever it is horny teenagers do when thrown into a closet.

Debate politics? compare brand of clothes? Make deep philisophical statements about the fundamental nature of life, the universe and humanity?

Do we know if the Go matches in March will be public?

I'd watch that.

CSGO pro matches are always publice.

Oh wait you meant that GO.

Ayyyyeee. ;D

Yeah, I meant the good one.

OH SNAP

I'd also be very interested to know that. Albeit all I know about Go is that it is:
a) a game with black and white pieces
b) supposed to be very hard for computers, as even great human players in a way have no formal idea what they are really doing to be as great as they are.

GO is actually a pretty great game. I play on occasion whenever my brother is around since he has a GO board and stones and I don't since those things cost a hundred dollars (for the board and pieces each) if you want a decent quality board and stones. It's definitely a lot more than just two colored pieces facing off. It gets pretty intense and the variety is a lot of fun to say the least

;, ((((

Edit: That looks terrible in this font.

This graph is incorrect. I always lose at poker against my computer.

So basically a human playing Go isn't much different from this dog playing Monopoli?

The way humans play is very similar to the way the Go bot plays. We develop abstract evaluation functions and heuristics that we may not be able to articulate, but it helps us decide the best moves. Of course, we are in a sense restricted by our reasonably strategies, but a computer is less so. It can expand randomly on the state space and get more meaningful learning done than a human could.