IBM AI wins jeopardy round, badguys ftw, and curbing misbehavior online
IBMs AI "Watson" wins Jeopardy practice round against human champions. This is pretty fucking sweet.
http://www.engadget.com/2011/01/13/ibms-watson-supercomputer-destroys-all-humans-in-jeopardy-pract/
More promo videos of it here and here.
And I was beginning to think their wasn't any decent advances in AI recently. Now we are truly beginning to see advances that could match up to theories and predictions of what is expected of AI advancement.
On another note. How many times have I seen comments along the likes of "instead of Watson, they should call it SKYNET!".
Are you people idiots? Clearly, it should be renamed after the rogue AI (in 2001 space odyssey), "H.A.L." They created the name for that movie, based off IBM... (H->I, A->B, L->M). Come on people, get your computer/cultural references right! silly n00bs....
This is my favorite Cracked.com article to date. Woohoo for bad guys! They say a great movie is defined by a good protagonist.
I was just talking about this with a friend. How online anonymity on the Internet turns people into fuckwads. Blizzard Entertainment was going to try to curb this by making players use their real name on the forums. That never panned out. However, Riot games might have found the right system for their own game. This sounds like a brilliant idea to that problem. It's something that could be tweeked and used in other communities as well. I've always love the little feature on forums where you can "thank" or Digg someones post to show your support for what they said without having to repeat them. But I really like this voluntary jury system. Wikipedia has similar systems in place, why shouldn't it work for other online communities as well?