NYT article on Deep Blue, chess, art, & intelligence
There's a good article that Ted Karoglou brought to my attention. You may have to be signed up (for free, I think) to read it though.
http://www.nytimes.com/2003/01/25/arts/25TANK.html?ex=1044517650&ei=1&en=7e31c3f801b8bd44
Comments?
ZMG
- I think creativity and art mean something very different than some AI researchers believe. The article above says something to the effect of... if its artistic it doesn't matter how it was created. I dont quite agree with that. For some fields the functionalist stance is not the correct one. I think chess playing AI using the brute force approach has set back the field significantly by purposely not trying to find intelligent/creative/innovative ways of playing chess. Marty
- I found this website today in the AJC. It is about Kasparov beating the computer Deep Junior in chess. Does this have any effect on the AI community? Did they build a computer that wasn't as smart as Deep Blue or has man simply conquered machine or did he conquer the programer? Search and conquer isn't the only way to solve a problem. The artistic factor must be considered. Here is the chess article http://www.accessatlanta.com/ajc/news/science/0103/27chess.html. Brooke Shepard
I was fortunate enough to see a AAAI conference panel on this subject many years back (I think '97) where the central topic of the discussion was weather AI was making progress because of brute force chess strategy. On the panel were both deep blue affiliates as well as chess champions without computer knowledge or aspirations. Interestingly one of them had said that Kasparov and others have to completely change their strategies against computer programs in chess. Instead of trying to outwit them they play purposely backwards trying to think of the least anticipated move. Of course when using brute force with some lookup moves, taking the least anticipated move is the best strategy because it will require the computer to do the most 'thinking' on the fly and thus fail or not reach a good enough solution fairly often. I guess thats why I think chess playing may not have helped AI greatly. If chess players have to completely change their strategy to compete against computers then programs are doing fundamentally such different things that we understand nothing more about cognition or for that matter intelligent knowledge representation. Marty
It's interesting that the humans adapted which is the one thing this particular computer program could probably not do at all as a strategy. A found another web site that has some background info on Big Blue.
http://whyfiles.org/040chess/index.html
ZMG