|Sent on:||Tuesday, August 30, 2011 10:19 PM|
I went through chapter two quickly last week. Rereading it more carefully this week, I can see that it has foundational concepts in it that though straightforward, aren't advisable to gloss through. I added some more flash cards for some of these concepts. Welcome new member Kendall, you can find the groups wiki here: http://freedombluesky.com/aiml
Reading this book reminds me a lot of working through math books, one has to go slowly and think carefully and exactly about each sentence to get the real meaning in the big picture. This book is a bit trickier though, in that the math like concepts instead of being in math symbols, are in English, which makes it tempting to go faster than is advisable.
Yahoo had a front page video story today on two Loebner Turing test contest AI bots talking with each other:
-the story makes the mistake of saying the bots are from Cornell, they are from http://cleverbot.com
The bot uses the pure statistical approach explained in chapter 1 of the AI book. I think it's strange that a pure statistical bot instead of a hybrid would be a front runner, and that that means the opportunity for some unknown programmer to go down in history as the creator of the first Turing test winner is still wide open. The grand prize winner also gets $100k and an Olympic sized solid gold medal.