(no subject)
This is an article with a conversation with the winner of this year's Loebner prize. That's the prize given out to the AI that "most passes" a Turing test (i.e. the computer that best fools a human in conversation that it is, in fact, human).
His name is George.
George apparently learned by having people talk to him on some website (over 5 million conversations). It amuses me that this is the article's conclusion:
"We have spent decades attempting to design computers to carry out our will politely, with docility. But we have, perhaps, come closest to simulating humanness with a program that gets all huffy when it thinks you've called it stupid, and which doesn't flinch, when the occasion arises, from calling you a dingbat."
His name is George.
George apparently learned by having people talk to him on some website (over 5 million conversations). It amuses me that this is the article's conclusion:
"We have spent decades attempting to design computers to carry out our will politely, with docility. But we have, perhaps, come closest to simulating humanness with a program that gets all huffy when it thinks you've called it stupid, and which doesn't flinch, when the occasion arises, from calling you a dingbat."