Human Dominance in Jeopardy

In 1997 (the same year, incidentally, in which the Terminator series projected that mankind would be wiped out by artificially intelligent machines), reigning world chess champion Gary Kasparov met his match in IBM's Deep Blue.  The supercomputer beat him in six games, with two wins, a loss, and three draws.  Citing unexpected flashes of brilliance from his opponent, Kasparov accused IBM of cheating.  The pride of man, it seems, is not so easily beaten.



But chess is a game of finite proportions.  Raw computing power was bound to outpace the best efforts of human beings eventually.   Kasparov just happened to be the unfortunate bastard who was around at the time to prove it.  And yet, simulated virtuosity at chess is nothing compared to the pattern recognition feats of ordinary people doing ordinary things everyday, making everyday, ordinary distinctions.

IBM intends to encroach on that territory with its new system, "Watson", which is being prepared to compete against human contestants on the popular television quiz show, Jeopardy!  The show's format offers unique challenges for programmers to overcome, with its often obscure trivia categories wrapped up in riddles and puns.  Watson will be operating entirely on its own internal memory, and have less than a second to provide correct responses at least 85% of the time.

Good luck with that.