In a move reminiscent of their "Deep Blue" chess match-ups with Garry Kasparov, IBM has once again created a machine meant to beat our puny human brains. This time around, it's named "Watson" and it aims to take on the most popular question and answer show on television - Jeopardy.
According to IBM, the human-machine contest will take place February 14-16, 2011 and will pit Watson against two of the game's greatest biological players, Ken Jennings and Brad Rutter. Rutter has won the most of any Jeopardy contestant - over 3.3 million dollars, and Jennings won 74 matches straight in 2004-2005, leading to a change in the rules about consecutive wins.
Both players have the wits, speed and skill to answer questions about everything from the obvious to the obtuse, so why does IBM think its Watson has a chance?
In what was ultimately an unsatisfying highlight of his career, chess Grand Master Garry Kasparov was beaten in 1997 by another IBM creation named "Deeper Blue", though Kasparov had beaten the original machine, Deep Blue, in 1996. This new 2010 iteration of computer power is named after company founder Thomas J. Watson, rather than the fictional Conan Doyle character, and the machine has some big and successful shoes to fill.
This time around, IBM has built a machine that they think will have the wherewithal to compete with some of the best human minds in the business, and they might just be right. Though the game might seem skewed to give the computer a higher chance of victory than either human, there are a few points to bear in mind. First, the array of potential questions posed on Jeopardy cover a vast amount of human knowledge, and it will take Watson some time to access the correct subsection of its storage space for any given query. In addition, many of Jeopardy's questions rely on wordplay, puns, or outright misdirection, all things that human players will be on the lookout for but that Watson might take literally, like an anti-social friend that just doesn't understand sarcasm.
Even if the matches are a total blowout for man or machine, the experiment is in an interesting one. If we humans take the day, it will mean that tech companies have a long way to go before robotic bulters can answer our doors and mechanical nannies can care for our children. Should Watson prevail, however, it may signal that we're finally getting close to understanding just how these big brains of ours work.
While we'd love to say we're impartial on this one, we're hoping the computer comes out as more the bumbling doctor than malicious Moriarty when this all goes down.