Millennium Checkmate

And yet as we still attempt to train these soulless arrangements of electrical charges to be all-the-human-they-can-be we are met with limited success. Take "Deep Blue," an IBM super computer programmed to excel at one specific task: chess. Many suggested that computers were far too predicable to beat humans at chess. Yet during a match between Deep Blue and world chess champion Garry Kasparov the computer made a move that Kasparov described as "eerily human."

In another field that has been evolving for half a century, computational linguistics, researchers are attempting to solve perhaps the greatest problem of all time. Teaching computers to translate from one language into another with superlative accuracy. The current models in use are far from flawless. It seems for computers to adequately translate, say Shakespeare's sonnets, while retaining their true essence, a challenge for even a human translator, computers would have to understand Shakespeare's message. But will machines ever perceive poetry? Can they ever perceive poetry?

Perhaps in a hundred years hence computers will read this very essay to one another and mock it while rolling in laughter. Still many critics, including this one, feel that day will never arrive. Never-- much like those critics that suggested machines could never defeat a human chess champion after "Deep Blue" lost its first major match against Kasparov in 1996. The same critics that just one year later, on May 11, 1997, watched in awe and shock as Garry Kasparov conceded 19 moves into the final and deciding game of a six-game rematch. Alas, "Big Blue" was victorious. Checkmate!

Go To Page: 1 2