This coming February television and technology are going to collide in a bizarre, yet totally awesome way. IBM’s Watson, a JEOPARDY playing supercomputer, is going to take on the game’s two greatest champions, Ken Jennings and Brad Rutter, over three shows airing February 14th, 15th, and 16th. It’s man versus machine in the biggest thing to happen to the venerable quiz show since Neil Patrick Harris.
But the question is, “Does Watson stand a chance?”
Those who follow IBM’s drive to create the next Skynet know they have entire teams devoted to making computers think better. In the 90s, IBM’s Deep Blue lost to chess legend Gary Kasparov in 1996. A year later, however, Deep Blue took Gary in a rematch to became the first computer to ever defeat a real-life grand champion.
Watson’s development hasn’t been a secret, however few predicted IBM would be able to make it compete with living, smart-as-a-whip humanoids. After all, chess is one thing, but JEOPARDY is a completely different beast.
I won’t say creating a computer to beat a chess king like Kasparov is an easy thing, because it isn’t. There’s an art to the game that is difficult to capture in binary. However, chess (as a game) can be calculated mathematically, and it doesn’t need to interpret anything but the pieces position on the board. That lends itself to being played by a computer; strategic models can be programmed and a ridiculous number of outcomes can be calculated nearly every move.
JEOPARDY is a game that relies on being able to think like a human being, something that doesn’t lend itself to binary processes. One must be able to interpret the answer, deriving clues from pop culture references and riddles, then applying that information to some vast amount of obtained knowledge. If you were to ask a computer to tell you the caloric content of Skippy chunky peanut butter, it would probably be able to tell you straight away. Telling it “cereal lovers know it’s the high-in-fiber outer casing of the oat,” isn’t much for a computer to go on.
Furthermore, like a person, Watson has to calculate probabilities and take a guess if it’s right before it chooses to buzz in. When was the last time your computer said, “I think I may be right, but I’m not sure enough to risk it?”
Alas, IBM has done it. Apparently, Watson is good enough that it’s regularly beating really smart people in Jeopardy. So, they’re putting their money where their mouth is and are taking on the greatest JEOPARDY players in history. The goal, of course, isn’t to win or get in the entertainment business. It’s to show that IBM is ushering in a new era of artificial intelligence; one with computers that can interpret, analyze, and draw conclusions in natural language. The applications for such a technology are immense. Whether Watson wins or not, no doubt businesses of all types (medical, diagnostics, legal, military, etc) will be pounding on IBM’s doors with checkbooks in hand.
I highly recommend you tune in this February, because you’ll have a chance to witness a groundbreaking moment in human history. If Watson wins, man will have created an artificial intelligence arguably more capable than any natural human intelligence.
Just think about that.
Want to learn more about IBM’s Watson? Check out their DeepQA site for videos and stuff.
Satisfy your inner geek while fueling your TV addiction… TV Tech Fix is a column by Matt Whitlock, editor of the TechLore.com Consumer Electronics Community (plus several other gadget-focused community websites), and lover of both technology and TV. In this column, he’ll cover a wide variety of tech topics aimed squarely at the TV addicts of the world – from tips and tricks to help you better your TV experience, to gear recommendations, to the impact technology is having on the TV shows we love.