I’ll admit, I’m not the best at Jeopardy. I’d say my favorite week out of the year is Jeopardy Kids Week, which is about the only time when I can answer most of the questions. However, like many of you, I tuned in last week with a little skepticism to see IBM’s creation, Watson, take on the two most decorated Jeopardy players in the game’s history: Ken Jennings and Brad Rutter.

Sure, Watson had terabytes of information (21.6 to be exact) packed away, including several dictionaries, encyclopedias and a downloaded version of Wikipedia, not to mention IMDB and a copy of the Bible, but the real question was how it would do at understanding the subtlety of the Jeopardy clues.

As it turned out, Watson had no problem taking on the Jeopardy champs, scoring $41,413 to Jennings’s $19,200 and Rutter’s $11,200. Even Jennings was impressed by Watson’s intelligence, saying that he had no problem losing to silicon and that Watson had many qualities of a successful Jeopardy contestant: “It’s very smart, very fast, speaks in an uneven monotone, and has never known the touch of a woman.”

However, the challenges for Watson and the typical Jeopardy participant are completely different, which makes me think it might not have been a fair competition.

For the typical participant, the challenge is sheer trivia knowledge. It’s all about knowing which NBA team won the championship in 2008, their first in 22 years (Who are the Boston Celtics?), or what constellation is Latin for “table” (What is Mensa?). This is the essential challenge of any trivia game show. For Watson, on the other hand, the challenge is language recognition and understanding. It’s not that hard for a computer to find the answer to trivia questions, especially with the terabytes of information that Watson had available to him. The tough part is for a computer to be able to take the clue without any help and figure out what it means.

Don’t get me wrong; this is an incredibly difficult task, especially in a game like Jeopardy where clues are very short and contain subtle contextual hints. Once a computer can understand, it can do what it has always been good at: calculating and remembering things quickly.

This combination, however, opens an entirely new set of possible applications. Already, people are saying Watson represents the biggest computing advance of the century, and its technology could have important applications, such as digesting a patient’s medical records and analyzing them for a doctor. Even professors at Northwestern are trying to create computer programs that are able to do mathematical proofs. If a computer is almost able to do my math homework, with Watson’s smarts, what else will they be able to do?

Just as computers and automation replaced factory and other jobs, will Watson ever be able to accomplish the everyday tasks of lawyers, doctors or consultants? Perhaps he will redefine the way we find information, a sort of AskJeeves.com in the real world that actually works. Of course, there will always be differences between what humans and computers can do. But a new age of computer automation might be upon us, and who knows where this can lead us?

As Pradeep Khosla, dean of engineering at Carnegie Mellon, explained, Watson represents a huge advancement in artificial intelligence, but the future of AI lies in this question: “Can Watson decide to create Watson?”

Let’s hope that day never comes.

Ben Rothschild is a Weinberg senior. He can be reached at [email protected].