Watson Wins: Computer Beats Human Trivia Champs

Two of the top "Jeopardy!" champions of all time, Ken Jennings, who won 74 consecutive games, and Brad Rutter, who won more than $3 million in prize money, competed against Watson in a three-day challenge testing the contestants' knowledge of history, science, geography, sports and literature.

The final score?

Watson: $77,147
Ken Jennings: $24,000
Brad Rutter: $21,600

A computer that can outplay humans

	In 1997, a computer named "Deep Blue" beat chess legend Gary Kasparov at the game.

In 1997 a computer named “Deep Blue” beat world chess champion Garry Kasparov. Skeptics downplayed the win because chess is essentially complex math problems.

So IBM researchers took on the grand challenge of developing a computing system that can master the complex natural human language on "Jeopardy!" which clues include jokes, puns, analogies and slang.

More than 200 IBM researchers throughout the world worked for four years to develop the super computer. Watson is named after Thomas J. Watson, who served as IBM’s CEO for 38 years. On television, it was represented by a "smarter plant" avatar with color-coded levels of confidence.

Programmers taught the computer to "get" human jokes

	One of the hardest parts of programming Watson was making the computer able to understand nuances in language like jokes and puns.

The programmers had to write dozens of formulas and algorithms to sort through the nuances of language, looking for patterns and learning from mistakes.

"There's not going to be one algorithm that just understands language," explained the head of the project, David Ferrucci. "It's going to be a lot of different algorithms. They're going to look at and interpret the language from different perspectives. And somehow, we're going to be able to combine them."
 
Once Watson understood the question, it searched through memory banks filled with the entire "World Book Encyclopedia," Wikipedia, the Internet Movie Database, much of the New York Times archive and the Bible.

Can Watson help doctors diagnose disease?

	Next, Watson will head to Columbia University to see if its computing power can help the health care industry.

Watson is not perfect and does make mistakes. When a final "Jeopardy!" question asked for the name of a U.S. city, Watson answered Toronto, a city in Canada. But each wrong answer helps programmers refine the formulas that lead Watson to the correct conclusion.

Watson's next challenge takes it to Columbia University Medical Center to explore and develop improvements in the health care industry.

Watson’s technology has the potential to develop formulas that could identify patterns on diagnostic images that radiologist can’t see with the human eye. Researchers also think Watson might be able to help financial advisors, government agencies, transportation workers and retail services.

Artificial intelligence: good or bad?

	Science fiction movies often exaggerate the dangers that artificial intelligence, like supercomputers and robots, could present for mankind.

Of course, in many science fiction movies smart computers don't always have humankind's best interests in mind; think "The Matrix" or "2001: A Space Odyssey."

"Certainly, artificial intelligence can be destructive. It's already used in our weapons. We have smart weapons that have their own intelligence," says futurist and author Ray Kurzweil.

"Technology can be destructive, particularly in the wrong hands. The positive side is that these tools can help overcome human suffering, help cure disease, alleviate poverty, solve the energy problem, clean up the environment."