Two Jeopardy! champions contend with AI computer

AddThis

Students and faculty watched a historic "Jeopardy!" match Monday between two human champions and a machine made possible, in part, by the research of three UT professors.

Watson, a deep question-and-answer computer, competed against "Jeopardy!" champions Ken Jennings and Brad Rutter. Jennings has the longest winning streak at 74 games in the show’s history, and Rutter won more money than any previous contestant at $3.25 million.

At the end of the first show, Watson was tied for first place with Rutter. The match will continue through Wednesday.

UT alumnus James Fan, an IBM researcher and one of the Watson developers, spoke at the watch party to answer audience questions during the show’s breaks. He was one of the 25 researchers who worked on the project since 2007 and one of the first researchers to test Watson.

“I’m actually one of the early contestants or test subjects for Watson,” Fan said. “We were stay-in "Jeopardy!" players. We knew about 40 percent of the trivia.”

Fan said that there are four challenges that computers have in the "Jeopardy!" game: the wide range of topics, the difficulty of language, time constraints and confidence.

IBM Corp. recognized three UT computer science faculty members for their research that helped with the development of Watson. Computer science department chair Bruce Porter, who helped with the research, said natural language is one of the most difficult concepts for computers.

“The challenge in natural language processing is ambiguity. Every word in English has multiple meanings. Every sentence has many interpretations, and Watson has to figure out of all these possible meanings,” Porter said.

Research scientist Ken Barker said IBM made a great contribution through the help of many projects and research from all over the world in one system.

“There’s a lot of research going on in the world right now on artificial intelligence,” Barker said. “Putting Watson on "Jeopardy!" is a good example of applying that research to a task that captures the public’s imagination.”

Electrical engineering freshman Andrew Wiley said he was aware of artificial intelligence before Watson, but now it is more public.

“Suddenly we have Watson,” Wiley said. “It’s like taking Google to a whole new level. You’re taking a search engine, but now you’re able to link knowledge.”

Aerospace engineering senior Nathan Knerr said he was amazed by the children who attended the watch party and asked fan questions about Watson.

“More importantly than what it actually does or does not do is how all the 6- to 8-year-olds in the audience will someday make this all work in its more general form so we can all use,” Knerr said. “It’s ultimately perhaps the greatest factor that’s put into it, is getting [kids] interested in this type of thing."