ABSTRACT
128Engineers at IBM have developed a computer named Watson that is capable of understanding and answering questions that are posed in natural language. The ultimate test came for Watson when it was a contestant on Jeopardy, competing against the two biggest winners in the game show's history: Ken Jennings and Brad Rutter. Jeopardy is a game that requires not only a broad base of knowledge about a variety of topics, but also an ability to untangle sometimes convoluted and backwards statements. Although accessing unlimited amounts of information should be easy for a computer, deciphering the cryptic questions and mastering the art of answering in the form of a question should be more difficult. Despite the inherent challenges of the game, Watson easily beat its competitors, bringing in over $77,000 in winnings compared to Jennings' $24,000 and Rutter's $21,000. Despite its intellectual prowess, Watson was not perfect. In Final Jeopardy the category was U.S. Cities and the clue was “Its largest airport is named for a WWII hero, its second largest for a WWII battle.” Watson answered, “What is Toronto,” which is not a U.S. city and its main airport was named after the winner of a Nobel Peace Prize, not a WWII hero. (The correct answer is “Chicago.”) Although Watson is a great specimen of artificial intelligence, scientists still have a lot of work to do to develop a computer that can replicate the complexity, flexibility, and nuance of verbal communication in interpersonal interactions.
