• In what ways is Watson “thinking” like a human being, and in what ways is it not? “This facial wear made Israel’s Moshe Dayan instantly recognizable worldwide.”

Such was the $1,600 clue in the category

“The eyes have it” of a mock round of the popular television game show, Jeopardy .

The contestants had mere moments to think about the question that appeared in writing in front of them—as long as it took the host to read the words aloud. As soon as he was done, it was time for the contestants to compete to be the fi rst to buzz in and give an answer.

The three contestants quickly performed these mental gymnastics to come up with the correct answer, but only one hit the buzzer fi rst. “What is an eye patch?” he correctly replied, adding $1,600 to his pot and handily beating his opponents—

opponents who happened to be former Jeopardy champions. But this competitor was special in his own right: he wasn’t even human. He was a computer named Watson, and he was winning (Baker, 2011).

This wasn’t the fi rst time human intelligence was pitted against artifi cial intelligence and lost. More than a decade earlier another computer named Deep Blue had beaten the reigning world champion of chess at a chess match. But while Deep Blue’s accomplishment was impressive, it was limited in key ways—it didn’t have to understand language, recognize subtle clues, or search through giant information databases to ferret out relevant facts and put them together to make a solution. But Watson did.