IBM Watson wins Jeopardy!

Background

*Jeopardy!* is an American television quiz show in which contestants are read a clue and must reply in the form of a question. As a challenge for a machine, the show differs from chess: the clues are written in puns, wordplay and oblique references rather than drawn from a fixed set of rules, and the reply has to be phrased correctly as well as found. IBM says it had been searching for a new human-versus-machine challenge since its chess machine Deep Blue defeated Kasparov in 1997 ^[1]. The quiz show was the challenge it settled on.

What happened

In February 2011, in a televised contest viewed by millions, Watson defeated the show's two foremost all-time champions, Brad Rutter and Ken Jennings ^[1]. The test was not a single game: Watson was put to the test in two *Jeopardy!* matches played over three days ^[1].

Key details

The money told part of the story. Watson won USD 77,147, which was donated to various charities; Ken Jennings won USD 24,000 and Brad Rutter won USD 21,600 ^[1].

The most memorable moment came in the first match. Watson missed the Final Jeopardy! clue in the US Cities category, whose correct response was "What is Chicago?" ^[1]. Instead, Watson answered "What is Toronto?????" — with five question marks, which IBM says indicated a substantial lack of confidence ^[1]. The response made the limits of the system visible in a way a correct answer never would: the machine had not simply been wrong, it had registered its own uncertainty and still had to commit to something.

The blunder did not pass quietly inside IBM either. It prompted an IBM engineer to wear a Toronto Blue Jays jacket to the recording of the second match ^[1].

On the board, the contest stayed close for a while. Watson entered the second match in a tie with Brad Rutter before pulling into a commanding lead ^[1].

Impact

Ken Jennings, one of the two champions beaten, commented that "Quiz show contestant" may be the first job made redundant by Watson, but he was sure it would not be the last ^[1]. The line captured how the match was received: not as an isolated stunt but as a public demonstration that a machine could take on a task built around ordinary language, cultural reference and quick recall — and win against the best humans the format had produced. The prize money going to charity also framed the contest as a showcase rather than a commercial product launch ^[1].

Within IBM's own account of its history, the match is the second landmark in a line that starts with Deep Blue in 1997 ^[1]. Where the chess machine had worked inside a closed world of rules, the quiz show required the machine to deal with clues written for people, and to answer them in front of an audience that could see every mistake. The Toronto answer, complete with its five question marks, became the part of the contest people remembered most — a reminder that the same system could look dominant in one round and openly unsure in the next.