Watson - television star and supercomputer
Meet Watson – a computer and, in all likelihood, next year’s big TV star.
After all , what’s not to like? Watson – something like the love child of Dev from Slumdog Millionaire and HAL from A Space Odyssey – has been designed in an IBM lab in New York to compete in America’s favourite quiz show, 'Jeopardy!'
If all goes to plan, he’ll make his debut sometime next year, and clash wits with human contestants on the show.
It’s not a new idea. As far back as 1997, an IBM supercomputer called Deep Blue played and beat the world champion Garry Kasparov at chess (though Kasparov claimed the match was unfair, and subsequently drew with a different version of the program).
But Jeopardy is a far bigger challenge for computers. While chess is a game of logic –each piece having clearly defined powers – answering Jeopardy questions requires a subtle understanding of language.
A successful player needs to be able to recognise puns, double entrendres and analogies.
To give you an idea, here's CNNMoney.com's David Goldman, pitting his wits against Watson:
Alright. Ready to go. Finger on the buzzer. Watson picks geography for $400. This should be easy, I'm thinking.
"Eighty percent of Algeria is covered by this desert."
I know this one! I buzz in quickly.
"Player two, Watson."
Huh?
"What is Sahara?" says the computer's simulated voice.
Too fast for me.
Waton’s software needs to interact with human on human terms and, to beat the buzzer, do so fast.
It’s not exactly a thinking computer, but an understanding one. Not connected to the Internet, he will make his answers based on text previously “read”, or processed and indexed.
As a contestant, he has his limits. For one, he is deaf so the quiz show rules will be adapted so that he can receive questions as electronic text while the human contestants will both see the written question and hear it spoken by the host.
But he can speak to give his answers, albeit it with a synthesised voice, and - like countless reality TV stars before him - winning this game show could open up a future of great possibility for him: helping doctors treat patients, or banks assess risk for example.
That’s one small step for a computer, one giant leap for artificial intelligence.
