"I for one welcome our new computer overlords" -- that's a quote from Ken Jennings, the guy who used to be the world's biggest "Jeopardy" egghead until IBM's Watson supercomputer waxed the floor with him and Brad Rutter, his fellow puny human.
He was joking, but not by much. In a piece commissioned by Slate, the winner of 74 consecutive Jeopardy matches (until he met his match in Watson) writes:
[ Also on InfoWorld.com: For a humorous take on the tech industry's shenanigans, subscribe to Robert X. Cringely's Notes from the Underground newsletter. ]
Watson seems to represent a giant leap forward in the field of natural-language processing -- the ability to understand and respond to everyday English, the way Ask Jeeves did (with uneven results) in the dot-com boom. "Jeopardy" clues cover an open domain of human knowledge -- every subject imaginable -- and are full of booby traps for computers: puns, slang, wordplay, oblique allusions. But in just a few years, Watson has learned -- yes, it learns -- to deal with some of the myriad complexities of English. When it sees the word "Blondie," it's very good at figuring out whether "Jeopardy" means the cookie, the comic strip, or the new-wave band.
Yeah, sure, it helps that success in "Jeopardy" depends a great deal on how fast you can press a buzzer with your thumb -- and there is no faster thumb than a "electromagnetic solenoid trigged by a microsecond-precise jolt of current." And yes, we can take some small solace in the fact that Watson occasionally got things very wrong -- confusing the cities of Chicago and Toronto, for example, and "The Elements of Style" with author Dorothy Parker. (The New Atlantis has a detailed analysis of what Watson missed and why -- presumably written by a human.)
But mostly, we're toast, and it doesn't even necessarily take a supercomputer with 2,880 CPUs and a 15-terabyte database to spread the butter. As uber math geek Stephen Wolfram points out, your average search engine will do.
While average "Jeopardy" contestants -- nearly all of them already among our species' best and brightest -- answer correctly 60 percent of the time (and Jennings pulls off an impressive 79 percent), Google doesn't do too badly either: 66 percent of the time the correct answer to a "Jeopardy" question can be found in the first search result on each page. Bing was but 1 percentage point behind (possibly because -- ahem -- it uses Google search results as "signals" for its own algorithms), with Ask a few points behind that.
Wikipedia, created and edited by humans? 29 percent accuracy. Ouch.