IBM's Watson computer plays Jeopardy!

[achali]

SOURCE
http://m.newyorker.com/online/blogs/news...atson.html

A DATE WITH WATSON

Today is Valentine’s Day and, unlike previous February the fourteenths, this year guys who spend too much time on the computer to get dates will have an excellent excuse for staying home. For tonight, Watson, a computer designed by IBM, begins its historic bid to defeat two of the greatest “Jeopardy!” champions ever, Ken Jennings and Brad Rutter, in a three-day, two-match throwdown. It’s the nerd Super Bowl, and unlike the actual Super Bowl the outcome of this contest may influence the course of civilization. (More on that later.)

Watson, as you may have heard, is the result of a four-year research project to build the smartest machine on earth. Deep Blue, the IBM computer that defeated the world chess champion Garry Kasparov in 1997, was smart, but Watson is playing a game that is orders of magnitude more difficult for a machine. Chess has complex rules and logic, but it requires no understanding of language, no knowledge of history, science, music, literature, film, or pop culture. “Jeopardy!” contestants, on any given night, have to contend with thirteen disparate categories of knowledge, from the Baby Jesus to Babe the Pig.

Here are some things you should know about your date: Watson is big, about the size of ten refrigerators. You’ll see him onstage in his air-conditioned glass server chamber, which contains his two thousand eight hundred core processors and fifteen terabytes of RAM. His database holds about two hundred million pages of information, or roughly a million books. He can’t hear; all the questions will be fed to Watson as text, at the same time Alex Trebek reads them to his human opponents. But he does speak, in a voice reminiscent of HAL from “2001”—a deliberate choice by IBM engineers, who do not want to pretend that Watson is human. He won’t be connected to the Internet during the game, and he never has been. He doesn’t worry about getting things right; he’s more concerned about understanding the clues, especially if they involve puns. Wordplay is his weak spot. Another one is that, while “Jeopardy!” contestants will hit the buzzer before they know the answer, and try to pull a correct response out of their posteriors before time sounds, Watson signals only when he knows the response. (Exploit this, Ken and Brad.) Watson is represented by a swirly avatar that looks like the logo representing IBM’s smarter-planet campaign. But watch it carefully, because it will give hints about Watson’s mood—green means he’s confident, orange means he’s struggling. Don’t treat him callously, because he’s not just a search engine: Google wouldn’t qualify as a contestant. Watson could signal a whole new direction in question-based search.

According to the researchers at IBM, there isn’t any one category in which Watson is vulnerable, but here’s hoping “Jeopardy!” ’s Clue Crew will pose a special Valentine’s Day query. For one thousand dollars: “Freud was famously stumped by this question.”

Deafening silence. Smoke curling out of the server cabinet…

The answer: What do women want?

Photograph: Watson competes against Jennings and Rutter in a practice match, on January 13th.




SOURCE
http://m.newyorker.com/online/blogs/news...-mean.html

WHAT WOULD WATSON’S VICTORY MEAN?

Barring a miracle, it looks like Watson will defeat the two human “Jeopardy!” champions tomorrow night and claim the crown of world’s biggest nerd. Going into the contest, I had been telling myself that I was for the computer, because a win for Watson was a win for humanity. But last night, when Watson started kicking serious ass, I began to feel queasy. Not for any Skynet-became-self-aware type of reason, but because the sight of Brad Rutter and Ken Jennings being sacrificed to a machine for the greater glory of I.B.M. was a sorry spectacle. My twelve-year-old son couldn’t bear to watch, and went downstairs, no doubt to use the computer.

All that remains is to figure out what Watson’s victory means. That’s one question Watson can’t answer, so it falls to us, the vanquished, to grapple with it. Is Watson a great breakthrough in science—a Sputnik moment—or an elaborate parlor trick? When I asked Steven Pinker this question, he responded:

I don’t rule out the possibility that some components of Watson could both provide insight into human cognition and lay the groundwork for more sophisticated artificial intelligence applications, such as natural language processing (the fancy term for understanding human languages like English, as opposed to computer languages). On the other hand, when a system is designed to meet a highly specific challenge like playing Jeopardy, and one where the reputations of the designers are on the line, there will be enormous pressure to tailor the system to succeeding at that challenge by any means whatsoever, including kludges that are specific to the rather peculiar requirements of the game of Jeopardy.
Pinker went on:

The real problem is that we may never know. It will depend on whether the I.B.M. team divulges the methods in technical publications or keeps them as trade secrets. In the golden years of A.I. (1960s and 1970s), there was a lot of back-and-forth between academia and industry. Labs at Xerox, I.B.M., B.B. & N., and A.T. & T. were among the best research departments in the world, and people and ideas flowed in and out of them. Then A.T. & T. lost the free money from its telephone monopoly, and the other companies realized that their openness was just helping their competitors (e.g., the Macintosh GUI, which was basically stolen from Xerox PARC), and they forced their scientists to work on applied projects and kept the details out of the public domain. The result is that A.I. has become disengaged from cognitive science, the old A.I./philosophy gurus like Minsky, Papert, Simon, and Schank have not been replaced, and questions like yours may be impossible to answer, if I.B.M., as seems likely, will keep the specs secret. That is, we won’t be able to know how much of the program’s success to attribute to humanlike or superhuman intelligence, and how much to Jeopardy-specific hacks.

I.B.M. spent, by some reports, more than a billion dollars to build Watson. The project has been a brilliant branding campaign, but are there business applications that would allow I.B.M. to recoup its investment? Watson is not going to be the next Google. Watson can understand nuances of language and syntax, unlike Google. But Google, though commercial and cluttered, is far more in tune with what humans want (including sex) than Watson is.

In some ways, Watson is a throwback to I.B.M.’s beginnings in computing: a large, stand-alone mainframe that can be programmed for specific tasks. There is talk of a “Watson M.D.”: a remote terminal that will be installed in doctors’ offices and can help to diagnose patients’ illnesses and prevent future ones. There’s also discussion of a Watson-enabled device for your car that could identify what’s causing that rattle more accurately than an unfamiliar mechanic or the hosts of “Car Talk.” Darren Hayes, a professor of computer science at Pace University who specializes in computer forensics, told me that he foresees applications for Watson in intelligence-gathering, specifically in improving the quality of analysis at the seventy-five “fusion centers” that the Department of Homeland Security set up around the country after 9/11.

I do hope Watson finds something useful to do, and does not become, like so many other instant TV celebrities, future fodder for a “Where are they now?” segment. Otherwise, I.B.M. just bought three very expensive half-hour infomercials.