Man ties machine on Day 1 of "Jeopardy!" showdown

NEW YORK Tue Feb 15, 2011 12:16am EST

Photo

Party at Coachella

Scenes from the Coachella Music Festival.  Slideshow 

NEW YORK (Reuters) - A man-versus-machine showdown on popular U.S. quiz show "Jeopardy!" ended in a tie on the first day of a three-day challenge, when an IBM computer showed off its knowledge of the Beatles, as well as a few glitches.

The machine, dubbed Watson after International Business Machines Corp's former president Thomas Watson, is the computer and software maker's latest display of its expertise in advanced science.

Watson's ability to understand language and solve problems through complex algorithms makes it even more evolved than Deep Blue, an IBM chess-playing supercomputer that beat world champion Garry Kasparov in 1997.

"Any time you feel the pain, hey -- this guy -- Don't carry the world upon your shoulders," was one clue posed by host Alex Trebek on Monday's show.

"Who is Jude?" answered Watson, following the show's trademark question-as-an-answer style.

It also gave correct responses to clues about the Italian Renaissance artist Michelangelo and U.S. Olympic swimming champion Michael Phelps.

But Watson also made some errors, coming up with the wrong Latin word for "terminal" and repeating another contestant's mistake because it can't interact with other players.

In an earlier practice session, it mistakenly named Beethoven instead of actor and musician Jamie Foxx, prompting contestant Brad Rutter to quip: "I get the two mixed up all the time." Watson didn't laugh, revealing what some say is another flaw -- the lack of a sense of humor.

At the end of Monday's half-hour show, Watson and Rutter, who has in previous appearances won a total of $3.3 million, were tied with $5,000 each.

In third place with $2,000 was Ken Jennings, who won 74 games in a row during the show's 2004-2005 season.

The final showdown airs on Wednesday, with contestants vying for a grand prize of $1 million. IBM plans to donate all of Watson's winnings to charity.

IBM spends around $6 billion a year on research and development. An unspecified part of that goes to what it calls "grand challenges," or big, multiyear science projects such as Watson and Deep Blue.

IBM executives have said Watson's linguistic and analytical abilities may eventually help develop new products in areas such as medical diagnosis.

(Editing by Eric Walsh)

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/
Comments (3)
annihilatorx wrote:
While not being able to use clues about contestant’s wrong answer is a design flaw, asking a machine to having a sense of humour is asking as much as God to create a setinent machine at this right moment.

Until AI is capable of self-conciousness and theory of human mind, humour is not implementable.

Feb 15, 2011 6:23am EST  --  Report as abuse
PeterMelzer wrote:
I am only going to be convinced of the superiority of A.I., when a computer invents something we humans have not dreamed about. Even Watson still lacks a memory of the future. Henry Molaison could attest to that.

Read more about H.M.’s memory here:
http://brainmindinst.blogspot.com/2009/02/theory-of-mind-ii-hms-memory.html

Feb 15, 2011 10:53am EST  --  Report as abuse
WRL wrote:
Computers are beginning to gain the ability to learn from trial and error, which one half of the process of evolution. The other half is random mutation or alteration of the organisms (or programs in this case) involved. If a computer can maintain multiple versions of its programming and those versions can be duplicated with random alterations then pruned down based on effectiveness, then A.I. will be able to improve itself to some degree.

The largest obstacle will likely be giving the computer a purpose and an ability to evaluate its own effectiveness in fulfilling that purpose. Things will get really interesting if and when computers can alter that purpose and the system by which they evaluate their own progress.

Feb 15, 2011 11:44am EST  --  Report as abuse
This discussion is now closed. We welcome comments on our articles for a limited period after their publication.