WATSON ON JEOPARDY


The last two nights on Jeopardy I watched IBM’s supercomputer, Watson, kick the butts of former Jeopardy All-Time winners Ken and Brad. After tonight’s episode, Watson leads Brad (who’s in 2nd Place) by $25,000 and Ken by $30,000. Tonight’s episode is the final meeting in the “IBM Challenge.” Back in the 1990s, IBM’s Deep Blue computer took on Russian Grand Master Garry Kasparov and beat him. After two nights of competition, it looks like history is about to repeat itself. If you want to view this historic final showdown between humans and a very smart machine, check your local listings for tonight’s broadcast time. As Bill Crider might say: “I welcome our new supercomputer AI masters.”

22 thoughts on “WATSON ON JEOPARDY

  1. Deb

    During the show, there were several interviews with the people who had developed & programmed Watson. As we watched, my husband (facetiously?) commented, “Mark down these names so that when the computers take over, we’ll know who to thank.”

    What I’ve found most interesting about the show is watching the array of possible answers that Watson comes up with for each answer. The three top responses and the percentage alloted to each response possibly being correct is shown on the screen even if Watson doesn’t ring in. I was also surprised that Watson missed Final Jeopardy last night–although being so far ahead, it was hardly a problem!

    Reply
    1. george Post author

      I don’t think we have to worry about much, Deb. The fact that Watson came up with “Toronto” in answer to “What U.S. city…” shows we humans still have some wiggle-room.

      Reply
  2. Jeff Meyerson

    Exactly, Bill. All the doors in the studio suddenly lock, the lights go out (except Watson) and the music from “Jaws” starts playing.

    Reply
  3. Patti Abbott

    I didn’t see it but human memory could never compete with computer memory. Their only chance was if a category had never been programmed into it or a malfunction I think.

    Reply
    1. george Post author

      The supercomputer also has to “decode” the JEOPARDY question, Patti. From time to time in the competition, Watson completely misunderstood the question.

      Reply
  4. Drongo

    George, I don’t think we have to worry about a HAL-like scenario anytime soon, but one wonders about the far future. Their speed and power is increasing all the time, and maybe one day they won’t need us at all.

    Of course, if our new robot/computer overlords all look like Summer Glau, life might be better than we think…

    Reply
    1. george Post author

      Roy Kurzweil wrote all about this in THE SINGULARITY IS NEAR, Drongo. I use Kurzweil’s book in my BUSINESS SYSTEMS course. It’s Kurzweil’s contention that super-AI would consider us like goldfish.

      Reply
  5. Richard R.

    Can the thing explain a Jackson Pollack painting, or why Mahler symphonies are so loved, or why people love Tolkien’s writing so much? Of course not. Yes the programing is a big deal, it was quite an accomplishment to program a computer to use it’s logic in this way, but we’re not at the I, Robot stage yet.

    Reply
  6. Deb

    I just watched the final show–and I had to laugh at Ken Jennings writing, “I, for one, welcome our super computer overlords” on his final jeopardy response.

    Reply
  7. Lauren W

    Hi Dr. Kelley!

    Did you get a chance to check out the NOVA special on Watson? After watching it, you really appreciate all of the obstacles they had to overcome in designing Watson. You also end up appreciating the natural capabilities of the human mind (though I wouldn’t mind having Watson’s ability to store information, especially for my Japanese class).

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *