• http://blog.agrawals.org rakeshlobster

    Great analysis of the underlying technologies, Danny.

    I did a question-by-question comparison using the first two nights.

    Even with just keywords, Google came up with the correct response in the first position 56% of the time. If you include the first 10 results, Google had the correct response 79% of the time. Watson had the correct answer 79% of the time.

    Both struggled on similar types of questions.

    See the breakdown here:

  • wilkit

    “The kind of search we do on a search engine today is much more keyword oriented and this is way beyond that … If we can search with intelligence, it could open up all sorts of new fields and possibilities.”

    The IBM guys I saw on a panel before the show gave a similar outlook on search engines and that was the most lasting impression I got from them after three days of seeing them. I suppose they were trying to upsell their own technology, but starting by marginalizing perhaps one of the most impressive algorithms we have yet to achieve really turned me off to Watson. The whole time they were presenting I couldn’t shake the thought that “You guys made a complex relational database of text that finds an answer in 3 seconds, Google searches the universe, EVERYTHING, and delivers 87m results in 0.13s”.

    I couldn’t get over how confident they were that Watson completely changed the game. People were even asking questions about how intelligent it was or how accurate it could analyze historical data and predict a future outcome. Maybe I’m getting too worked up over marketing speak.

  • http://2helixtech.com matthiaswh

    Thanks for your coverage of this!

    I was really disappointed when the IBMers marginalized search. The way they talked about it you would think Google and Bing and Blekko just run a database query for the keywords you enter and randomly show you everything that’s a match. I realize the whole point of this tournament was a marketing ploy, but there was no need for the proverbial “trash talking.”

    Scrutinize Google all you want for some of the dirty things it’s done, or for being too powerful, or letting slide too much spam but one thing not even its competitors should do is say that its search technology is not an amazing feat.

    Enough ranting! I enjoyed the first two days of the tournament, had to miss the third. I honestly hoped that it would have been a closer competition, even if Watson won in the end. It sort of left the feeling in my mind that sure, Watson is a marvel of technology, but all they did with the Jeopardy tournament was prove that the computer is faster on the buzzer. (Duh)

  • Thembelani Kwetane

    Wow, why are you so defensive? Watson is way more interesting and advanced than any search engine including Google, more so given the spam Google search returns.
    Google returns results in less than a second because there is more computer power behind it than Watson has access to. But so what if Google is faster if it returns junk 90% of the time? I have given up on searching sometimes because I got tired of filtering the junk Google returns as results.
    Watson is 100X better than anything Google has or have shown. Live with it.

  • http://setandbma.wordpress.com setandbma

    There is an interesting TED talk featuring Steven Baker (author of Final Jeopardy!), Kerrie Holley (IBM Fellow looking for Watson’s next job), Dr. Herbert Chase (Columbia University Professor of Clinical Medicine) and Dr. David Ferrucci (IBM Watson Principal Investigator).

    What surprised me most was what Kerrie Holley said. The use of Watson suggested by him does not have any relation to the core technology. He is talking about analytics, route optimization etc – but Watson is about language comprehension.