You can play against Watson online: http://www.nytimes.com/interactive/2010/06/16/magazine/watso...
The flash game is self-contained and Watson's answers are pre-calculated. So it isn't querying Watson in real-time.
IBM's Watson AI trumps humans in "Jeopardy!"
21–30 of 56 posts
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#22In 2006, I was working at a startup that was trying to nail this space. It's amazing how far this sort of tech has come.
What's the market?
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#23How is this possible? I was very impressed but could guess at how it worked with the first several answers/questions listed, but this one blew my mind: “Classic candy bar that’s a female Supreme Court justice” — “What is Baby Ruth Ginsburg?” [Of course, Google now knows the answer :) ] Can someone explain how it can do that? Can it solve cryptics too (way harder than crosswords)? Amazing...
It said in the article that the machine analyzes the category. My speculation is that if it correctly deduces that the category is one of those "answer-hybrid" ones, then it can try to break down the question into two parts, answer each, and cross-reference to find an overlap. In this case it seems pretty reasonable that such a sophisticated machine could break the problem down to "classic candy bar" and "female Supreme Court justice", combined with the knowledge that they share a word this would actually become "straightforward".
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#24"The only way to program a computer to do this type of mathematical reasoning might be to do precisely what Ferrucci doesn’t want to do -- sit down and slowly teach it about the world, one fact at a time."
I'm wondering what the resistance to this is. Each and every one of us has been through exactly that process. Do we really think that we can create a knowledgeable mind out of whole cloth?
It seems what we need to do is teach an artificial-mind everything we know, slowly, once, and then it can teach all the artificial minds everything it knows in the blink of an eye.
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#25"The only way to program a computer to do this type of mathematical reasoning might be to do precisely what Ferrucci doesn’t want to do -- sit down and slowly teach it about the world, one fact at a time." I'm wondering what the resistance to this is. Each and every one of us has been through exactly that process. Do we really think that we can create a knowledgeable mind out of whole cloth? It seems what we need to…
There's a little company over in the UK called True Knowledge http://www.trueknowledge.com/ which is taking this exact approach to Question Answering.
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#26The article is split into 8 pages. Eight. Pages. The worst part is that only one every several sentences had meaningful information. And that is the New York Times. Did someone else had the same thoughts or am I acquiring ADHD?
If you're on Firefox, Auto-pager will make it all better.
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#27Re: IBM's Watson AI trumps humans in "Jeopardy!"
#28How long before we have a tournament with Wolphram Alpha vs. Google vs. Bing vs. Watson, etc?
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#29How is this possible? I was very impressed but could guess at how it worked with the first several answers/questions listed, but this one blew my mind: “Classic candy bar that’s a female Supreme Court justice” — “What is Baby Ruth Ginsburg?” [Of course, Google now knows the answer :) ] Can someone explain how it can do that? Can it solve cryptics too (way harder than crosswords)? Amazing...
As albertni suggests, they have a series of heuristics, some of which do word matching and others which are specialized to common Jeopardy idioms -- such as the before-and-after clue. Each heuristic gives a list of candidate answers and probabilities, and Watson replies with the highest-probability answer, if its certainty is high enough.
How does Watson "know" with high probability to apply the before-and-after heuristic in this instance? Because the category is explicitly "Before and After", assuming they're recycling the clues from the following show:
http://www.j-archive.com/showgame.php?game_id=3258
(I see a wag on a discussion board suggested "Pay Day O'Connor" as an alternate solution, which is awesome. Curiously, Watson had the same solution as the contestant that day; Alex seemed to be expecting "Baby Ruth Bader Ginsburg".)
So, no, Watson probably wouldn't do too well on cryptics, but a similar approach with the right set of heuristics would probably work.
Re: IBM's Watson AI trumps humans in "Jeopardy!"
#30Two days ago, for some reason I was thinking that we would see IBM dead in the following couple of years. You know, SGI, Sun, IBM... It appears that I might be clueless. Maybe they will revamp the company into the default search engine of the internet. Maybe they will own a cloud service for companies to automatic data mining.
I am perplexed why the comment above ("revamp the company into the default search engine") got down voted.
I saw this research as evidence that IBM is quite interested in search. I'm not so sure they want to go head to head with Google or Bing or the others but if this research isn't about search then I don't know what would be.