A phrase mistakenly credited to the fictional character Sherlock Holmes... may now be given a new meaning by the blue boys at IBM.
"Watson", is the next generation of 'Deep' AI computers from IBM, creators of Deep Blue, the first AI computer to successfully beat World Chess Champion Garry Kasparov in 1997.
Watson's task is to compete (and eventually beat) the World's best Jeopardy champions.
Now, most people might be thinking that chess is a more difficult game than Jeopardy, considering that a computer can access all the information that it could need to compete in Jeopardy. True... but access to information isn't the goal of Watson.
It's understanding human language. For those of you who have been using voice-recog apps, you know how frustrating this can be...even through Google's network, it doesn't quite understand simple queries. And usually, only key-word speech is understood well.
Unlike Deep Blue that had a human operator tell it what moves Kasparov made, and in return Deep Blue told it's human operator what moves to make in return, Watson will have to listen, understand, buzz in (yes, buzz-in), and give the correct question (using the proper syntax) in order to be awarded the points.
The fact that Watson has to listen, buzz, and speak is a big step forward for computer evolution since it effectively makes the machine adapt to the human, rather than the other way around like typing, gesturing, clicking, swiping, etc.
Consider what this means in principle. Currently, when we do a search, we are presented with hundreds of thousands of results. In the end, we still have to sort through it all and select the answer that is relevant to our search. Ask.com (or Ask Jeeves for those that reached him) was the first crude attempt at automated language comprehension. Although not much more than a keyword analyzer, it was an important step in computer-human interaction.
In Jeopardy, Watson will have to understand the topic, then the Jeopardy answer, before sorting through all its information, then giving its final response. And take note, even for humans, recognizing context isn't as easy as it looks... hehehehe...