What Is Watson The Computer Doing Now? The 'Jeopardy!' Champ's Been Unlucky
In 2006, perhaps the world's greatest mind was born in New York. Well, not exactly born — more like developed. Watson the Computer, otherwise known as the opponent who took down Ken Jennings, was created by researchers at IBM eight years ago, and is widely considered the greatest achievement of artificial intelligence to date. Since thenn, other than winning Jeopardy!, Watson has received little other press or attention, and seems to have fallen by the wayside. Which is pity, considering Watson's truly superhuman abilities to effectively answer essentially any question posed to it.
In the years since its initial introduction to the public, Watson has found little popularity or application, a testament to the idea that being too smart is, unfortunately, not exactly the way to people's hearts. But with increasingly creative ideas as to how to apply Watson's genius, perhaps the Supercomputer has its best and brightest days ahead. Let's take a look back at how Watson got its start, and what lies in store.
What, who, how is Watson?
A very good question that deserves a much more technical answer than this: Watson is a genius supercomputer created in 2006 by IBM in response to Ken Jennings' unbelievable winning streak on the beloved quiz show Jeopardy! Whereas Google and other search engines could pull up pages of information in response to a query, Watson would have to do better — it would have to not only understand a question, but also be able to answer it, all in a matter of seconds in order to beat its human competitors.
And boy, did it ever. In a competition with Ken Jennings and fellow Jeopardy! superstar Brad Rutter, Watson walked away the champion by a margin of more than $53,000, proving the ultimate dominance of machine over man.
Is this the first supercomputer IBM has created?
Far from it. In fact, Watson came about as a result of an ongoing series of IBM challenges known as "Grand Challenges." With these projects, the company intentionally tested the limits of modern technology, pitting artificial intelligence against human intelligence. IBM's first major victory for a mind of metal came in 1997, when their supercomputer Deep Blue defeated grand master Garry Kasparov in a game of chess.
While this earned IBM considerable publicity, having a chess-whiz supercomputer wasn't exactly a scalable solution to any real world problem. After all, you can't end world hunger with a check-mate. As such, they sought a more marketable solution to a real-world problem with their next "Grand Challenge," which led to the creation of Watson.
OK, how does this thing work?
Well, here it is in IBM's terms — in order to play Jeopardy!, Watson employed "more than 100 different techniques...to analyze natural language, identify sources, find and generate hypotheses, find and score evidence, and merge and rank hypotheses." Essentially, Watson stores a huge amount of information and then parses through it at a rate equivalent to reading a million books per second. Because of its incredible rate of processing, Watson is able to come up with several potential answers, and then evaluate their validity based on the amount of evidence that supports each hypothesis.
Even more impressive is Watson's ability to "learn." Each wrong answer it gives informs the machine what not to do in the future, and as such, Watson effectively never gives the same incorrect answer twice.
Can I get one of these as a laptop?
Nope, sorry. When Watson played against Jennings and Rutter, it wasn't even in the same room as the other contestants, mostly because Watson, at that time, took up an entire room. Comprised of 750 servers, Watson had to be kept separate not only because of its enormous size, but also because of the noise it generated from its cooling system. Today, Watson has shrunk considerably, and is now approximately the size of "three stacked pizza boxes," or as the Watson business unit's vice president Manoj Saxena says, "the size of the vegetable drawer in your double-drawer refrigerator."
Of course, this still won't allow you to lug one around, as even this new and improved Watson weighs about 100 pounds.
So why haven't I heard more about Watson if it's so impressive?
This is a question that IBM likely wants answered as well. Despite Watson's incredible question-answering capabilities, the machine has found little traction in any industry, including healthcare and finance. Though Watson was tested at M.D. Anderson Cancer Center and Citigroup, doctors and investors found little use for the machine because while Watson is a great generalist, mastering specifics of different businesses is a different story.
When IBM created an entirely separate business unit for Watson, they did so with the belief that Watson would generate "$10 billion in annual revenue within 10 years." At least, this is what IBM's CEO Virginia "Ginni" Rometty told fellow executives last October, as reported by the Wall Street Journal. But so far, Watson has managed to bring in only a tenth of its annual expectations, and it doesn't seem as though this will improve much in the near future.
Most problems, unfortunately, are much more complex than Jeopardy! clues, and as such, Watson fails to impress when it comes to curing cancer or developing effective investment strategies. As one Watson client told the Wall Street Journal, Watson just took too long to learn.
Alright, then what can Watson be used for?
Of course, with continued trials and experimentation, Watson will likely develop much greater capacities and be able to aid in solving several different issues. But right now, one of its greatest accomplishments is in a very strange field indeed: cooking. By generating unique flavor combinations, Watson has become a phenomenal chef, or at least, recipe maker. Watson even has its own line of barbecue sauce, and from what we hear, it's pretty phenomenal.
ABC has also compiled an extensive list of possible Watson applications, including shopping, travel, home buying, and potential solutions to joblessness.
With a mind like Watson's, the sky is the limit.
Images: Wikimedia Commons, IBM, NBC