Watson as you pass the danger, but shows weakness (NewsFactor)

On the second day of his appearance in TV SHOW Jeopardy, IBM Watson supercomputer devastated its human opponents. But even Watson, with 90 IBM Power 750 servers, have weak points.

On Tuesday, the world's smartest Computer blew past his opponents based on carbon, Ken Jennings samples and Brad Rutter. At the conclusion of the evening, the computer has racked up $ 35,734 while Rutter had $ 10,400 $ 4800 and Jennings. At the end of Monday's game, Watson was linked with Rutter $ 5,000 each, with Jennings in 2000 dollars. The game ends Wednesday.

Is fundamentally different from humans

The winner gets $ 1 million. As Watson is not designed for shopping, IBM will donate the prize to charity if he wins his creation. The two men said they'll donate half if they manage to pull off on Wednesday.

Eric Nyberg, Carnegie Mellon University Professor of computer science who led a CMU team that helped IBM, commented that Watson's answers "fundamentally different" from human beings. For example, a competitor can buzz and takes few seconds to respond, while Watson will buzz only when you have a high degree of confidence that the answer is correct. Watson may develop and evaluate two million pages of information in three seconds. But if the question is short, Watson is a disadvantage, because it takes three seconds to complete on every issue.

Most difficult questions can throw out Watson. In a brief moment of happiness for human whiffed badly partisans, Watson on final Jeopardy Tuesday. The category was the city of United States, and the symptom response was: "largest airport of the city is named after a hero of World War II; the second largest, for a battle of World War II. " The correct answer was "what is Chicago?" Both humans responded correctly. Watson, who never spent much time in airports, much less U.S. ones, replied "what is the Toronto?"

On his blog smarter planet, IBM David Ferruci, project manager, Watson was quoted as saying that the computer had learned the names of the category "only weakly suggest the type of answer that is expected and, therefore, the machine would despise their meanings." Furthermore, he noted, there are cities in the United States called Toronto and Toronto, Canada has an American League baseball team.

Career Post-Jeopardy Watson

But the good news for Terminator-wannabes is that Watson is continuously learning how categories work in danger. He also learned to be wary of betting. As his level of confidence for the final Jeopardy question of Tuesday was low, it only bet $ 947 gain riled.

On the blog of IBM, IBM Steven Hamm noticed some of the uses for intelligence Watson after calm from the fame of danger. In medicine, he predicted, "a patient a doctor can describe a certain symptom or a high level of pain, which on the surface it may seem like an important clue to the cause of the noise." But could discover Watson, sifting through and its huge repository of data, another symptom might be the real key for diagnosis of weighing.

Other applications might include sifting through case law relevant to find previous, helping governments examine the requests from clients or social services, help intelligence agencies in the analysis of information flows, or reach out to the call center.