As most of the world knows by now, IBM has created a computer
called Watsonthat is, arguably, the smartest, most human-like computer ever built.
Most folks know Watson as the computer that won the Jeopardy game some years ago. Watson had to understand verbal language to win, a hard thing for a computer to do well (as Siri users will attest).
Since the days of Jeopardy, Watson has been helping doctors fight cancer. “It’s ingested 2 million pages [of medical information] and understands medical language,” Rometty says. Today it helps doctors verify a diagnosis and pick the statistically best treatments.
But IBM CEO Ginni Rometty says that we haven’t seen anything yet.
Watson 2.0 will “see” she says, meaning it will be able to look at pictures like x-rays, understand them and interpret them.
Watson “3.0 is one that can debate and reason,” she says. Maybe to argue its point if a human disagrees?
Making Watson that human-like sounds scary, said Fortune journalist Stephanie Mehta, who was interviewing Rometty on stage. The audience murmured agreement. Perhaps they were thinking of movies like “2001: A Space Odyssey,” in which a self-aware super-computer named HAL 9000 starts killing humans that threaten it.
So Rometty defended the frighteningly powerful machine IBM is building:
“It’s a service. Do not be afraid. It is really, truly an advisor to a decision-making process. There are many things the human brain does that is not imitated … Think of it as an assistant.”