TORONTO—IBM’s artificial intelligence computer, Watson, may have impressed television audiences on the quiz show Jeopardy in February when it competed and won against two human contestants. But at the eHealth 2011 conference this week, IBM’s chief medical scientist for care delivery systems, Martin Kohn, is quick to point out that, while Watson has enormous potential to improve health care, the goal is not for it to replace physicians nor will it mean the elimination of future medical errors committed by physicians.
“One: It’s not achievable. Two: It’s not desirable,” said Kohn during a keynote at the conference.
What Watson can do is to reduce medical diagnostic errors, deal with ambiguous and incomplete medical information, compile both structured and unstructured data, and make medical recommendations.
With technology like Watson, the expectation is that future health care will be evidence-based, more collaborative, safer and less wasteful, said Kohn.
Yet, he continued, the goal is not for Watson to actually practice medicine. It’s one reason IBM didn’t name the artificial intelligence computer “Dr. Watson.”(That and the fact that the name was already trademarked).
Watson will merely “assist” physicians in their jobs, said Kohn, but the challenge will be to identify a use that is not intrusive to the health care process given that frequent, system alerts are disruptive to medical professionals.
Follow Kathleen Lau on Twitter: @KathleenLau