A few months after IBM's Watson won Jeopardy, the IBM Research team came to the Texas Medical Center in Houston for a dog and pony show about its possibilities for health care. They presented for two hours to an auditorium of a hundred doctors and administrators from TMC, one of the premiere medical communities in the world. From a technical point of view, it was pretty impressive what Watson can do, reading everything, never forgetting, assigning weights to knowledge over time, able to explain where it got an answer. But when the time came for the doctors to respond ... ouch! One male doctor said "I already have access to a $500,000 decision support system and I don't use that. Why would I start using your stuff?" Another woman cardiologist said she could see using Watson on about ten percent of her office appointments, but she had a hard time seeing how to integrate it and not be slowed down for the other 90% where it was unnecessary. Doctors are a tough crowd.
The Ruby Rogues podcast had Conrad Barski, author of Land of Lisp, a while ago:
http://rubyrogues.com/043-rr-book-club-land-of-list-with-con...
Barski is a medical doctor who gave up medicine to code. He explained how it just didn't fit the medical culture to be interested in thinking deeply about an interesting problem, House MD not withstanding. Medical training is designed to give real time answers to presented problems, and it is, er..., 'highly discouraged' to tell a patient "That's an interesting case you have there. I'm going to go google that tonight and I'll have more information in the morning."