The Association of American Medical Colleges has projected that in 2015 there will be 63,000 fewer physicians then are needed in the US. A Microsoft executive speculated recently that future artificial intelligence applications might include a "medical doctor in a box" that could help with basic medical issues in the face of this shortage.

For example, radiologists basically focus on looking at visual images from medical devices, like X-Rays and CAT scans. Machines are getting much better at analyzing that type of visual information. Some people are thrilled at this prospect, while others are more concerned that the quality of their medical care will decrease if it is farmed out to artificial intelligence. The question is whether AI can be taught to evaluate and interpret an "abnormal pattern" on an X-Ray, and to differentiate between "abnormal" and an "abnormal finding" that is artificially created by technique and positioning. 4-5 years of training after medical school is typically required for radiologists to be able to make this distinction accurately.