I've been reading about expert systems and noticed started reading about MYCIN.

I was astonished to find that MYCIN diagnosed patients better than the infectious diseases physicians.


Since, it had such a good success rate, why did it fail?


The MYCIN Experiments of the Stanford Heuristic Programming Project were indeed showing impressive diagnostic performance back in 1983. By now, had the medical community and lawyers not met to raise ethical concerns with the university, it is likely that a few billion dollars in unnecessary office visits would have been saved and the average insurance premium would be significantly lower in cost today.

The ethical issue that was raised in the periodicals of that time was that ancillary or unrelated chronic conditions would be missed and people would die. Of course that is a bit of a misstatement, since MYCIN could be altered to ask additional questions and order the kind of blood work a general practitioner would order.

If the MYCIN project objective encompassed all of the Hippocratic Oath and was further developed, no rational ethical argument exists supporting a higher risk to life and limb for a well designed software system than the current conveyor belt style office visit approach of post-modern medicine. Add current technology, and the effectiveness of the pre-1965 home visit from doctors would likely be restored.

The only valid ethical question is whether it is reasonable for people who have undergone seven years of study under the assumption that they could see a high return on investment not see the full return. Of course that's the ethical question for all of the jobs that have been replaced by automation. Labor has characteristically united to save careers in several, but not all, of those cases.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.