Charles (Charlie) Safran, MD, MS: 50 Years of AI in Medicine

Fifty years of medical AI show reasoning got solved fastest; data bias, rule maintenance, and healthcare payment economics remain the real unsolved bottlenecks.

• Started in 1973: Charlie Safran, a Tufts logician, joined MIT's medical decision-making group and expected a Turing-capable AI doctor by the year 2000, using early Bayesian/decision-theory tools and LISP on PDP-10s.
• Hodgkin's disease paper: His first New England Journal of Medicine paper (lead author, before med school finished) applied Bayes' theorem to decide whether lymphangiography tests should change treatment choice between radiation and chemotherapy.

Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.