Why are humans still in the loop with advancing AI capabilities
? That's the provocative title of a fun BMJ Digital Health article by Jonathan H. Chen, Michael Pfeffer and Christopher Longhurst, MD, MS. Before pushing back in gasping horror, it is more an exercise in understanding where the clinician's value lies than a statement that AI systems are ready for the job (although they are getting there on some dimensions). My favorite nugget: "If a pleasant bedside manner were all that was needed, we could strap an AI-powered talk box onto a nice Golden Retriever and call it a doctor." The authors offer a simple (and no doubt simplistic but useful) framework: a human brings judgment, negotiation skills and accountability. I think this last piece, accountability, is the only one that will stand the test of time across many areas of human endeavor: negotiation skills will be the first to fall, judgment and "taste" will follow (with different activities falling on their own timelines) but accountability is just a legal, moral and financial requirement.
