> but it's the wrong conclusion to make that if doctors were wrong about smoking they are also wrong about everything else. Usually they are not. Indeed. And u

> but it's the wrong conclusion to make that if doctors were wrong about smoking they are also wrong about everything else. Usually they are not. Indeed. And usually, there is no significant objection to the medical consensus, so it is irrelevant the "all medical advice" space. The interesting (from my perspective, but also from an information theoretic one) things are those about which there is disagreement. And the "usually right" on those subjects goes down significantly - perhaps it's still "usually", but maybe it's just 60% rather than the 99% of the "right" things which inspire people to blindly believe doctors about everything. And perhaps it's just 20%; We'll likely know about today's status only in 30 years. For example, there is still consensus about statin use to reduce cardiac problems through cholesterol lowering. I looked a little into that; the evidence is severely lacking, especially if you consider all cause mortality rather than just "lowering myocardial infraction risk" (which you can achieve 100% with Cyanide!). I have informally surveyed about 20 doctors, of which one was a cardiologist. Everyone other than the cardiologist was convinced statins are the best thing since sliced bread, have no risks, and excellent efficiency. Not one of them actually knew much about, e.g. statin effect on all cause mortality, or much of anything else related to statins except the official line and up-to-date (the service) summaries. The cardiologist still thought statins were a good thing, but was much more reserved about recommending them to anyone with high blood lipids, because he actually read the research and had actual experience (and discussions with informed colleagues).