A lawyer is not allowed to stop learning about new laws; he would eventually be disbarred. An architect is not allowed to construct buildings that would only have been up to code at the time she finished her degree. And doctors should likewise not ignore the state of the art in their field.
Doctors invest huge effort in keeping up with the state of the art in their field, the problem is that to be a competent doctor you also need considerable amount of intuition and heuristic.
New advances sometimes requires that you unlearn things and this i a completely different process; when you simply keep up with the state of the art you have a set of known facts and techniques and slowly you discover more and more of them. Unlearning things requires you to review entire branches of your knowledge tree to eradicate every trace of the now wrong fact
I'm afraid this hasn't been my usual experience.
You cannot be serious. So if someone tells you that you’ve got a fact wrong, you have to review entire branches of your knowledge tree to eradicate every trace of the now wrong fact? So if you get the directions to a party wrong it’s a project to get rid of them and learn the correct one?
https://en.wikipedia.org/wiki/Principle_of_explosion
Let's say your "knowledge" is a set of propositions which you believe to be true. This set is continually expanding for two different reasons:
1. You're acquiring new facts about the world through direct observation (emperical knowledge)
2. You are thinking: making valid arguments which start from your current knowledge and lead to new propositions (analytic knowledge)
At some point, a false fact creeps in. Your system is now "exploded" - the whole system is inconsistent and with a few lines of perfectly valid argument, you can now arrive at any conclusion. You now believe both "A" and "not A" for all values of A in the set of all well-formed formulas. Obviously this is not ideal. So we need to remove the false fact. But in the mean time, your mind has been at work, adding new propositions to your knowledge set. It is necessary to remove not only the offending bad fact that crept in, but any new proposition resulting from a valid argument which used that bad fact as a premise. Computationally, the easiest way to do this is to maintain a graph where the nodes are propositions and edges connect to their premises. For example, we would record in this graph structure that we now believe X because we deduced it from A, B, and C. If this structure, which the parent comment called "knowledge tree," exists, we can quickly remove all invalid propositions from our knowledge set and return to an consistent system. In practice its a very good idea to also store the arguments themselves. The metamath project does this, and can verify the entirety of its knowledge (which includes set theory, some calculus, complex numbers, trignometry, etc.) all the way back to axiomatic first principles in just a few minutes.
Humans, of course, do not do this in their ordinary lives. They often believe propositions without being able to recall or reproduce the arguments, or cite the facts, which first led them to believe that proposition. This leads to a great deal of intellectual inertia. In mathematics and the sciences the situation is a little better: researchers certainly strive for this as an ideal, and largely succeed - a good research scientist has a deep understanding of the crucial experiments, theorems, and approximations that underpin their own work, and a rigorous practice of citing sources helps keep track of these dependencies. Therefore, if one paper is found fraudulent, other papers which cite it can be reviewed. But it is a difficult, even overwhelming project and no one can do it outside their own narrow area of expertise.
In my example, X was "tell patient they have a death sentence". The fact that yours is closer to a specific situation in the article does not make it "a different failure mode" with different dynamics and causes.
Did you miss that, or did you not consider it a good reason?
If you can take a moment to review the thread, I only brought it up to distinguish valid vs invalid reasons for not proposing a kind of treatment, and thus why a doctor's intuition would vs wouldn't be defensible given the state of the literature.
Where specifically is the article saying the opposite of any of that? Unless the article talks about the topic of appropriately classifying failure modes, which it isn't, it wouldn't bear on that topic.
And I don’t see what that would bear on my original point (which was just distinguishing between valid vs invalid reasons not to recommend a treatment).
haaah. the things people believe about lawyers. merely being bad at your job isn't nearly enough to actually be disbarred.
If a lawyer were to consistently act illegally as a result of not keeping up with their knowledge of the law, they definitely would eventually be disbarred.
The case of doctors is interesting as they are a well defined, skilled, educated, and reasonably well-measured population engaged in activities for which outcomes can be quantitatively assessed; outcomes, health & wellness measures, or that ultimate outcome metric: deaths.
Which means that in studying doctors there are valuable lessons which might be transferred to other fields. Such as software developers, venture finance, executive management, and end-user practices. Or in addressing social ills, economic inequality, education, opportunity, respecting others, expanding worldviews.
The habit of looking past the immediate to see potential analogues and metaphors is a powerful one, though many seem resistant to acquiring it.
> Procedures live on even after they’ve been proved ineffective. It can lead to harms and wasted resources.