There is little evidence that LW can do better on factual questions than mainstream medicine or dietology. I believe that there are few posters who are significantly better than an average doctor, but there are also others who are worse and the readers have not much data to decide who is which. Given that the LW advisors have little or no chance to examine the patient the way normal doctors can and thus are basing their diagnoses solely on verbal communication with the patient, the most rational advice would be “go and see your doctor”. Since there are a lot of different answers under that post, I wonder whether the future readers would benefit from that at all, rather than being harmed.
More generally, learning well settled fields is in many respects preferable to pondering difficult unsolved problems. At least you get a definite answer.
There is little evidence that LW can do better on factual questions than mainstream medicine or dietology.
To collect evidence (whichever way it might point) we should encourage factual questions on these topics. I’m interested in such evidence because some LWers have indeed claimed that a rational approach such as a good LWer might be capable of should indeed do significantly better than mainstream experts.
Since there are a lot of different answers under that post, I wonder whether the future readers would benefit from that at all, rather than being harmed.
It’s not clear one way or another, but they might benefit, e.g. from being exposed to certain opinions and references. Certainly, to extract concrete advice and follow it, they have to judge the many options discussed themselves, as well as look for other options not mentioned. That’s what LW rationalist training is supposed to be all about :)
More generally, learning well settled fields is in many respects preferable to pondering difficult unsolved problems.
In real life you don’t get to choose which problems to solve. As a basic example, for me to save my own life, I have to answer questions about medicine (and perhaps diet), even if no human has been able to answer them before.
To collect evidence (whichever way it might point) we should encourage factual questions on these topics.
That’s true. But the advice asking posts don’t do that particularly well. I have no idea how to collect meaningful evidence on LW’s expertise from what’s been written there.
There is little evidence that LW can do better on factual questions than mainstream medicine or dietology. I believe that there are few posters who are significantly better than an average doctor, but there are also others who are worse and the readers have not much data to decide who is which. Given that the LW advisors have little or no chance to examine the patient the way normal doctors can and thus are basing their diagnoses solely on verbal communication with the patient, the most rational advice would be “go and see your doctor”. Since there are a lot of different answers under that post, I wonder whether the future readers would benefit from that at all, rather than being harmed.
More generally, learning well settled fields is in many respects preferable to pondering difficult unsolved problems. At least you get a definite answer.
To collect evidence (whichever way it might point) we should encourage factual questions on these topics. I’m interested in such evidence because some LWers have indeed claimed that a rational approach such as a good LWer might be capable of should indeed do significantly better than mainstream experts.
It’s not clear one way or another, but they might benefit, e.g. from being exposed to certain opinions and references. Certainly, to extract concrete advice and follow it, they have to judge the many options discussed themselves, as well as look for other options not mentioned. That’s what LW rationalist training is supposed to be all about :)
In real life you don’t get to choose which problems to solve. As a basic example, for me to save my own life, I have to answer questions about medicine (and perhaps diet), even if no human has been able to answer them before.
That’s true. But the advice asking posts don’t do that particularly well. I have no idea how to collect meaningful evidence on LW’s expertise from what’s been written there.