Does that mean there is a terrible ignored risk? No, when there is a real risk, the brightest people of extreme and diverse intellectual accomplishment are the ones most likely to be concerned about it (and various “autodidacts” are most likely to fail to notice the risk).
Besides, being more concerned is not the same as being more likely to be concerned. Just as being prone to panic doesn’t automatically make you better at hearing danger.
being more concerned is not the same as being more likely to be concerned
True, and I see that this distinction undercuts one of the ways there could be more autodidact concern than expert concern. But there is at least one more way, which I suggested earlier.
Imagine a world populated by a hundred experts, a hundred autodidacts, and a risk. Let E be the number of experts concerned about the risk, and A be the number of concerned autodidacts.
I interpret you as saying that E is greater than A. Is this a correct interpretation?
To the claim that E > A, I am saying “not necessarily.” Here is how.
Since the risk is a genuine risk, we assume that nearly all the experts are concerned. So we set E = 95. Now suppose those without formal training all suffer from the same common pitfalls, and so tend to make errors in the same direction. Suppose that due to these errors, autodidacts with their little learning are even more likely to be concerned. If they were all better trained, they would all relax a bit, and some would relax enough to cross the line into “not concerned” territory.
The above scenario seems perfectly plausible to me; is there some problem with it that I have missed? Does it miss the point? It is not the most likely scenario, but it’s far from impossible, and you seem to have cavalierly ruled it out. Hence my original request for a source.
The above scenario seems perfectly plausible to me
Seems highly unlikely for some risk the properties of which you don’t get to choose. Therefore in no way contradicts the assertion that experts are more likely to become aware of risks.
To large extent everyone is an autodidact, without scare quotes—a lot of learning is done on your own even if you are attending an university. It’s just that some people skip exercises and mistake popularization books for learning material, and so on. Those aren’t more likely to make correct inferences, precisely due to their lack of training in drawing inferences.
edit: and of course there are people who were not able to attend an university, despite intelligence and inclinations towards education, due to factors such as poverty, disability, etc. Some of them manage to learn properly on their own. Those have their work to show for it, various achievements in technical fields, and so on. I wouldn’t put scare quotes around those. And the brightest aren’t going to ignore someone just because they don’t have PhD, or listen to someone just because they do.
Seems highly unlikely for some risk the properties of which you don’t get to choose. Therefore in no way contradicts the assertion that experts are more likely to become aware of risks.
OK, so maybe this turns on how likely “likely” is?
Well, one can always make some unlikely circumstances where something generally unlikely is likely. E.g. it’s unlikely to roll 10 sixes in the row with this die. You can postulate we’re living in a simulator set up so that the die would have 99% probability of rolling 10 sixes, that doesn’t actually make this die likely to roll 10 sixes in the row if its unlikely that we are living in such a simulator. This is just moving improbability around.
Yes, that’s true. So, is that what I was doing all along? It sure looks like it. Oops. Sorry for taking so long to change my mind, and thanks for your persistence and patience.
I understand the scare quotes.
I agree that autodidacts “are more likely to fail to notice some genuine risk, than experts are.”
But autodidacts are also more likely to exaggerate other genuine risks than experts are, are they not?
If (3) is true, then doesn’t that undermine the claim “Experts are more likely to be concerned over risks than autodidacts are”?
What I said was:
Besides, being more concerned is not the same as being more likely to be concerned. Just as being prone to panic doesn’t automatically make you better at hearing danger.
True, and I see that this distinction undercuts one of the ways there could be more autodidact concern than expert concern. But there is at least one more way, which I suggested earlier.
Imagine a world populated by a hundred experts, a hundred autodidacts, and a risk. Let E be the number of experts concerned about the risk, and A be the number of concerned autodidacts.
I interpret you as saying that E is greater than A. Is this a correct interpretation?
To the claim that E > A, I am saying “not necessarily.” Here is how.
Since the risk is a genuine risk, we assume that nearly all the experts are concerned. So we set E = 95. Now suppose those without formal training all suffer from the same common pitfalls, and so tend to make errors in the same direction. Suppose that due to these errors, autodidacts with their little learning are even more likely to be concerned. If they were all better trained, they would all relax a bit, and some would relax enough to cross the line into “not concerned” territory.
The above scenario seems perfectly plausible to me; is there some problem with it that I have missed? Does it miss the point? It is not the most likely scenario, but it’s far from impossible, and you seem to have cavalierly ruled it out. Hence my original request for a source.
Seems highly unlikely for some risk the properties of which you don’t get to choose. Therefore in no way contradicts the assertion that experts are more likely to become aware of risks.
To large extent everyone is an autodidact, without scare quotes—a lot of learning is done on your own even if you are attending an university. It’s just that some people skip exercises and mistake popularization books for learning material, and so on. Those aren’t more likely to make correct inferences, precisely due to their lack of training in drawing inferences.
edit: and of course there are people who were not able to attend an university, despite intelligence and inclinations towards education, due to factors such as poverty, disability, etc. Some of them manage to learn properly on their own. Those have their work to show for it, various achievements in technical fields, and so on. I wouldn’t put scare quotes around those. And the brightest aren’t going to ignore someone just because they don’t have PhD, or listen to someone just because they do.
OK, so maybe this turns on how likely “likely” is?
Edit: fixed quotation marks
Well, one can always make some unlikely circumstances where something generally unlikely is likely. E.g. it’s unlikely to roll 10 sixes in the row with this die. You can postulate we’re living in a simulator set up so that the die would have 99% probability of rolling 10 sixes, that doesn’t actually make this die likely to roll 10 sixes in the row if its unlikely that we are living in such a simulator. This is just moving improbability around.
Yes, that’s true. So, is that what I was doing all along? It sure looks like it. Oops. Sorry for taking so long to change my mind, and thanks for your persistence and patience.