It’s a perfectly legitimate dismissal. People talking about how awesome they are has well known spiral qualities, and isn’t that useful. As with the start of this comment chain: The burden of why we should talk about this is on people who think we should, as the reasons why we shouldn’t are obvious: It doesn’t appear to serve any use and is at risk of becoming nothing more than backpatting.
Mistake 2 - It’s useless to talk about a trait that makes you “awesome”.
Discussing how to capitalize on a positive trait seems quite useful to me, and the more “awesome” the trait, the more potential mileage of exchanging information on how to exploit it. It’s a trait where we have a comparative advantage (and some disadvantages) far out on the tail of the distribution. It’s stupid not to discuss it in the rare circumstance where we have a large sample size far out on the tail of the distribution, just where we need the data.
I think discussions of how to capitalize on intelligence might be interesting and worth while. That’s got nothing much to do with discussing what the actual statistical distribution of IQ is among the LW population; all that needs saying about that is that it turns out that LW has quite a lot of high-IQ people.
(What would be said in a discussion of how to capitalize on IQ-style intelligence? Non-obvious careers in which such intelligence turns out to be highly advantageous. Typical weaknesses that accompany high IQ, and how to use abilities that correlate with IQ to overcome them. Classes of problems that turn out to submit more or less readily than you’d expect to the kinds of thinking that high-IQ people are good at. I don’t know whether any of this would be unobvious enough to be worth it, but it seems at least an order of magnitude more productive than yet another Oh Grandmother, What A Big IQ You Have discussion.)
I think discussions of how to capitalize on intelligence might be interesting and worth while.
So do I.
But “Quite a lot of high-IQ people” strikes me as vague to the point of useless. The real distribution matters. Numbers matter. Being top 50% is different than being top 10% is different than being top 0.1%. The value of our population as a data sample is dependent on how rare the sample is. How we might capitalize would also depend on the numbers.
I don’t know whether any of this would be unobvious enough to be worth it
IMO, a particular weakness of the very intelligent is behaving as if reality is obvious and rational, something that one can figure out, instead of something that will surprise you if you take the time to actually look at it, or ask other people about.
yet another Oh Grandmother, What A Big IQ You Have discussion.
I guess I missed the “who’s got the biggest swinging IQ” discussions. Have you really had a lot of those? That does sound tiresome.
I’m more interested costs/benefits, and making the most of what should be an asset.
How we might capitalize would also depend on the numbers.
How, specifically, would it depend on the numbers? What are the important differences in the type of content we would want to encourage if the average IQ were 140 as opposed to if it were 130?
How we might capitalize would also depend on the numbers.
How any given person capitalizes on their mental powers will depend on just what those mental powers are. But it seems pretty clear to me that there’s much more variation between different LWers than there is uncertainty in just where the average is. (From what I’ve seen it seems pretty much certain that for most measures of IQ, (1) the average among active LWers is somewhere between 120 and 140 and (2) the range is at least ~120 to ~160.)
What discussions would we have that would need to be much different depending on whether the average IQ here is 120 or 140, but that wouldn’t end up being useless for most of LW in either case?
And suppose we nail down very accurately what the average IQ here is by some particular measure. That still won’t tell us much, because whatever you may say about g there’s plenty of variation in how smart people are smart.
I guess I missed the “who’s got the biggest swinging IQ” discussions.
That isn’t what I meant; sorry if I was unclear. I was intending to cover “who’s got the biggest IQ”, “what is our average IQ”, “how much does our IQ vary”, and basically the whole range of discussions that are about characterizing LW IQ measurements. I just don’t see that there’s much value to those discussions beyond the level (which we’ve already reached) of observing that the LW population is at least one or two standard deviations smarter than average according to typical IQ measures, because anything that would be invalidated by not having more precise knowledge of the LW population average is equally invalidated by the fact that the average is only an average and actual LWers are different in many different directions.
(1) the average among active LWers is somewhere between 120 and 140
The OP referred to the SAT/ACT score averages as 0.11%, giving 3sigma+, while the IQ numbers averages I saw were 140 to 146, being almost 3sigma.
Certainly this data set isn’t equally useful to everyone, but I think significantly useful to many. It depends on how significant intelligence level is in life. Out at 3sigma and beyond, I’d think pretty significant. I wouldn’t expect that estimate to be universal, but I would expect it to get at least a good plurality.
OK, so “between 120 and 140” was on the low side. That’s not the point; the point is that whatever measure you pick, the remaining uncertainty about the mean is less than the known variation between people.
It’s a perfectly legitimate dismissal. People talking about how awesome they are has well known spiral qualities, and isn’t that useful. As with the start of this comment chain: The burden of why we should talk about this is on people who think we should, as the reasons why we shouldn’t are obvious: It doesn’t appear to serve any use and is at risk of becoming nothing more than backpatting.
Mistake 1 - I am intelligent = I am awesome.
Mistake 2 - It’s useless to talk about a trait that makes you “awesome”.
Discussing how to capitalize on a positive trait seems quite useful to me, and the more “awesome” the trait, the more potential mileage of exchanging information on how to exploit it. It’s a trait where we have a comparative advantage (and some disadvantages) far out on the tail of the distribution. It’s stupid not to discuss it in the rare circumstance where we have a large sample size far out on the tail of the distribution, just where we need the data.
I think discussions of how to capitalize on intelligence might be interesting and worth while. That’s got nothing much to do with discussing what the actual statistical distribution of IQ is among the LW population; all that needs saying about that is that it turns out that LW has quite a lot of high-IQ people.
(What would be said in a discussion of how to capitalize on IQ-style intelligence? Non-obvious careers in which such intelligence turns out to be highly advantageous. Typical weaknesses that accompany high IQ, and how to use abilities that correlate with IQ to overcome them. Classes of problems that turn out to submit more or less readily than you’d expect to the kinds of thinking that high-IQ people are good at. I don’t know whether any of this would be unobvious enough to be worth it, but it seems at least an order of magnitude more productive than yet another Oh Grandmother, What A Big IQ You Have discussion.)
So do I.
But “Quite a lot of high-IQ people” strikes me as vague to the point of useless. The real distribution matters. Numbers matter. Being top 50% is different than being top 10% is different than being top 0.1%. The value of our population as a data sample is dependent on how rare the sample is. How we might capitalize would also depend on the numbers.
IMO, a particular weakness of the very intelligent is behaving as if reality is obvious and rational, something that one can figure out, instead of something that will surprise you if you take the time to actually look at it, or ask other people about.
I guess I missed the “who’s got the biggest swinging IQ” discussions. Have you really had a lot of those? That does sound tiresome.
I’m more interested costs/benefits, and making the most of what should be an asset.
How, specifically, would it depend on the numbers? What are the important differences in the type of content we would want to encourage if the average IQ were 140 as opposed to if it were 130?
How any given person capitalizes on their mental powers will depend on just what those mental powers are. But it seems pretty clear to me that there’s much more variation between different LWers than there is uncertainty in just where the average is. (From what I’ve seen it seems pretty much certain that for most measures of IQ, (1) the average among active LWers is somewhere between 120 and 140 and (2) the range is at least ~120 to ~160.)
What discussions would we have that would need to be much different depending on whether the average IQ here is 120 or 140, but that wouldn’t end up being useless for most of LW in either case?
And suppose we nail down very accurately what the average IQ here is by some particular measure. That still won’t tell us much, because whatever you may say about g there’s plenty of variation in how smart people are smart.
That isn’t what I meant; sorry if I was unclear. I was intending to cover “who’s got the biggest IQ”, “what is our average IQ”, “how much does our IQ vary”, and basically the whole range of discussions that are about characterizing LW IQ measurements. I just don’t see that there’s much value to those discussions beyond the level (which we’ve already reached) of observing that the LW population is at least one or two standard deviations smarter than average according to typical IQ measures, because anything that would be invalidated by not having more precise knowledge of the LW population average is equally invalidated by the fact that the average is only an average and actual LWers are different in many different directions.
The OP referred to the SAT/ACT score averages as 0.11%, giving 3sigma+, while the IQ numbers averages I saw were 140 to 146, being almost 3sigma.
Certainly this data set isn’t equally useful to everyone, but I think significantly useful to many. It depends on how significant intelligence level is in life. Out at 3sigma and beyond, I’d think pretty significant. I wouldn’t expect that estimate to be universal, but I would expect it to get at least a good plurality.
OK, so “between 120 and 140” was on the low side. That’s not the point; the point is that whatever measure you pick, the remaining uncertainty about the mean is less than the known variation between people.