It's not a trend so much as it is a redefinition of intelligence. When you get older, you realize that somebody can be bright/wise/funny despite being fairly stupid in so-called "quantitative" terms. Somebody who can't read and isn't very logical can still accumulate common sense, mature with age, and become a better person in a lot of ways. So while people are stupider under the strict sense of "Lots of people can't do very simplistic things", they're brighter than the rote negativism would let on.

Put another way, I've got a friend who's a poet and who's training to be a nurse who insists Sarah Palin was a brilliant politician maligned by the liberal media. Online it's fun to suggest that a single stupid thought invalidates a person's entire existence, but - both fortunately and unfortunately - it's possible to think stupid things and still be a smart person, and vice versa.

Isaac Asimov, one of the most brilliant men I've had the pleasure of reading, wrote a short article on this and on the fallibility of lots of intelligence measures. You can find it here: http://www.wanderings.net/notebook/Main/WhatIsIntelligenceAn...

I think people are as smart as they need to be. You don't need to be able to summarize tabular data if you work at a fast food shop (e.g. as a waiter).

And it definitely has to do with the circles you hang around in. As a programmer you're more likely to be in a rather smart circle.

dono. I've met more than a few smart people working in the service industry. (I've hired a few, too, with good results.)

Personally, when I hire someone, I rate hobby experience as more desirable than paid experience. I can give you paid experience. At least for SysAdmins, I can train you in all the gotchas of production. But I can't make you like it.

I think Robin Hanson and you are both right: I tend to overestimate the intelligence of random samples in the population (the receptionist), but, in trying to correct for that, I end up underestimating the knowledge/capabilities of at least some of my peers.

I think that estimating about non-peers is most dangerous. How smart/dumb is the guy who fixes your car? Or the lady who owns a flower shop?

I think it's quite a stretch to assume too much about intelligence simply on the basis of vocation, etc. And the tests that Hanson cites are highly based on reading comprehension (a skill not needed in many vocations)... the seeming logic problems are all simple enough to be understood by a chimp, so if humans miss them it's probably due to some other factor.