People thinking they're smarter than they are is on full display here, because stupid shit that "sounds right" leaves facts and logic in the dust 100.000% of the time. See, ten peeps with an IQ of 102 plus ONE with an IQ of 80 would have an "average" IQ of ((102 * 10) + 80) / 11 = 100, yet ten out of those eleven WOULD BE ABOVE THAT. Stupid fucking quote is stupid. Then again, what do I know - I'm just a dumb internet rando...
At the risk of maybe - just maybe - taking this way too seriously, this comment is at the same time correct, incorrect and irrelevant. And I myself always have to refrain hard from doing it despite knowing that because I love being a pedantic person about this kind of stuff. So let me elaborate:
1) It is correct - your example is easily one which shows what can happen. A more common example is not about intelligence: most people in the world have an above average number of fingers. Almost everybody has 10 fingers. Some, due to genetic irregularities, start off with more or (arguably more often) less. And then some lose fingers during their lifetime, putting the average at 9.9something, and everybody with 10 fingers is above average.
2) It is incorrect - while common use of the term "average" refers to the arithmetic mean, technically it doesn't have to be. There's quite a lot of others, less in use but with reasons to be used once in a while, amongst them the median, which by definition provides precisely what Carlin says: half of the people are below it, half are above. (In practice both halves will include a certain percentage which is exactly at the median)
3) It is irrelevant - while the calculation as a thought exercise is correct, intelligence - if we accept the IQ as measurement for that, but that'd be a different debate why that actually is a bad idea - is by definition on a normal distribution, the N(100,15)-distribution to be precise. And there strange things like in your example do not happen, because of symmetry reasons all the averages fall together (well, almost all, harmonic and geometric mean don't, but certainly arithmetic mean and median do).
But I don't think I took it too seriously, I mean who in their right mind doesn't want to know more about statistics? Amirite folks, amirite?