this post was submitted on 15 Aug 2023
95 points (100.0% liked)
SneerClub
982 readers
8 users here now
Hurling ordure at the TREACLES, especially those closely related to LessWrong.
AI-Industrial-Complex grift is fine as long as it sufficiently relates to the AI doom from the TREACLES. (Though TechTakes may be more suitable.)
This is sneer club, not debate club. Unless it's amusing debate.
[Especially don't debate the race scientists, if any sneak in - we ban and delete them as unsuitable for the server.]
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
You have just demonstrated that you don’t understand Taleb’s critique. Admittedly, his basic critique is buried in tons of verbiage, but your response here is an irrelevance.
IQ measurements are next to useless on an individual level because a) IQ measurement is terrible & non-repeatable with very large variance between successive tests for any given individual & b) IQ doesn’t measure the thing you actually want, which is task-specific performance: it has terrible correlation with any given task-specific measure, barely rising above “vaguely related”.
At the population-level, IQ suffers from terrible statistical issues, including circularity affecting outcomes (SAT tests in the US are a particular problem), and inter-population differences that make comparisons extremely noisy. The field is also historically full of charlatans who literally made up data out of thin air, even before you start in with the problems with the actual data they drew upon & the stats they applied to it.
Ultimately, It doesn’t matter that you can measure some “factor” and show that there’s a weak correlation with lifetime wealth, or prison likelihood or whatever if that measurement is an otherwise useless one: Using IQ as a measure of an individual is wildly inappropriate. Using it as a population measure is next to useless because of widespread issues with both the input data & the statistical analysis done to torture some kind of correlation out of said data & call the job done.
Finally, when you’ve done all these population level stats on your so-called “g-factor” and squeezed some kind of vague relationship between various groups & your “g-factor” out of the data, what are you doing that /for/? What good do you expect to do in the world with that information? Because the only real-world use seems to be advocating for blocking the immigration or education of specific groups of people, despite the fact that, as has already been pointed out, you cannot use IQ on an individual level because it has extremely poor predictive value at the individual level. Sounds ... kind of racist don’t you think?