Nassim Taleb: IQ is largely a pseudoscientific swindle

[Yonis]

From the author of The Black Swan - Nassim Taleb -Lebanese Christian completely debunks IQ as a measure of intelligence

Haven't seen a single right wing youtuber debunk this

Very telling

 

[NotreDame]

IQ was always a scam. Anyone with a high enough IQ can tell you that.

It is impossible to quantify something like intelligence. There's practical intelligence, logical intelligence, emotional intelligence, memory intelligence, the ability to learn and understand quickly, ability to think creatively and originally. Literally impossible. IQ can only test how unintelligent you are.

 

[Prime Minister]

Can anyone summarise what he means?

 

[Octavian]

Can anyone summarise what he means?

iq is bs

 

[Prime Minister]

iq is bs

Thanks.

 

[x z]

white people are so, so insecure in their ethnicity and peoplegroups future (or lack thereof as they are onset to become extinct cause their women ever want some colored dlck or only 2 kids, training one of them to become LGBT and not reproduce) that they sit around and use white mans age old BS pseudoscience to make themselves feel better. This whole IQ bs was stupid back when it was made in the first place, to justify colonizing africans and asia, and is stupid now when they have had experience and seen africans taken from shi*holes like burco and khartoum and congocity, who supposedly born with low intelligence, and live in cadaan countries and study in oxford and cambridge alongside their racially superior intellegent white offspring.

iq is based of resources and education access, which is why whites arent even on top, its thpuse east asians who start revising for college at the age of 4.

 

[The alchemist]

The IQ test is an effective toolkit to reinforce old racist ideology, and it's fundamentally wrong. It's just factor analysis superimposed on made-up premises to give the illusion of a standard.

 

[Unlock]

Can anyone summarise what he means?

He’s saying that IQ is formally thought to be accurate because it falls into a natural distribution; meaning mathematically what is predicted to happen from randomness. This means that it’s believed to be true because it’s so random it has to be unbiased, and thus an accurate measure. However, what he points out is that this is a false positive because when you test it mathematically or in a real world application and test for correlation, it doesn’t predict anything and that is why it’s rubbish. I wish I could explain the prior argument about it mathematically not being accurate in layman terms but I’m not a statistician. Basically the IQ test is just a test but doesn’t predict intelligence, just how well you can take the test.

I should also clarify, he states the test tests for stupidity and assumes that the inverse, intelligence, is also tested for. But the rules of statistic tell us it’s not.

I would like to add my own two cents. I disagree that it even tests for stupidity because what’s stupid in one environment is not always the same in another. That has always been an historical problem with the test. Our neural network are programmed by things outside of just genetics. Since a white person designed the test they’re brain is wired to say a thought process is stupid when in a certain environment it can be beneficial. Smart is smart but stupid isn’t as linear. The IQ test more accurately can maybe test for mental incapacity though.

 

[Prime Minister]

He’s saying that IQ is formally thought to be accurate because it falls into a natural distribution; meaning mathematically what is predicted to happen from randomness. This means that it’s believed to be true because it’s so random it has to be unbiased, and thus an accurate measure. However, what he points out is that this is a false positive because when you test it mathematically or in a real world application and test for correlation, it doesn’t predict anything and that is why it’s rubbish. I wish I could explain the prior argument about it mathematically not being accurate in layman terms but I’m not a statistician. Basically the IQ test is just a test but doesn’t predict intelligence, just how well you can take the test.

I should also clarify, he states the test tests for stupidity and assumes that the inverse, intelligence, is also tested for. But the rules of statistic tell us it’s not.

I would like to add my own two cents. I disagree that it even tests for stupidity because what’s stupid in one environment is not always the same in another. That has always been an historical problem with the test. Our neural network are programmed by things outside of just genetics. Since a white person designed the test they’re brain is wired to say a thought process is stupid when in a certain environment it can be beneficial. Smart is smart but stupid isn’t as linear. The IQ test more accurately can maybe test for mental incapacity though.

Very intelligent people also score high on the test so is it that it's able to differentiate the very intelligent from the very stupid but for the vast majority of people it does not accurately portray how they would do in real life?

 

[Unlock]

Very intelligent people also score high on the test so is it that it's able to differentiate the very intelligent from the very stupid but for the vast majority of people it does not accurately portray how they would do in real life?

It’s a test so yes very intelligent people usually can break down the test to figure out the cues and patterns to do well, which is technically what it’s testing you on anyways. But just because you don’t do well doesn’t necessarily mean you’re not very intelligent, or vise versa. So it’s somewhat accurate but not statistically accurate. A good analogy is taking a multiple choice test on a subject that you didn’t study for. Most tests have verbal and organizational patterns that you can use to differentiate choices to guess an answer even if you didn’t study, and they’ll teach you that for standardized tests. Essentially you’re reading the mind of the test maker. Now imagine that you did the same test in a language you don’t understand. You could still figure out the organization pattern to not fail but it becomes extremely unlikely you’ll do well compared to someone who also has the verbal cues. What the video addresses is that the statistics used for the test says because it follows a natural distribution then it means 99.9% are taking the same test in a language they understand. But off of what we know about cultures you can then argue that because some people speak that same language as the test maker it would be easier for them to understand the verbal cues of the test. Which is what racists disagree with because they say if that’s the case the results wouldn’t fall into a natural distribution, but the video points out it’s a false distribution.

Also the fact that intelligent people score well isn’t as much proof it works as you think. Here’s why. The test follows a bell curve so that means the vast majority of people are relatively 50/50 right and wrong, meaning they’re not overly right or overly wrong. Which means if there are four options, if you’re only intelligent enough to exclude the two that are wrong you’re going to do decently. The test only proves that the vast majority of people can point out the obvious most times. If you can’t then something is either wrong with you, or you fall into the group where wrong to you is different from the test makers, but (importantly) you still had a logical path arriving at the wrong conclusion. As far as being very intelligent, all you need to be very intelligent is guess correctly or be right more times than not out of the two remaining options.

So to break it down an average IQ of 100 just means you weren’t stupid enough to choose the obviously wrong answers too many times. An IQ of 130 usually either means you either actually figured out the pattern or you weren’t stupid enough to score below 100 but out of the remaining choices you guessed properly. The man in the video is arguing that statistically the IQ test doesn’t do an accurate job of differentiating the two for a >110 score, and that’s evident by the real world application.

 

[Prime Minister]

It’s a test so yes very intelligent people usually can break down the test to figure out the cues and patterns to do well, which is technically what it’s testing you on anyways. But just because you don’t do well doesn’t necessarily mean you’re not very intelligent, or vise versa. So it’s somewhat accurate but not statistically accurate. A good analogy is taking a multiple choice test on a subject that you didn’t study for. Most tests have verbal and organizational patterns that you can use to differentiate choices to guess an answer even if you didn’t study, and they’ll teach you that for standardized tests. Essentially you’re reading the mind of the test maker. Now imagine that you did the same test in a language you don’t understand. You could still figure out the organization pattern to not fail but it becomes extremely unlikely you’ll do well compared to someone who also has the verbal cues. What the video addresses is that the statistics used for the test says because it follows a natural distribution then it means 99.9% are taking the same test in a language they understand. But off of what we know about cultures you can then argue that because some people speak that same language as the test maker it would be easier for them to understand the verbal cues of the test. Which is what racists disagree with because they say if that’s the case the results wouldn’t fall into a natural distribution, but the video points out it’s a false distribution.

Also the fact that intelligent people score well isn’t as much proof it works as you think. Here’s why. The test follows a bell curve so that means the vast majority of people are relatively 50/50 right and wrong, meaning they’re not overly right or overly wrong. Which means if there are four options, if you’re only intelligent enough to exclude the two that are wrong you’re going to do decently. The test only proves that the vast majority of people can point out the obvious most times. If you can’t then something is either wrong with you, or you fall into the group where wrong to you is different from the test makers, but (importantly) you still had a logical path arriving at the wrong conclusion. As far as being very intelligent, all you need to be very intelligent is guess correctly or be right more times than not out of the two remaining options.

So to break it down an average IQ of 100 just means you weren’t stupid enough to choose the obviously wrong answers too many times. An IQ of 130 usually either means you either actually figured out the pattern or you weren’t stupid enough to score below 100 but out of the remaining choices you guessed properly. The man in the video is arguing that statistically the IQ test doesn’t do an accurate job of differentiating the two for a >110 score, and that’s evident by the real world application.

Thanks I appreciate this breakdown.

 

[The alchemist]

Basically, the IQ test only measures how well you do on the IQ test – it's not a good indication of life outcomes, I.e., negative correlation. Dweebs claim it's the best estimates for overall intelligence, and extrapolate a lot of nonsense from it, thinking it gives us a good insight, shortsightedly and reductively narrowing down the human complex mind into some digits, without ever doing molecular biological research on the gene-specific relationship on environment environmental pressures, which actually could give us some value.