In Paul Fussell’s book on class (I think), he says that people are really worried about differentiating themselves from the class immediately below them, but largely ignorant of the customs and sometimes even existence of the classes above them. When I found SSC, and then The Motte, and stuff like TLP, I was astonished to find a tier of the internet I had had no idea even existed. The quality of discourse here is . . . usually . . . of the kind that “high brow” (by internet standards) websites THINK they are having, but when you see the best stuff here you realize that those clowns are just flattering themselves. My question is, who is rightly saying the same thing about us? Of what intellectual internet class am I ignorant now? Or does onlineness impose some kind of ceiling on things, and the real galaxy brains are at the equivalent of Davos somewhere?
Jump in the discussion.
No email address required.
Notes -
Not sure if you know this but IQ and its validity has a very deep history as a widely held belief in this very forum we are on. Im not being sarcastic. In fact one can make an argument that the discussions of IQ and its intersections with HBD being a very contentious CW topic is what partially led to the creation of The Motte.
Anyways. IQ is ultimately a metric. But its the best metric of among all others as a proxy for g, the thing we are really trying to measure. Im not gonna reiterate the literature.
But, in a discussion about intelligence it would be rather obtuse to not mention the best proxy of it would it not? Its not a value judgement.
Somebody else will hopefully respond with a more detailed breakdown of the actual studies but IQ:
Is very predictive of success in things people care about
Cannot be trained
Is a very well studied and consistent measure
Varies a lot person to person and has at least a large genetic component
To the degree that some tests meant to measure IQ can be studied for is a flaw in those tests but the ones we have are pretty good at not being gamed.
HBD stands for "Human Bio Diversity" and is the recognition that populations of people vary genetically. Uncontroversially in situations like the Dutch being taller on average than the Spanish. Incredibly controversially in situations like American Jews having a higher IQ on average to American Natives. People here disagree wildly on whether it is true and what it would imply if it were. But we do allow its discussion which brings quite a bit of heat from the sort of people who find the idea dangerous. We originally started out as a weekly thread on slatestarcodex.com run by Scott who now blogs on astralcodexten.substack.com, but he was harassed by people who didn't like the discussion so he asked us to got to his subreddit, then our own subreddit and finally this shiny new offsite. Although it's not just HBD that got us targeted, it's the willingness to entertain dangerous ideas with HBD being the most usually cited example.
Not who you are asking but I will provide a response to your last paragraph. And all your other concerns will probably be responded to by others anyways because the validity of IQ is an extremely popular opinion here on the Motte. Whilst blankslatism is extremely unpopular. I am surprised you posted here for so long without noticing as much.
No. Because.
If we want to make people smarter. We would need a working definition of smartness and a solid way to measure it to know if our interventions are working. This absolutely cannot be ignored. We would also need to know the mechanisms behind it let that be environmental, genetic or whatever to actually address the problem. In short, we can't just look away from it.
You can do things with what you learn. If what you learn is true, then you have a better predictive model of the world than someone who thinks otherwise. The purpose of discussing things is not only to improve something but to understand it as well. You can't understand something and its nth-order effects by.. ignoring it. Just knowing that IQ is group stratified has made my personal model of the World orders of magnitude better are being able to predict things.
So what? "Interesting people" is a very subjective term and it's best kept out of supposedly objective discussions.
That's literally what a probability is for. Some things are likely, some things are not.
If you are 5'7" tall. You might dream of playing in the NBA but it wouldn't be a lie to tell you that you are overwhelmingly not likely to achieve that dream and that your time and effort is better spent elsewhere.
Why shouldn't we apply the same level of sober pragmatism to matters of intellectual interest? It is exceedingly unlikely that someone with an 85 IQ can obtain a Ph.D. in Physics. If someone with 85 IQ tells me about his theoretical physics dreams, I would be feeding into his delusions if I told him anything other than it's unlikely.
Yes, it sucks for them. It sucks for everyone. But the truth sometimes sucks.
Thousands of people do this every day. As they get rejected from their programs of choice or get an F grade in calculus. They get told they are not good enough and that's the end of it. 50% of engineering students drop out.
The truth is worth whatever it is worth regardless of how much it hurts to accept it. The fact that it is hard to accept it isn't a very good reason to not seek it out.
Sure. But I'm not really one to keep papers or sources on hand. I arrived at my conclusions over years of reading various books, articles and posts and updating my conclusion over and over again until they reached a settling point.
I suggest you read the Wikipedia page on IQ to get a rough understanding of what it is exactly.
Else this article by Scott is a good primer. Its an introduction to just how useful IQ is as a metric.
Because it's a metric that can predict a lot of things. Such as social mobility, education attainment, wealth, crime rate, etc. And it predicts those things for groups and individuals.
IQ score is a better predictor of Job performance than Education level, College grades, and Interview scores by a mile. It's not even close.
IQ is correlated with so many things consistently in the same direction that the summed correlation is extremely strong. This is the telltale sign of a signal in the noise.
Of course, it exists.
We don't know how to make people smarter. We really don't. Millions of dollars were spent on improving educational outcomes for certain groups to no avail. Refer to Arthur Jensens most infamous paper.
Everything you suggested helps one reach their natural limit, they do little to go past that limit.
Sure.
IQ has a r of 0.82 with National GDP/capita. That information is immensely useful for me. I can spot nations that are doing better or worse than their expected outcome and analyze further better than anyone who isn't aware of this, because.. I know where to look!
Seriously, I challenge you to find me something that explains this much variance that isn't an economic metric or just another proxy for IQ.
I don't care at all. The Motte is a place for adults to discuss adult things. And I will speak the truth even if it really really hurts because that's just my value system. I value the truth more than not offending people with it.
Exceptions don't disprove the rule.
That's on you, has nothing to do with the truth value of IQ as a metric. I reiterate, It's not a value judgment or a prescription, its a description.
Profound misunderstanding of the central HBD claim.
Group differences give you little information on the individual. Imagine you have two normal distributions
Na(100,15)andNb(85,15). Na refers to the White IQ probability distribution, Nb refers to the black. There will be a lot of black people who score higher on an IQ test than a lot of white people and vice versa. BUT on average, white people score higher.Also the fact that there is a race based IQ score gap is... not controvertial. The reason for that gap and what to do about it is. Just look it up.
The idea behind it is simple. When you track various different abilities such as mathematical test scores, language abilities, musical abilities, memory recall under stress, and many more. All those results tend to correlate. Meaning people who are good at one test are also likely to be good at another test. Keeping in mind care is taken to make sure people can't learn or study for these tests, they are trying to test of inherent ability.
This suggests the existence of there being some kind of latent variable/factor that is a feature of all the above functions. This latent factor is called the 'g factor'. Or the general intelligence factor. Which is the working definition of "intelligence" or more colloquially "smartness" according to most psychologists and especially psychometricians.
You will need to understand factor analysis for the "light bulb moment". IQ tests are the best metric we have as trying to proxy the g factor. Because of their strong 'statistical reliability' (Explanation on Wikipedia/IQ).
Those links assume that you know that IQ is tracking the g factor. So look at them in retrospect now.
Here are the mechanisms.
IQ tracks g.
g cannot be increased and is hereditary (probably genetic).
g correlates with many success metrics.
I can't do the thinking for you. But if you spend some time reflecting on those 3 facts. It will be evident why IQ is so useful as a metric.
Not at all. Psychometricians are well aware of those weak points and take great care to account for those. The strongest studies showing IQ's heredity (studies that track identical twins across their lifetimes) show that IQ is strongly hereditary.
These discrepancies also hold in different regions, different times, and different places. There really isn't any other explanation but to accept the signal that some groups of people score differently.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link