How Stifling Debate Around Race, Genes and IQ Can Do Harm (2018)

I find different cultures/religions to be quite interesting in their characteristics of families. Mormons seem to have larger families. They require 10% of income goes to the church which increases the church's influence, so the church gets stronger the more people are in it. It's like a self-sustaining organism that gets stronger with each generation.

As an intelligent individual belonging to no particular culture or religion, I feel less inclined to have children than others I know since it's not part of my values. It's likely that people like myself are going to have a difficult time passing on our genetics. If you want to encourage smart people to have kids, create a religion/culture that makes sense to them and give them that directive.

You can characterize and predict how humans will perform in many, many games with this single variable. It's the one that works best. So the current thinking is that there is only one measurement of intelligence. They all converge into a single one. It's called IQ because that was the first one that worked well enough, not because there's just one way to measure it.

(Technically there is also age. Younger people are dumber than older ones up to a pretty high age. But that's the IQ game: age and "raw iq". Nothing else really matters in predicting if someone can succeed at something. IQ, the measure used, is corrected for age, so it should take that out. But a 10 year old with 100 IQ is, in absolute terms, dumber than a 20 year old with IQ 100)

So your criticism is rather a lot beside the point. We have worked, a long and hard and often on developing better systems. We have just never actually succeeded in creating a truly different system (this of course is a "technically incorrect" statement. There are recent research results, there are ... blah blah blah ... there is no halfway decent reason to assume those research directions will lead to anything, but we're still trying. We've been doing that for >100 years, so the odds of the next big thing being in the pipeline right now is small).

The basic observation behind IQ is that "all skills converge" (I would argue this has even been observed in machine learning). If you know one skill, whichever one it is, you become faster, better and quicker to learn almost any other skill. Not much, a little. In machine learning there are papers that point out that the ability to distinguish poodle races in pictures help with music composition. Not "much", but the key is what happens over time. It helps something like 2-5% (measured in time to learn a certain proficiency in music composition). The issue is what happens with kids when you give them new skills regularly over a 10 year period. Then they have 100 more skills, or 100 times 2% easier time to learn a new skill compared to someone who learned nothing. 100 times 2% easier ... is (1.02^100)-1, which is an enormous number (which is why IQ is an exponential score, not a linear one. Someone with 120 should be about double as smart as someone scoring 100. By which we mean that if you take a random intellectual skill, the odds of the person with 120 having some proficiency at that skill is about double that of the person with 100). To add insult to injury IQ compounds on itself. The higher your IQ, the easier it is to get it even higher.

What people hate about IQ is the underlying theory clearly shows that human brains are designed to amplify differences in intelligence over time, not to minimize them. That is a clear and incontestable fact at this point. This is presumably also not a bad thing, or it would have been selected out. This also won't change because of any amount of changes occurring in psychology or the measurement of it. We can prevent ourselves from knowing about it, but we can't change it.