I suspect that high IQ / high g is particularily important in understanding the natural world - which is rather more consistent than people. The STEM fields seem to have field specific minimum intelligence levels to master a given field - but 'success' in general tends to be more defined by mastering interactions with people, particularily in social groups. And people are irrational and inconsistent enough that facility with natural patterns is of limited utility (useful, but not dominant).
IQ is a better predictor of “success” than social skills, although I’m sure being impaired in social skills or IQ to a certain extent past the other makes either metric useless. We are all irrational, but our perceived rationality is influenced by reduced uncertainty of irrational reasoning, which having a higher IQ will allow you to do more of.
Social skills are a function of intelligence, personality and life experience while STEM mastery is a function of intelligence and interest. From my personal experience, people that are "pleasant but dummies" tend to be less successful socially, as being apt at social situations requires quite a comprehensive theory of what other people think and what motivates them.
A challenge I proposed for Taleb was for him to come up with a better measure than IQ which has 1) at least as much predictive validity for various important life outcomes as IQ, and 2) fulfills all of his absurd statistical standards that no one else has.
As you said, the other aspects of intelligence aren't captured by the IQ tests in part because they are abstract and difficult to measure; nonetheless, I'm curious to hear what you think the other salient components of intelligence are that aren't captured by IQ tests.
It’s been a long time since I read his writing on this, so if I am completely off base please call me out.
What I remember being most compelling was that the variation was not uniform. There was less variation in the lower end, but the variation increased with IQ, that is there were massive heteroskedasticity issues across the data when it came to IQ level and outcomes.
This is what fueled the r vs r-squared question. Since BLUE conditions are not satisfied for the regression, r-squared is not a meaningful statistic.
You can still look at correlation or r, hence why that would be the preferred statistic here. However its still missing a lot of critical information, but all conditions are met.
The main takeaway was that for IQs under 100, scores were highly predictive of outcomes and thus very meaningful. Tests are useful at identifying below average individuals. The problem was that above average individuals were much less easily identified as the variance between IQ and outcome was always very high.
>What I remember being most compelling was that the variation was not uniform. There was less variation in the lower end, but the variation increased with IQ, that is there were massive heteroskedasticity issues across the data when it came to IQ level and outcomes.
This does occur in the relationship between IQ and income, if I recall correctly. This has to do with IQ being a multiplier, and not an additive component (and income itself being multiplicative).
«A popular definition I’ve frequently seen is that “intelligence is the ability to get what you want out of life”.»
Life outcomes also would scale with "number * value of actual problems solved. And that's definitely not intelligence itself. However intelligence/IQ is often sold as the weaker "general problem solving ability", which need not necessarily imply good life outcomes, as the potential ability to do so might simply remain an underexploited capacity, due to temperament and circumstance. But I think even that is an overestimate and I think your definition of «ability to infer information from information» is much more accurate.
Equating this with "general problem solving ability" neglects that the difficulty of a problem need not lie in "unknown/uncertain information" at all, or only to a certain point.
As Ayn Rand puts it:
“Action without thought is mindlessness, and thought without action is hypocritical.”
Being more intelligent (as in your definition), one will be tempted towards more thought, even long past the point, where action is the optimal move. A "hypocrite" cannot be truly capable.
I suspect that high IQ / high g is particularily important in understanding the natural world - which is rather more consistent than people. The STEM fields seem to have field specific minimum intelligence levels to master a given field - but 'success' in general tends to be more defined by mastering interactions with people, particularily in social groups. And people are irrational and inconsistent enough that facility with natural patterns is of limited utility (useful, but not dominant).
IQ is a better predictor of “success” than social skills, although I’m sure being impaired in social skills or IQ to a certain extent past the other makes either metric useless. We are all irrational, but our perceived rationality is influenced by reduced uncertainty of irrational reasoning, which having a higher IQ will allow you to do more of.
Social skills are a function of intelligence, personality and life experience while STEM mastery is a function of intelligence and interest. From my personal experience, people that are "pleasant but dummies" tend to be less successful socially, as being apt at social situations requires quite a comprehensive theory of what other people think and what motivates them.
A challenge I proposed for Taleb was for him to come up with a better measure than IQ which has 1) at least as much predictive validity for various important life outcomes as IQ, and 2) fulfills all of his absurd statistical standards that no one else has.
I hope the grand theory includes the lexical hypothesis
As you said, the other aspects of intelligence aren't captured by the IQ tests in part because they are abstract and difficult to measure; nonetheless, I'm curious to hear what you think the other salient components of intelligence are that aren't captured by IQ tests.
It’s been a long time since I read his writing on this, so if I am completely off base please call me out.
What I remember being most compelling was that the variation was not uniform. There was less variation in the lower end, but the variation increased with IQ, that is there were massive heteroskedasticity issues across the data when it came to IQ level and outcomes.
This is what fueled the r vs r-squared question. Since BLUE conditions are not satisfied for the regression, r-squared is not a meaningful statistic.
You can still look at correlation or r, hence why that would be the preferred statistic here. However its still missing a lot of critical information, but all conditions are met.
The main takeaway was that for IQs under 100, scores were highly predictive of outcomes and thus very meaningful. Tests are useful at identifying below average individuals. The problem was that above average individuals were much less easily identified as the variance between IQ and outcome was always very high.
>What I remember being most compelling was that the variation was not uniform. There was less variation in the lower end, but the variation increased with IQ, that is there were massive heteroskedasticity issues across the data when it came to IQ level and outcomes.
This does occur in the relationship between IQ and income, if I recall correctly. This has to do with IQ being a multiplier, and not an additive component (and income itself being multiplicative).
«A popular definition I’ve frequently seen is that “intelligence is the ability to get what you want out of life”.»
Life outcomes also would scale with "number * value of actual problems solved. And that's definitely not intelligence itself. However intelligence/IQ is often sold as the weaker "general problem solving ability", which need not necessarily imply good life outcomes, as the potential ability to do so might simply remain an underexploited capacity, due to temperament and circumstance. But I think even that is an overestimate and I think your definition of «ability to infer information from information» is much more accurate.
Equating this with "general problem solving ability" neglects that the difficulty of a problem need not lie in "unknown/uncertain information" at all, or only to a certain point.
As Ayn Rand puts it:
“Action without thought is mindlessness, and thought without action is hypocritical.”
Being more intelligent (as in your definition), one will be tempted towards more thought, even long past the point, where action is the optimal move. A "hypocrite" cannot be truly capable.