27 Comments
Jan 8, 2023Liked by Emil O. W. Kirkegaard

Even if the adoption of genetic enhancement technology for cognitive ability is slow or mostly among elites who can afford it, it looks like it would probably have significant society-wide effects. It's conceivable that this cognitive elite 5% becomes extraordinarily different, and hopefully, the relationship still holds.

Expand full comment
Jan 8, 2023Liked by Emil O. W. Kirkegaard

>The story of the theory is that it was originally proposed 100+ years ago by Terman, and then mentioned by Jensen and Weyl half a decade later. 

Half a century

Expand full comment
Jan 8, 2023Liked by Emil O. W. Kirkegaard

"...mentioned by Jensen and Weyl half a decade later..."

(Should be 'century' not 'decade'.)

Expand full comment

Elon Musk has like 10 kids from different women. He knows he has good genes and is doing his best to spread it. People reading this blog are probably at least 1 SD above the mean. Are any of you doing your part?

Expand full comment

The obvious conclusion is that we need a new "space race" but for germline engineering (or similar) of intelligence.

Expand full comment

"Our findings suggest that parents and teachers often fail to recognize the potential of poor and minority students and those with limited English proficiency."

Maybe some parents want there kids to help out with the family, ie: chores, child-bearing, farming, you know building a legacy and future together, and so sending your gifted Amish kid off to an elite boarding school so they can become a professor at Harvard would not be in the best interests of the parents or society.

Expand full comment

Argentina is surprisingly low (given their supposedly bulk of German/Italian ancestry)

Israel low average and not amazing 95th percentile - yet as a small country it does amazingly well in science and tech on a global scale

Yet NZ/Aus should be doing much better given their 120+ 95th percentile

I think the data is mostly the issue. But regardless in this case I dont see "smart fraction theory" being vindicated here. Actually the contrary

Expand full comment

Is there a way to get the SPI without the blatant leftist values (e.g. acceptance of gays, gender parity of education, discrimination against minorities etc.)

It would be interesting to know if it changes the scores materially

Expand full comment

"Our finding that intellectual class ability has incremental validity over and above average ability suggests that funding elite education may be the most effective way to boost national performance."

I think this is quite an intellectual jump. You haven't shown that elite education is causing the dispersion visible in any of the countries under consideration. You tend to have an environmentalist mindset (as do I) and are generally suspicious of claims that policy interventions in education have large effects at least at the margin. So why assume the opposite here?

I think elite education is useful and has its place. But we live in a world now where anyone with a curiosity about anything can access close to the sum total of human knowledge with a $200 smartphone. This is very different from when I was young when information was expensive and carefully rationed - yet still we educated ourselves!

Anyway you may well be right that elite education is good for all, but it's not considered in your paper.

Expand full comment
Jan 10, 2023·edited Jan 10, 2023

Hypothesis testing like this languishes for decades due to dumb-fraction midwit responses like "correlation doesn't imply causation" teaming up with the "smart fraction" among midwitticisms like "ecological correlations are invalid" teaming up with the truly smart fraction sociopaths that I've elsewhere called "genius birds of a different feather".

https://wiki.chadnet.org/the-parable-of-the-smart-birds

While it's all well and good to be a realist, the situation calls for a bit of meta-realism about the aforementioned situation.

Here's a suggestion:

Let's compile a corpus of public domain data sets in use by realists in their attempts to gain attention to the decline of societies. Submit them as pull requests to the my github repository Ockham's Guillotine, and then embark on constructing a unified macrosocial model based on the algorithmic information criterion for causal model selection.

https://github.com/jabowery/OckhamsGuillotine

If any serious activity along these lines were to occur, financing would likely flow to those who measurably improved the unified macrosocial model by that criterion. Once that started happening, the Ivy League priesthood would most likely be forced to respond with an explanation for why they couldn't contribute to such improvements.

Expand full comment

This matches my own view, countries with divergent genetic subcastes will have a smart fraction that pulls far away from its average. The order of important seems right. Average IQ is the most important, but smart fraction is non-trivial driver.

Also, the bottom 5% can probably be segregated to not cause too much trouble to the engines of growth. They can be unpleasant (ask Sweden about Muslims), but until they grow into enough of a mass to start pulling down the average and acting as a political force they can be isolated away from things that matter. This was what happened with blacks in the USA, some instances of local demographic superiority aside. Problem of course is 5% chain migrates the next 50%.

If the underclass is the majority you end up like India, which yeah has some good people in its smart fraction but is basically a shithole.

Expand full comment
Jan 8, 2023·edited Jan 8, 2023

Thanks for making these summaries. I wonder how much the HR lady effect - identity-based admission and hiring hurts the smart fraction if they are more male and white in countries prone to that. Maybe it just filters through the more conformist ones.

Expand full comment
Jan 8, 2023·edited Jan 8, 2023

Yes, it is not that non-obvious.

The hierarchy has always been: theoreticians, researchers and innovators, system maintainers (cognitive), leaders and managers, system maintainers/diagnosticians (physical), operators (cognitive), operators (physical), followers and consumers and useless in terms of the structuralization of society.

Theoreticians > +4Z or more = no coherent theory or unified theory of abstract structures - > higher trial and error rates, more time needed for innovation without clear 'path' for investments

No researchers and innovators > +3Z-+5Z= no new technologies, no new discoveries -> more automation, more efficient designs, higher energy-on-return-to-investment yields

No system Maintainers (cognitive) +1.75-2Z= = no one to copy, understand, design products and services associated with those new technologies, no system architects, no principles/doctrines, higher error rates,

higher failure rates

No leaders and managers (High-level) +1.5Z = inefficient allocation of resources, slow progress, more prone to error

No system maintainers/diagnosticians/technicians + 1Z-1.25Z = no one to maintain infrastructure of processes of systems, no manual followers

No operators (cognitive/physical) +0.5-1Z = no one to operate the infrastructure of said processes or comprehend the nature of them (i.e. airline cockpit buttons)

Followers/consumers -0.25Z-0.25Z = provides resources by doing inefficient labour. Capable of using end-products streamlined for ease-of-use by 3Z+ designers (i.e. compare computer punch-cards and internet in 80s to present smartphones [prev selected for +1Z or higher])

Useless = -0.75Z + = needs assistance to follow instructions or live day-to-day life in society given A level of technologies of certain complexity class

Without a certain fraction of such classes of people of moderately intelligent people (at absolute levels of IQ >95), regardless of how much more tail deviations IQ of the elite class is, it is meaningless. You cannot indoctrinate them into some political theory or religion to follow your whims, they do not follow instructions, are prone to making errors and generally do not accept authority without force, and corruption is everywhere because people cannot sacrifice present-day gains for future-day gains.

On the other hand, provided with a sufficient population of moderately intelligent people (i.e. IQ 95) the gains are synergistic with a higher fraction of high-intelligence people because the latter can design technologies and make systems that are more immune to error/failure, and the general population can use those services/products and maintain them without issue (i.e. able to assemble things from an Ikea manual).

Collinearity does need to be addressed by seeing how much more relative importance a higher proportion of threshold IQ is too, not just absolute difference from mean since what matters more is absolute IQ (g) rather than the relative distribution of relative abundances between smart-to-mean IQ normalized by X population though since moderately most countries are ethnically unified, these residual differential effects of differentially weighted populations are less pronounced (i.e. Jews/Chinese running the country, rest are Muslims/Blacks).

I'd like to see an R-index of civilizational achievement at X levels of IQ, or raw scores of intellect, given a sufficiently large enough population to maintain and operate those systems/processes/designs. I'd like to see a new religion/ideology for higher IQ people that self-stratify so such peoples can make their own societies in Space/other places and not be bothered constantly by having to maintain a population of lower IQ people (i.e. taxes and political theatre) which should reduce cognitive penalties (i.e. equivalent to aerodynamic drag).

Values such as wanting-power and wanting-wealth are also partially inherited which is why there are significant populations of higher IQ individuals who do not go into elite schools as it requires a continuous cumulative distribution of choosing X>Y peoples over generations to accumulate said wealth/status and also maintain, which also is mediated/attentuated by the selectiveness of each generation and the availability of fractions of equally stratified mates (i.e. caste pools and degree of horizontal socialization). Though with AI, brain-scans and black-box reverse engineering we may no longer need proscribed IQ tests or lengthy verbatim school degrees; probably just a DNA pGS-polygenetic score is enough combined with some externalizing variables to add confidence to such models.

The common core in the US and removal of standards are intentional dysgenic policies, although Microsoft and other tech companies are moving into the sphere of evaluating psy-behavioral social profiles of individuals en masse through BCI/rAI technologies while providing life-term learning tailored courses anyways. That has always been the goal to stratify the economically useful (by said elites). The modern education system has been deemed useless (and rightfully so). Knowledge multiplies many times every year and there is no possibility for any normal human being to master every field or even one field at a high level given our cognitive limitations.

Anyways even if in theory there is a population of IQ 200 (relative to present-day peoples), of some quantity like 10,000 (even though the observed abundance in a probablistic Gaussian distribution is infinitesimally minimal), it's unlikely anything they develop or made could be used by modern civilization without an increasing mean IQ population to comprehend the underlying principles behind such advanced technologies. (i.e. suppose you need to be able to operate your mind on a third-fuzzy level logic order to minimally know how to operate the systems, but you only had a 99.99% abundance of 120 IQ people; there is also a power-law distribution in my view given a certain level of complexity of class of problem to resolve, which is non-resolvable because lower IQ people goes beyond the 'memory buffer' when trying to evaluate all the variables in tandem -> similar parallel to why neural networks were discovered at least four decades ago but had little utility due to insufficient computing power. Likely that we need a higher DOF system than just silicon chips if we want to move to the next computing era)

Expand full comment

The figures here for Israel don't support the idea that Ashkenzi have IQs over 100 (having a very similar 50th and 95th percentile Israel(93,116) as Bulgaria(92, 115).

It would be great to see a post about Israeli IQs by sub population and smart fraction.

Expand full comment

At first I thought this said "simulation theory vindicated"

Expand full comment

More population means more cognitive elite. Didn't you also need to control this? For example China should be best in this model.

Expand full comment