If I recall correctly, black people tend to score higher than whites on lots of Good Person qualities. They tend to have high self esteem, low neuroticism, tend to be more relaxed and happy etc. If AI or other "intelligence assistive" technologies can raise the floor, then I'm sure black people will benefit a lot. If AI makes everyone a good writer and good reasoner and so on, then wouldn't the remaining qualities become more important? I.e. if both job candidates can write and reason just as well, I'd rather hire the chill black guy over the twitchy white guy.
From what I can tell reading you and other Substack writers, the polygenic scores for intelligence presume that the genes add up linearly to produce intelligence. This is the equivalent of a neural net model which has a single neuron.
A more complex net (or scientific hypothetical framework) could take into account the possibility that some genetic variants add to IQ in some contexts and subtract from IQ in other contexts, where by context I mean other gene variants.
To illustrate, let's replace gene variants with traits. A high energy trait could enhance intellect by resulting in more awake time in the classroom, more study time, etc. Or this trait could enhance impatience with the system, lead to delinquency, etc. Or consider a trait of thinking out of the box. It could lead to creativity, inventions, etc. Or it can lead to picking the wrong box on multiple choice tests by using reasoning not anticipated by the creator of the test.
----
I'm using a non-g model of intelligence. IQ is a measure of ability to do well on IQ tests. There are many separable traits which contribute to the ability to score well on a standardized test: good pattern matching skills, cultural similarity to the creator of the test, ability to abstract, interest in doing puzzles, absence of typos and related brain hiccups, introversion, raw brain power, interest in reading widely, ability to reason while under the pressure of a timed high stakes test.
Let's focus on just that last trait. Back when I was young I did well in standardized tests and courses with high stakes exams. My aggression was wired into my ability to think. I thus craved activities like speed chess and competitive debate. On the other hand, many otherwise smart people lose it under such pressure. Adrenaline results in fight or flight instead of turbo charged thinking. The valedictorian in my high school scored below 1000 on the SAT for this reason; she was beaten by 200 points by a C student.
Keep this line of causality in mind when looking at the correlation between IQ and criminality. The same emotional response that can lead to a lower IQ score can also lead to a crime of passion. (Separately, cleverness can lead to not being caught for doing criminal acts, which throws off the stats.)
You're onto something here. I tried explaining it in the other thread as emergent properties of complex systems but the analogy of a neural network is good. Of course it will only work for someone who knows neural networks. I don't think you need to explain it in terms of countervailing traits. We could just compare the node weights of GPT with a similar model like Claude. I guarantee you will not find similarity in the weights. it probably depends on where you start and what you train on and each model will not have 10 key nodes/weights. It won't have 100. It won't have a thousand. No human mind would ever be able to tell you that node 10010176 is weighted 0.7888 because node 3662918 is at 0.23332 and, BTW if the first node had a slightly different value, chaos theory would predict that you couldn't predict if the other would go up or down let alone how much.
We might be talking past each other. I'm not saying we need to model the human brain at the neuron level. I'm saying that the analysis of the mapping of genes to intelligence is equivalent of pattern matching using the most primitive neural net possible -- basically modelling everything down to one dimension.
Neural networks of the feed forward variety [assuming I'm keeping the terminology correct] are basically a curve fit in a potentially very high dimensional space using crude wavelets. The space starts off as the number of dimensions equal to the number of features, in this case the number of genes considered. It makes sense to take this down to a much smaller dimensional space with the first layer, but taking it immediately down to one dimension can lead to overly simplistic fits.
Point being: you can have plenty of genetic determinism and miss most of it if you try to fit the data to too simple a curve family.
While genes don't map as cleanly to traits as high school biology used to teach, the concept of thinking of IQ as a function of multiple traits is useful.
Including good looks -- as the phrenologists of yore taught.
I didn't think you meant we needed to model the brain at the neuron level. I knew you were talking about computer neural networks. WRT overfitting, I'm not an expert but I think of that as more a problem with other machine learning techniques.
That said I thought you were just making an analogy about the complexity, not overfitting. I think it's a good analogy if you are talking about the complexity and how impossible it is to relate any given set of nodes in the network to any given output, because it's the insanely complex interaction between the nodes that does it. There's no good nodes and better nodes in a neural network, just like intelligence might not be the summation of a million slightly better intelligence genes.
The polygenic score model is an extreme under fit. It's rather like trying to do linear regression on a sine wave for an older metaphor.
(I spent a couple of years applying neural nets and other machine learning algorithms at financial data. Over fitting is a big issue with neural nets.)
I guess I know you can have overfitting with neural nets. It's more the problems I'm used to people solving with them, I guess already take this into account and don't do it without lots of training data. It's intuitively obvious to me that you can't, e.g, train a neural net to predict the stock market.
Back in college a friend told me his professor was working on predicting the stock market by treating it as a wave function and using fourier transforms to fit it. I told him no way that could work and he yelled at me for my arrogance, a student, questioning a professor! I stood by it You only have a fraction of the curve. how can you figure out the wave function?
Back to the original point, I don't care whether or how much a polygenic score works in predicting intelligence. It only matters to me if it helps us find intelligence genes that would could, perhaps exploit in some way.
I'm questioning whether such genes exist. The idea people have is that there are a million intelligence genes and each codes for a protein or regulator slightly better than average and the summation of all these slightly better proteins is more intelligence.
What would these genes be? Faster sodium channels? Better vesicles full of neurotransmitters? I doubt it. I think intelligence is more down to architecture and connections which implies that it's more embryology and development. I don't think we are close to having a handle on how genes relate to that sort of thing.
There was a point in Markieff and Marcus Morris’ careers where they had almost identical total games and minutes played and total points, and it was like 10 years (and 600 games~) into their careers despite playing in totally different situations in regards to team and scheme. They also graded out as basically identical players, big surprise!
I always loved this fact and bring it up to my dad whenever we talk about monozygotic twins and their traits.
I find the relatively small interracial gap in educational achievement far less important than the huge interracial gap in murder and other violent crimes.
African-american PISA scores are on par with various eurasian nations that have far lower murder and incarceration rates.
I get that IQ tests and school tests are far easier to create and administer than those that measure psychopathy or impulse control disorder or whatever drives this type of saturday night crimes so we will not be getting answers anytime soon about why blacks are so violent but this is one of the most important HBD questions.
The murder gap might be a lot more nurture than the IQ gap. Other races have gone through violent periods and honor killings. Lots have grown out of it.
I'm starting to wonder whether IQ testing, despite (or perhaps because of) its apparent accuracy and replicability, isn't more trouble than it's worth.
Just look at the enormous efforts inspired by it's refutation -- efforts that have had incalculable negative effects on us and our children.
Maybe it would be better to just say "ok, we give up, IQ is meaningless," and thereby take the wind out of the sails of the "environment is everything" religious zealots. If the alternative is spending 40 trillion on achieving parity, is it really worth arguing with these people?
Per Steve you'd still have the same argument about every even slightly g-loaded test like the SAT and the fireman lieutenant exam. There'd be no MCAT so you'd depend on the principles and bravery of med school professors to keep dangerously dull witted people from becoming your doctor. A recent article about UCLA (and anecdotes from someone I know who is a long time professor at a top medical school) suggest that wouldn't work out very well.
> Maybe it would be better to just say "ok, we give up, IQ is meaningless," and thereby take the wind out of the sails of the "environment is everything" religious zealots. If the alternative is spending 40 trillion on achieving parity, is it really worth arguing with these people? <
Bill, this is a confusion of correlation and causation. The anti-genetic nonsense, the 40 trillion and the endless busybodying have nothing to do with the invention of IQ testing. They are products of the managerial super-state and minoritarian (anti-majority, anti-white, anti-national) ideology.
And as Erik points out *any* testing shows the same gaps. Any sort of grades or evaluation of any kind of intellectual work shows the same gaps. As Murray and Herrnstein pointed out, pretty much any socio-economic measure of success shows the same gaps.
The war on genetics is a war upon white majorities, upon the right of whites to have white nations, upon civilization itself. There isn't any "let's concede X" that buys you anything. You have to defeat them--convince them or kill them--or the West dies.
Perhaps education for some people should deemphasize trying to get them into college and start trying to keep them out of the pen, addiction, obesity, and debt. The virtues of recycling could be an elective.
My daughter was an assistant teacher at a private Montessori school in Washington DC. The Montessori method is a great way to educate the children of the upper class. DC created a Montessori charter school and she got a job there. She quickly realized that the public school students were drastically different from the private school kids.
The issue is: If blacks, on average, have lower cognitive ability than non-Whites who have less than Korean-Americans what policies should be adopted. The Very right of center idea that one just assumes all blacks are idiots does not work.
Fabius beat me to it, but I think a lot of the "missing heritability" is going to be from complex non-additive effects.
Compared to something like height--hey, let's keep growing this bone longer--intelligence is really, really complicated. I'm not a bio guy, much less and neurology guy but just off the top of my head--brain volume, brain structure, size of various regions, neural density, neural connection density (I don't know the term, but how many connections tend to be made), neuron speed, all the various neurotransmitters amounts, relative proportions, any mutations.
This is all very complex. And the results of "neural package" works to make someone intelligent is complex. (I know my three kids, though all reasonably smart--with very good, to excellent SAT scores--are quite different in their personalities and in how their brains work and what they are good at.) Spearman's "g" strikes me as mostly the point that testing mostly just reveals some overall level of quality brain functioning--i.e. high functioning brains do all brain tasks well, low functioning ones, not so much. But that doesn't mean that all high-functioning brains are that way for the same reasons, much less that you can just toss all the alleles into a pile and count up their contributions.
But we have the data from these identical twins raised apart thing and it's pretty darn clear that given a "reasonable" environment--adequate nutrition, non-abuse, lack of impairment by disease, availability of education to learn to read and do math--mental *capability* is mostly genetic. Or at least genes plus womb development.
If the genetic studies aren't prying out the heritability it is not because the genetic contribution is not there. It means those studies--as yet--suck at pulling it out. Probably because they are picking at something quite complex.
Deleterious, rare variants (a.k.a. mutational load) need not have additive effects. The argument for additivity applies to functional variation, stuff that works correctly (for what it is).
Rare variants are also hard to see for these GWAS studies.
If you had different mutations of a bad kind slightly hampering the correct functioning of your long term memory, axons, brain development&architecture, plasticity, etc., the multiplicative effects for the success rate of having good thoughts could be big. Like the game "broken phone".
As for the missing heritability for disease... I think the "non-shared environment" for identical twins has to be smaller. Or should that be called gene-environment interaction? I don't know. Anyway, MZ twins are much more likely to pick up the same pathogens from their environment.
It would be very reasonable to propose focusing on just raising the performance of each demographic group rather than chasing parity between them. Unfortunately it also would mean society accepting permanent economic - among other - gaps and we are still too close in time to the civil rights era and its ideas for that to be acceptable to a critical mass of the public. But it’s likely to happen over time.
It is difficult to believe that this kind of insanity can persist for so many decades. Thanks to social media, race realism is becoming ever more common. I keep hoping that we will reach a tipping point when a sufficiently high fraction of the population are made aware of the realities of race.
As has been pointed out many times, the county in California with the third highest violent crime rate is an 80% white county. Yet, social media never seems to notice. Social media cannot even accurately place NYC in the correct ranking for homicides.
The NBA is probably my least favorite major sports league but Jason Collins, a twin, once played for the New Jersey Nets. Near the end of his career he came out of the closet then re-signed with the Nets who by then moved to Brooklyn. I remember making a point of watching his first game back, which was at the Los Angeles Lakers. Commentators Ian Eagle and Mike Fratello didn't say anything false about him but they both went out of their way to praise his play in a way they wouldn't for any other random signing.
His younger twin Jarron had a less successful (but still long) career as a player but has won three rings as an assistant coach with the Warriors. Both twins are alumni of Harvard-Westlake which Steve has some sort of connection to; their teammate was the actor Jason Segel
Obviously, no government program will come up with the "Missing Malleability" because mostly it does not exist.
However, what would help enormously--and would help blacks especially enormously--is to end all the progressive nonsense. The single most beneficial thing is to end the immivasion nightmare and firm up the labor market (and "firm down" the housing market). But beyond that simply "restoring order"--on the streets, in the schools--would be exceptionally helpful.
As I've noted before the biggest environmental--and "structural"--problem blacks face is that they mostly live around other blacks. Ergo live--on average--with much more crime, drugs and general disorder. A return to self-confident enforcement of white normative standards in public environments--the schools, the streets--would be huge boon to working class blacks who are trying to get on with life, but are continually dealing with crap imposed upon them by the criminal and layabout blacks around them.
So what have "progressives" been doing? Just the reverse, tell the most worthless blacks that they are entitled to act out.
```
Of course, the real public-policy win is not environmental at all but genetic. Stop the immivasion and start doing eugenics--stop the welfare cases from having kids, lockup, kill or expel the criminals so they can't have kids (in America). American blacks would benefit--the population would improve--more from reasonable--low hanging fruit--eugenic policy even than American whites.
As you pointed out, too many believe the solution is a Final Solution.
Conventional Wisdom: "How can we possibly live in a world where races are not exactly equal genetically without turning to mass genocide?"
Steve: "Uh, judging from what I see on ESPN, we live in such a world just fine."
If I recall correctly, black people tend to score higher than whites on lots of Good Person qualities. They tend to have high self esteem, low neuroticism, tend to be more relaxed and happy etc. If AI or other "intelligence assistive" technologies can raise the floor, then I'm sure black people will benefit a lot. If AI makes everyone a good writer and good reasoner and so on, then wouldn't the remaining qualities become more important? I.e. if both job candidates can write and reason just as well, I'd rather hire the chill black guy over the twitchy white guy.
AI only raises the floor. It is nowhere near genius levels.
From what I can tell reading you and other Substack writers, the polygenic scores for intelligence presume that the genes add up linearly to produce intelligence. This is the equivalent of a neural net model which has a single neuron.
A more complex net (or scientific hypothetical framework) could take into account the possibility that some genetic variants add to IQ in some contexts and subtract from IQ in other contexts, where by context I mean other gene variants.
To illustrate, let's replace gene variants with traits. A high energy trait could enhance intellect by resulting in more awake time in the classroom, more study time, etc. Or this trait could enhance impatience with the system, lead to delinquency, etc. Or consider a trait of thinking out of the box. It could lead to creativity, inventions, etc. Or it can lead to picking the wrong box on multiple choice tests by using reasoning not anticipated by the creator of the test.
----
I'm using a non-g model of intelligence. IQ is a measure of ability to do well on IQ tests. There are many separable traits which contribute to the ability to score well on a standardized test: good pattern matching skills, cultural similarity to the creator of the test, ability to abstract, interest in doing puzzles, absence of typos and related brain hiccups, introversion, raw brain power, interest in reading widely, ability to reason while under the pressure of a timed high stakes test.
Let's focus on just that last trait. Back when I was young I did well in standardized tests and courses with high stakes exams. My aggression was wired into my ability to think. I thus craved activities like speed chess and competitive debate. On the other hand, many otherwise smart people lose it under such pressure. Adrenaline results in fight or flight instead of turbo charged thinking. The valedictorian in my high school scored below 1000 on the SAT for this reason; she was beaten by 200 points by a C student.
Keep this line of causality in mind when looking at the correlation between IQ and criminality. The same emotional response that can lead to a lower IQ score can also lead to a crime of passion. (Separately, cleverness can lead to not being caught for doing criminal acts, which throws off the stats.)
You're onto something here. I tried explaining it in the other thread as emergent properties of complex systems but the analogy of a neural network is good. Of course it will only work for someone who knows neural networks. I don't think you need to explain it in terms of countervailing traits. We could just compare the node weights of GPT with a similar model like Claude. I guarantee you will not find similarity in the weights. it probably depends on where you start and what you train on and each model will not have 10 key nodes/weights. It won't have 100. It won't have a thousand. No human mind would ever be able to tell you that node 10010176 is weighted 0.7888 because node 3662918 is at 0.23332 and, BTW if the first node had a slightly different value, chaos theory would predict that you couldn't predict if the other would go up or down let alone how much.
We might be talking past each other. I'm not saying we need to model the human brain at the neuron level. I'm saying that the analysis of the mapping of genes to intelligence is equivalent of pattern matching using the most primitive neural net possible -- basically modelling everything down to one dimension.
Neural networks of the feed forward variety [assuming I'm keeping the terminology correct] are basically a curve fit in a potentially very high dimensional space using crude wavelets. The space starts off as the number of dimensions equal to the number of features, in this case the number of genes considered. It makes sense to take this down to a much smaller dimensional space with the first layer, but taking it immediately down to one dimension can lead to overly simplistic fits.
Point being: you can have plenty of genetic determinism and miss most of it if you try to fit the data to too simple a curve family.
While genes don't map as cleanly to traits as high school biology used to teach, the concept of thinking of IQ as a function of multiple traits is useful.
Including good looks -- as the phrenologists of yore taught.
I didn't think you meant we needed to model the brain at the neuron level. I knew you were talking about computer neural networks. WRT overfitting, I'm not an expert but I think of that as more a problem with other machine learning techniques.
That said I thought you were just making an analogy about the complexity, not overfitting. I think it's a good analogy if you are talking about the complexity and how impossible it is to relate any given set of nodes in the network to any given output, because it's the insanely complex interaction between the nodes that does it. There's no good nodes and better nodes in a neural network, just like intelligence might not be the summation of a million slightly better intelligence genes.
The polygenic score model is an extreme under fit. It's rather like trying to do linear regression on a sine wave for an older metaphor.
(I spent a couple of years applying neural nets and other machine learning algorithms at financial data. Over fitting is a big issue with neural nets.)
I guess I know you can have overfitting with neural nets. It's more the problems I'm used to people solving with them, I guess already take this into account and don't do it without lots of training data. It's intuitively obvious to me that you can't, e.g, train a neural net to predict the stock market.
Back in college a friend told me his professor was working on predicting the stock market by treating it as a wave function and using fourier transforms to fit it. I told him no way that could work and he yelled at me for my arrogance, a student, questioning a professor! I stood by it You only have a fraction of the curve. how can you figure out the wave function?
Back to the original point, I don't care whether or how much a polygenic score works in predicting intelligence. It only matters to me if it helps us find intelligence genes that would could, perhaps exploit in some way.
I'm questioning whether such genes exist. The idea people have is that there are a million intelligence genes and each codes for a protein or regulator slightly better than average and the summation of all these slightly better proteins is more intelligence.
What would these genes be? Faster sodium channels? Better vesicles full of neurotransmitters? I doubt it. I think intelligence is more down to architecture and connections which implies that it's more embryology and development. I don't think we are close to having a handle on how genes relate to that sort of thing.
There was a point in Markieff and Marcus Morris’ careers where they had almost identical total games and minutes played and total points, and it was like 10 years (and 600 games~) into their careers despite playing in totally different situations in regards to team and scheme. They also graded out as basically identical players, big surprise!
I always loved this fact and bring it up to my dad whenever we talk about monozygotic twins and their traits.
I find the relatively small interracial gap in educational achievement far less important than the huge interracial gap in murder and other violent crimes.
African-american PISA scores are on par with various eurasian nations that have far lower murder and incarceration rates.
I get that IQ tests and school tests are far easier to create and administer than those that measure psychopathy or impulse control disorder or whatever drives this type of saturday night crimes so we will not be getting answers anytime soon about why blacks are so violent but this is one of the most important HBD questions.
The murder gap might be a lot more nurture than the IQ gap. Other races have gone through violent periods and honor killings. Lots have grown out of it.
I'm starting to wonder whether IQ testing, despite (or perhaps because of) its apparent accuracy and replicability, isn't more trouble than it's worth.
Just look at the enormous efforts inspired by it's refutation -- efforts that have had incalculable negative effects on us and our children.
Maybe it would be better to just say "ok, we give up, IQ is meaningless," and thereby take the wind out of the sails of the "environment is everything" religious zealots. If the alternative is spending 40 trillion on achieving parity, is it really worth arguing with these people?
Per Steve you'd still have the same argument about every even slightly g-loaded test like the SAT and the fireman lieutenant exam. There'd be no MCAT so you'd depend on the principles and bravery of med school professors to keep dangerously dull witted people from becoming your doctor. A recent article about UCLA (and anecdotes from someone I know who is a long time professor at a top medical school) suggest that wouldn't work out very well.
> Maybe it would be better to just say "ok, we give up, IQ is meaningless," and thereby take the wind out of the sails of the "environment is everything" religious zealots. If the alternative is spending 40 trillion on achieving parity, is it really worth arguing with these people? <
Bill, this is a confusion of correlation and causation. The anti-genetic nonsense, the 40 trillion and the endless busybodying have nothing to do with the invention of IQ testing. They are products of the managerial super-state and minoritarian (anti-majority, anti-white, anti-national) ideology.
And as Erik points out *any* testing shows the same gaps. Any sort of grades or evaluation of any kind of intellectual work shows the same gaps. As Murray and Herrnstein pointed out, pretty much any socio-economic measure of success shows the same gaps.
The war on genetics is a war upon white majorities, upon the right of whites to have white nations, upon civilization itself. There isn't any "let's concede X" that buys you anything. You have to defeat them--convince them or kill them--or the West dies.
Perhaps education for some people should deemphasize trying to get them into college and start trying to keep them out of the pen, addiction, obesity, and debt. The virtues of recycling could be an elective.
My daughter was an assistant teacher at a private Montessori school in Washington DC. The Montessori method is a great way to educate the children of the upper class. DC created a Montessori charter school and she got a job there. She quickly realized that the public school students were drastically different from the private school kids.
"So, it’s been a two-generation long conundrum with no solution in sight."
Rejecting reality is hard.
The issue is: If blacks, on average, have lower cognitive ability than non-Whites who have less than Korean-Americans what policies should be adopted. The Very right of center idea that one just assumes all blacks are idiots does not work.
Fabius beat me to it, but I think a lot of the "missing heritability" is going to be from complex non-additive effects.
Compared to something like height--hey, let's keep growing this bone longer--intelligence is really, really complicated. I'm not a bio guy, much less and neurology guy but just off the top of my head--brain volume, brain structure, size of various regions, neural density, neural connection density (I don't know the term, but how many connections tend to be made), neuron speed, all the various neurotransmitters amounts, relative proportions, any mutations.
This is all very complex. And the results of "neural package" works to make someone intelligent is complex. (I know my three kids, though all reasonably smart--with very good, to excellent SAT scores--are quite different in their personalities and in how their brains work and what they are good at.) Spearman's "g" strikes me as mostly the point that testing mostly just reveals some overall level of quality brain functioning--i.e. high functioning brains do all brain tasks well, low functioning ones, not so much. But that doesn't mean that all high-functioning brains are that way for the same reasons, much less that you can just toss all the alleles into a pile and count up their contributions.
But we have the data from these identical twins raised apart thing and it's pretty darn clear that given a "reasonable" environment--adequate nutrition, non-abuse, lack of impairment by disease, availability of education to learn to read and do math--mental *capability* is mostly genetic. Or at least genes plus womb development.
If the genetic studies aren't prying out the heritability it is not because the genetic contribution is not there. It means those studies--as yet--suck at pulling it out. Probably because they are picking at something quite complex.
Deleterious, rare variants (a.k.a. mutational load) need not have additive effects. The argument for additivity applies to functional variation, stuff that works correctly (for what it is).
Rare variants are also hard to see for these GWAS studies.
If you had different mutations of a bad kind slightly hampering the correct functioning of your long term memory, axons, brain development&architecture, plasticity, etc., the multiplicative effects for the success rate of having good thoughts could be big. Like the game "broken phone".
As for the missing heritability for disease... I think the "non-shared environment" for identical twins has to be smaller. Or should that be called gene-environment interaction? I don't know. Anyway, MZ twins are much more likely to pick up the same pathogens from their environment.
Aren't we already relying on the principles and bravery of professors?
How's that working out, btw?
Maybe IQ really is overrated...
It would be very reasonable to propose focusing on just raising the performance of each demographic group rather than chasing parity between them. Unfortunately it also would mean society accepting permanent economic - among other - gaps and we are still too close in time to the civil rights era and its ideas for that to be acceptable to a critical mass of the public. But it’s likely to happen over time.
It is difficult to believe that this kind of insanity can persist for so many decades. Thanks to social media, race realism is becoming ever more common. I keep hoping that we will reach a tipping point when a sufficiently high fraction of the population are made aware of the realities of race.
As has been pointed out many times, the county in California with the third highest violent crime rate is an 80% white county. Yet, social media never seems to notice. Social media cannot even accurately place NYC in the correct ranking for homicides.
The NBA is probably my least favorite major sports league but Jason Collins, a twin, once played for the New Jersey Nets. Near the end of his career he came out of the closet then re-signed with the Nets who by then moved to Brooklyn. I remember making a point of watching his first game back, which was at the Los Angeles Lakers. Commentators Ian Eagle and Mike Fratello didn't say anything false about him but they both went out of their way to praise his play in a way they wouldn't for any other random signing.
His younger twin Jarron had a less successful (but still long) career as a player but has won three rings as an assistant coach with the Warriors. Both twins are alumni of Harvard-Westlake which Steve has some sort of connection to; their teammate was the actor Jason Segel
Obviously, no government program will come up with the "Missing Malleability" because mostly it does not exist.
However, what would help enormously--and would help blacks especially enormously--is to end all the progressive nonsense. The single most beneficial thing is to end the immivasion nightmare and firm up the labor market (and "firm down" the housing market). But beyond that simply "restoring order"--on the streets, in the schools--would be exceptionally helpful.
As I've noted before the biggest environmental--and "structural"--problem blacks face is that they mostly live around other blacks. Ergo live--on average--with much more crime, drugs and general disorder. A return to self-confident enforcement of white normative standards in public environments--the schools, the streets--would be huge boon to working class blacks who are trying to get on with life, but are continually dealing with crap imposed upon them by the criminal and layabout blacks around them.
So what have "progressives" been doing? Just the reverse, tell the most worthless blacks that they are entitled to act out.
```
Of course, the real public-policy win is not environmental at all but genetic. Stop the immivasion and start doing eugenics--stop the welfare cases from having kids, lockup, kill or expel the criminals so they can't have kids (in America). American blacks would benefit--the population would improve--more from reasonable--low hanging fruit--eugenic policy even than American whites.
Lyman Stone is on a crusade to convince everyone that twin studies do not prove genetic heritability https://open.substack.com/pub/lymanstone/p/more-evidence-twin-studies-are-bad?r=1yh61&utm_medium=ios
That doesn’t pass the common sense test.