>This has profound implications for the ability of the movement to work in the interests of all humanity. For example, EAs who believe it is important that the human population expand but who think that African people are likely to be stupider than the human average may divert resources away from them. EAs making hiring decisions may prioritise applications from people whom they have not subconsciously assigned to the schema "stupider than me."
Let's say that it is true that not only are africans today lower in cognitive ability than whites on average, but that this gap is also mostly heritable (the latter is debatable but there is strong evidence to support it - the former is indisputable unless you think literally every single measure of cognitive ability that has strong predictive validity is radically and fatally flawed) .
Should this be acknoledged? Should this affect how we view the world?
Assuming it's true, do you believe, as a rationalist, that ignoring fundmental characteristics of human ity that explain more about society than almost any other factors will lead you towards making better decisions?
Speaking only for myself here, I think the answer is an unqualified hell no.
If you assume African Americans are congitive equivlanet to Asian Americans on average, then you will assume that sending more of them to college (in excess of those who can get in on the basis of academic merit alone currently) will result in improved socioeconomic outcomes (even if no affirmative action in hiring existed). But the real result is that many black people will have wasted 2-4 years of their life accumulating debt and forgoing earnings to gain an "education" that did nothing to close the cognitive gaps between themselves and asian americans. The debt, loss of income and wasting of time that could have been used developing valuable skills will hurt these graduates.
If instead, given that we know that intelligence varies between people, and we're interested in maximising the earning potential of black kids from low income families, we had identified their IQ early on and put our effort into getting them onto the optimal career track for their specific level of cognitive ability.
(The same is true for kids of any race - the difference is people are generally much more accepting of the fact that some white kids are just plain less intelligent than other white kids and can't have the same careers, and much more of acepting income inequality within than between races. And so there is a much bigger push to get black people specifically into college rather than low income/IQ kids generally, which is why it makes sense to think along racial lines here).
The same logic applies to entire countries. If you assume that Nigerians are no less intelligent than Chinese people, then you're likely to expect that Nigeria is going to follow the same development path as China once some threshold is passed for e.g. education health and so on. This is almost certainly not going to happen, and expecting it to will mean you will make suboptimal decisions when thinking about how to help countries in Africa.
More controversially, it also would mean that people in Africa are likely going to be of much lower instrumental value to humanity than people in Japan. But choosing to ignore this is actually the decision that is being made on the basis of race. Choosing to factor this into your worldview and decisions means you're acting on the basis of utility alone and not making decisions aimed at helping people because of their race more than you would of.
Liberal americans obviously aren't categorically opposed to helping low income black americans more than low income white americans in rural areas/the south, and it's likely this is because they think that black americans are better people and have more to offer the country, whereas white americans are just dumb and ignorant and cannot be reformed (not everyone believes this entirely, but I think some implicit form of this belief colors much of liberal thought in this area). All I'm saying is this type of thinking also applies more generally, except in this case the difference is based on data rather than intution and in-group favoratism.
If you, as you say, care about "working in the interests of humanity", you need to think about the interests of humanity collectively. Maximixing the interests of humanity will necessarily involve not helping all groups equally.
>Across large racial groups, however, the observed differences are statistically significant but modest (as they are for other anthropometric characters)
The balck white IQ gap is approximately one standard deviation, which is by no means modest. If we could raise a population's mean IQ by one standard deviation, this would be utterly transformative in a way that almost no other intervention could be.
I'm not saying it's not modest because its not a small number - it's not modest because such a gap has profound social implications in a way that black people being slightly taller than asian people doesn't. Controlling for IQ literally eliminates most of the inequality between races in things like income.
And it's not just a difference in the mean. Black people have a lower variance in IQ, meaning there are less black people 3 standard deviations above the black mean than there are asian people 3 standards above the asian mean. So not only are black cognitive elites less intelligent than e.g. Asian cognitive elites on average (because +3σ from a lower mean is a lower number than +3σ from a higher mean), they're also less common. The further away we move above the mean, the greater the intelligence discrepencies between races become.
>and given the social and cultural disparities across racial groups
Cultural explanations have consistently failed to account for IQ gaps, either in the form of test bias or genuine intellectual ability and its development. And understandably so: Are Ashkenazi jews much more intelligent than asian on average because asian culture isn't intellectual enough?
And again for social disparities. A full two thirds (!) of the black white IQ gap exists in children before they even begin school (with the rest increasing in line with the general increase in the heritabilty of traits with age), and adoption studies show a much greater correlation between biological parental IQ than adopted parental IQ, largely eliminating the possibility that home environment explains a majority of IQ gaps.
>and the social plasticity of IQ the share of biological causality
No such plasticity exists.
Twin studies conclusively show that IQ is highly heritable. When test reliability is accounted for, this heritability may be as high in countries like the US as 80%.
Maybe it's different between races, but the burden of proof is on you to show that this general rule for heritability doesn't apply in this specific case. And all main environmental explanations have studied and convincingly argued against.
>is a difficult and controversial research problem.
It's controversial for ideological, not scientific reasons. As for difficulty, perhaps. But breakthroughs in molecular genetic studies of intelligence have not only identified many of the genes associated with cognitive ability, but it turns out the alleles are unequally distributed between races (and in a way that is congruent with observed IQ differences): https://www.mdpi.com/2624-8611/1/1/34/htm
>Bostrom has not scientific legitimacy to give an opinion in such a technical and disputed issue.
This is a textbook, slamdunk case of 'isolated demands for rigor'. Such objections are never, and I truly mean never, raised by people like yourself against people denying the existence of heritable differences in intelligence between races. Most people who object to the existence of such differences are not only not scientists, but don't even have a rudimentary understanding of the literature on the topic. And sorry to say, but this includes yourself. Cultural explanations for IQ differences lack any support in the literature and this has been the case for decades now.
If virtually anyone is allowed to deny these differences exist, I see no reason why Bostrom cannot similarly express (an almost certainly more well informed than average) opinion on the topic too.
>Regarding feelings and human sympathy, the big question is: whose feelings? How can a modest difference in the mean of the IQ distribution offend a black man with an IQ of 130? And if you are a white person with an IQ of 85, how useful is for you the white black IQ gap for which you make a negative contribution? IQ is a personal characteristic, and means are completely irrelevant for every single person.
Groups differences in intelligence explain group differences in social outcomes - that's why this matters.
Racially conscious black people aren't angry just because they personally have less e.g. money than white people. They think black people generally having less money than white people generally is a grave social injustice. And considering these differences are explained by things like group intelligence differences, that's why said differences are relevant and important. And it also explains why we should never expect African countries as a whole to achieve anywhere near the same levels of economic development pre-singularity as western or North-East Asian countries, even if plenty of high IQ Africans can be successful.
If everybody was as politically individualistic as you're implying, there largely wouldn't be any problem. But that's not the case.
>If we believe that IQ is important for some application, we can measure it directly at the individual level instead of looking at skin color. Anti-racist arguments do not depend on the distribution of psychometric variables.
Yes, they do!
All of anti-racist thought today as practiced by anyone with any kind of institutional legitimacy and/or power is dependant on biological racial egalitarianism. They believe that without "racism", whatever they happen to mean by that, the races would have equal outcomes because they're biologically equivalent in intellectual ability.
>They depend on treating each person as an individual, ignoring skin color, and exclusively considering the abilities and limitations relevant to the case considered of the person under scrutiny. Any “anti-racism” in excess of this, is, in fact, racism
This is a wonderfully admirable belief to have, but unfortunately it is an extreme minority position institutionally. Hence why affirmative action exists.
>If Black people are not heavily under-represented in a 'cognitively demanding' organization that is very strong evidence the organization is racist against White and Asian individuals!
Any way of measuring cognitive ability shows that blacks have lower cognitive ability on average than whites and asians. Not only that, they have a lower variance in IQ, meaning fewer people 3 standard deviations above the black mean than whites/asians 3 standard eviations above the white/asian means respectively.
Any sufficiently large and non-niche organization (e.g. not some black focused charity or anything else that may legitimately have some metirocratic basis for favoring black applicants, or is located somewhere where there's mostly black people in the local labor market) that hires people at least significantly on the basis of cognitive ability and does not consider race should hire far fewer black people per capita than whites or asians. This has to be the case.
If this isn't the case, then this organization simply then factors other than cognitive ability must be dominating the hiring criteria, especially if we average this out over a whole sector, type of organization or the whole economy to smooth out any idiosyncrasies.
And we know for an indisputable fact that A) Universities engage in affirmative action on the basis of race which results in black applicants being admitted to selective colleges with academic scores below that of other races who get admitted and B) Black college graduates at all levels of qualification have significantly lower IQ, literacy and numeracy than graduates of other races at the same qualification level. So yes, its trivially true that institutions discriminate against whites and asians on the basis of race. You may think this is a good thing, but if you deny it exists then you are objectively incorrect.
>Obviously this point of view is completely at odds with any sort of fair and inclusive community or organization.
Hiring people on the basis of their cognitive ability is necessary NOT inclusive. Hiring the person who is most intellectually capable is necessarily discriminatory, even if we're talking about picking between people of different races.
As for fair, what is "fair"? Who decides what that is? Is a high IQ asian applicant missing out on a position at harvard to a lower IQ black applicant on the basis of their race "fair"?
I think this is highly unfair. But perhaps even more importantly, this will lead to worse social outcomes because there is a strong correlation between job performance and IQ.
Anything other than picking people at random is going to be not "inclusive", but doing so would make the maintenance of a modern economy impossible.
>This has profound implications for the ability of the movement to work in the interests of all humanity. For example, EAs who believe it is important that the human population expand but who think that African people are likely to be stupider than the human average may divert resources away from them. EAs making hiring decisions may prioritise applications from people whom they have not subconsciously assigned to the schema "stupider than me."
Let's say that it is true that not only are africans today lower in cognitive ability than whites on average, but that this gap is also mostly heritable (the latter is debatable but there is strong evidence to support it - the former is indisputable unless you think literally every single measure of cognitive ability that has strong predictive validity is radically and fatally flawed) .
Should this be acknoledged? Should this affect how we view the world?
Assuming it's true, do you believe, as a rationalist, that ignoring fundmental characteristics of human ity that explain more about society than almost any other factors will lead you towards making better decisions?
Speaking only for myself here, I think the answer is an unqualified hell no.
If you assume African Americans are congitive equivlanet to Asian Americans on average, then you will assume that sending more of them to college (in excess of those who can get in on the basis of academic merit alone currently) will result in improved socioeconomic outcomes (even if no affirmative action in hiring existed). But the real result is that many black people will have wasted 2-4 years of their life accumulating debt and forgoing earnings to gain an "education" that did nothing to close the cognitive gaps between themselves and asian americans. The debt, loss of income and wasting of time that could have been used developing valuable skills will hurt these graduates.
If instead, given that we know that intelligence varies between people, and we're interested in maximising the earning potential of black kids from low income families, we had identified their IQ early on and put our effort into getting them onto the optimal career track for their specific level of cognitive ability.
(The same is true for kids of any race - the difference is people are generally much more accepting of the fact that some white kids are just plain less intelligent than other white kids and can't have the same careers, and much more of acepting income inequality within than between races. And so there is a much bigger push to get black people specifically into college rather than low income/IQ kids generally, which is why it makes sense to think along racial lines here).
The same logic applies to entire countries. If you assume that Nigerians are no less intelligent than Chinese people, then you're likely to expect that Nigeria is going to follow the same development path as China once some threshold is passed for e.g. education health and so on. This is almost certainly not going to happen, and expecting it to will mean you will make suboptimal decisions when thinking about how to help countries in Africa.
More controversially, it also would mean that people in Africa are likely going to be of much lower instrumental value to humanity than people in Japan. But choosing to ignore this is actually the decision that is being made on the basis of race. Choosing to factor this into your worldview and decisions means you're acting on the basis of utility alone and not making decisions aimed at helping people because of their race more than you would of.
Liberal americans obviously aren't categorically opposed to helping low income black americans more than low income white americans in rural areas/the south, and it's likely this is because they think that black americans are better people and have more to offer the country, whereas white americans are just dumb and ignorant and cannot be reformed (not everyone believes this entirely, but I think some implicit form of this belief colors much of liberal thought in this area). All I'm saying is this type of thinking also applies more generally, except in this case the difference is based on data rather than intution and in-group favoratism.
If you, as you say, care about "working in the interests of humanity", you need to think about the interests of humanity collectively. Maximixing the interests of humanity will necessarily involve not helping all groups equally.
>Across large racial groups, however, the observed differences are statistically significant but modest (as they are for other anthropometric characters)
The balck white IQ gap is approximately one standard deviation, which is by no means modest. If we could raise a population's mean IQ by one standard deviation, this would be utterly transformative in a way that almost no other intervention could be.
I'm not saying it's not modest because its not a small number - it's not modest because such a gap has profound social implications in a way that black people being slightly taller than asian people doesn't. Controlling for IQ literally eliminates most of the inequality between races in things like income.
And it's not just a difference in the mean. Black people have a lower variance in IQ, meaning there are less black people 3 standard deviations above the black mean than there are asian people 3 standards above the asian mean. So not only are black cognitive elites less intelligent than e.g. Asian cognitive elites on average (because +3σ from a lower mean is a lower number than +3σ from a higher mean), they're also less common. The further away we move above the mean, the greater the intelligence discrepencies between races become.
>and given the social and cultural disparities across racial groups
Cultural explanations have consistently failed to account for IQ gaps, either in the form of test bias or genuine intellectual ability and its development. And understandably so: Are Ashkenazi jews much more intelligent than asian on average because asian culture isn't intellectual enough?
And again for social disparities. A full two thirds (!) of the black white IQ gap exists in children before they even begin school (with the rest increasing in line with the general increase in the heritabilty of traits with age), and adoption studies show a much greater correlation between biological parental IQ than adopted parental IQ, largely eliminating the possibility that home environment explains a majority of IQ gaps.
>and the social plasticity of IQ the share of biological causality
No such plasticity exists.
Twin studies conclusively show that IQ is highly heritable. When test reliability is accounted for, this heritability may be as high in countries like the US as 80%.
Maybe it's different between races, but the burden of proof is on you to show that this general rule for heritability doesn't apply in this specific case. And all main environmental explanations have studied and convincingly argued against.
>is a difficult and controversial research problem.
It's controversial for ideological, not scientific reasons. As for difficulty, perhaps. But breakthroughs in molecular genetic studies of intelligence have not only identified many of the genes associated with cognitive ability, but it turns out the alleles are unequally distributed between races (and in a way that is congruent with observed IQ differences): https://www.mdpi.com/2624-8611/1/1/34/htm
>Bostrom has not scientific legitimacy to give an opinion in such a technical and disputed issue.
This is a textbook, slamdunk case of 'isolated demands for rigor'. Such objections are never, and I truly mean never, raised by people like yourself against people denying the existence of heritable differences in intelligence between races. Most people who object to the existence of such differences are not only not scientists, but don't even have a rudimentary understanding of the literature on the topic. And sorry to say, but this includes yourself. Cultural explanations for IQ differences lack any support in the literature and this has been the case for decades now.
If virtually anyone is allowed to deny these differences exist, I see no reason why Bostrom cannot similarly express (an almost certainly more well informed than average) opinion on the topic too.
>Regarding feelings and human sympathy, the big question is: whose feelings? How can a modest difference in the mean of the IQ distribution offend a black man with an IQ of 130? And if you are a white person with an IQ of 85, how useful is for you the white black IQ gap for which you make a negative contribution? IQ is a personal characteristic, and means are completely irrelevant for every single person.
Groups differences in intelligence explain group differences in social outcomes - that's why this matters.
Racially conscious black people aren't angry just because they personally have less e.g. money than white people. They think black people generally having less money than white people generally is a grave social injustice. And considering these differences are explained by things like group intelligence differences, that's why said differences are relevant and important. And it also explains why we should never expect African countries as a whole to achieve anywhere near the same levels of economic development pre-singularity as western or North-East Asian countries, even if plenty of high IQ Africans can be successful.
If everybody was as politically individualistic as you're implying, there largely wouldn't be any problem. But that's not the case.
>If we believe that IQ is important for some application, we can measure it directly at the individual level instead of looking at skin color. Anti-racist arguments do not depend on the distribution of psychometric variables.
Yes, they do!
All of anti-racist thought today as practiced by anyone with any kind of institutional legitimacy and/or power is dependant on biological racial egalitarianism. They believe that without "racism", whatever they happen to mean by that, the races would have equal outcomes because they're biologically equivalent in intellectual ability.
>They depend on treating each person as an individual, ignoring skin color, and exclusively considering the abilities and limitations relevant to the case considered of the person under scrutiny. Any “anti-racism” in excess of this, is, in fact, racism
This is a wonderfully admirable belief to have, but unfortunately it is an extreme minority position institutionally. Hence why affirmative action exists.
>If Black people are not heavily under-represented in a 'cognitively demanding' organization that is very strong evidence the organization is racist against White and Asian individuals!
Any way of measuring cognitive ability shows that blacks have lower cognitive ability on average than whites and asians. Not only that, they have a lower variance in IQ, meaning fewer people 3 standard deviations above the black mean than whites/asians 3 standard eviations above the white/asian means respectively.
Any sufficiently large and non-niche organization (e.g. not some black focused charity or anything else that may legitimately have some metirocratic basis for favoring black applicants, or is located somewhere where there's mostly black people in the local labor market) that hires people at least significantly on the basis of cognitive ability and does not consider race should hire far fewer black people per capita than whites or asians. This has to be the case.
If this isn't the case, then this organization simply then factors other than cognitive ability must be dominating the hiring criteria, especially if we average this out over a whole sector, type of organization or the whole economy to smooth out any idiosyncrasies.
And we know for an indisputable fact that A) Universities engage in affirmative action on the basis of race which results in black applicants being admitted to selective colleges with academic scores below that of other races who get admitted and B) Black college graduates at all levels of qualification have significantly lower IQ, literacy and numeracy than graduates of other races at the same qualification level. So yes, its trivially true that institutions discriminate against whites and asians on the basis of race. You may think this is a good thing, but if you deny it exists then you are objectively incorrect.
>Obviously this point of view is completely at odds with any sort of fair and inclusive community or organization.
Hiring people on the basis of their cognitive ability is necessary NOT inclusive. Hiring the person who is most intellectually capable is necessarily discriminatory, even if we're talking about picking between people of different races.
As for fair, what is "fair"? Who decides what that is? Is a high IQ asian applicant missing out on a position at harvard to a lower IQ black applicant on the basis of their race "fair"?
I think this is highly unfair. But perhaps even more importantly, this will lead to worse social outcomes because there is a strong correlation between job performance and IQ.
Anything other than picking people at random is going to be not "inclusive", but doing so would make the maintenance of a modern economy impossible.