News for hackers. Stuff that matters.
Re:MSU Vice President of research & innovation Steve Hsu has... by ciarannolan on Friday June 19, 2020 @10:59PM (#23581177)
From the campus group calling for his firing [1]:

>However, these categories overlap & are all rooted in Hsu’s beliefs in innate biological differences between human populations, especially regarding intelligence --an unscientific position which makes him unsuited to direct a research institution’s funding or its graduate studies

Humans are, biologically, extremely diverse. Visible attributes, disease risk, geometry of the physical body, etc. Some groups are more prone to certain genetic disease risk factors, for example.

Doesn't it follow that the brains of different groups of humans might group into certain traits, including intelligence?

I'm asking in all sincerity. I'd love to hear if there's some evidence that my thinking is wrong.

[1] https://twitter.com/GradEmpUnion/status/1270829006439485441

Re:MSU Vice President of research & innovation Steve Hsu has... by esyir on Friday June 19, 2020 @11:37PM (#23581407)
There really isn't.

The current argument is weasly though, approaching it as follows:

Race is a social construct and not a genetic one. Humans exist on a genetic spectrum. Therefore you can't box humans into distinct groups there. Therefore racial differences are moot.

This is based off the grad student unions linked sources. It's also weasly as all fuck. Race may be a social construct, but it does correlate strongly with a number of genetic factors. As such, race serves as an aggregate variable for a large quantity of genetic variation. And it can be shown that between populations of different race, there's an observable difference at a genetic level.

Re:MSU Vice President of research & innovation Steve Hsu has... by Tarq0n on Friday June 19, 2020 @11:20PM (#23581291)
It's complicated. One thing we do know is that the variance in intelligence is larger than any group differences. People therefore often assume that anyone that insists on emphasizing the average difference is using it as a dog-whistle for some kind of bias. Whether the principle of charity should be applied or not is up to you.
Re:MSU Vice President of research & innovation Steve Hsu has... by throwaway286 on Saturday June 20, 2020 @12:47AM (#23581750)
It doesn't matter if your thinking is right or wrong. Look what happened to Galileo. If you go against the current on a touchy topic you're done for.
Re:MSU Vice President of research & innovation Steve Hsu has... by staplers on Friday June 19, 2020 @11:34PM (#23581386)
There is no singular "intelligence" and measuring it is certainly not linear. Splitting hairs between a population of people (specifically blacks) who only 200 years ago were killed for being able to read certainly doesn't make for a compelling or worthy investment in research.
Re:MSU Vice President of research & innovation Steve Hsu has... by DuskStar on Saturday June 20, 2020 @12:04AM (#23581550)
> There is no singular "intelligence" and measuring it is certainly not linear.

Wanna bet? I think most intelligence researchers would disagree with you, and IQ is predictive of life outcomes.

There are areas where you can find fault with Hsu's work, but I don't think "IQ approximates g, and g is real" is one of them.

For that matter... If a population was killed according to a trait, wouldn't that make it more likely for there to be genetic causes of group differences in that trait?

Re:MSU Vice President of research & innovation Steve Hsu has... by throwaway4666 on Sunday June 21, 2020 @06:51AM (#23591278)
"Intelligence research" is, like most of psychology, an elaborate grift designed to give a handful of people cushy tenured jobs. Actual geneticists, that is, people who study how genes actually work, wouldn't touch the standards of rigor of "intelligence research" with a ten foot pole. Note how people supporting Hsu tend to be psychologists while people dismissing his research tend to be geneticists. It's also worth nothing that Hsu is also out-of-field himself (he used to be a physicist) and only got recently into genomics. That he got stuff wrong that's basically understoood by the whole genetics community isn't that surprising.
Re:MSU Vice President of research & innovation Steve Hsu has... by DuskStar on Sunday June 21, 2020 @10:42AM (#23592259)
> "Intelligence research" is, like most of psychology, an elaborate grift designed to give a handful of people cushy tenured jobs. Actual geneticists, that is, people who study how genes actually work, wouldn't touch the standards of rigor of "intelligence research" with a ten foot pole.

Wow, that's... Misguided. Most intelligence research doesn't have to do with genetics at all, to start with. Second, twin studies have shown that intelligence is highly heritable, with about half of variance explained by genetics (and half "non-shared environment", also known as random chance) - we might not know what genes are responsible, but you don't need to know the genes to compare identical and fraternal twins. Finally... IQ is real is one of a very small number of psychology studies that reproduces consistently, so the normal "psychology is bunk" objections don't apply.

Re:MSU Vice President of research & innovation Steve Hsu has... by throwaway4666 on Monday June 22, 2020 @03:05AM (#23597912)
>Second, twin studies have shown that intelligence is highly heritable, with about half of variance explained by genetics (and half "non-shared environment", also known as random chance) - we might not know what genes are responsible, but you don't need to know the genes to compare identical and fraternal twins.

Heritability does not imply the existence of a genetic mechanism. Twin studies have a number of flaws that further demonstrate my points about standards of methodology that geneticists wouldn't touch with a ten foot pole. Even not taking into account stuff like equal environment assumption, identical twins simply do not have the same DNA. Yes, even identical twins, look it up. There are also a number of things that are heritable without involving a genetic mechanism at all, such as the various microbiomes or epigenetics. And of course there are things like transmission of de novo mutations over up to two generations (e.g. affecting germinal cells of one's fetus).

Regarding the 'IQ is the best finding we have on psychologist' point. Being the best in a field that's a largely grift doesn't make one any less of a grift. IQ (and g) still has flaws that no one in the hard science would consider it. For instance, it is gaussian for no reason and despite most social outcomes performances being fat-tailed, meaning the theoretical covariance between IQ and any kind of relevant task is undefined (and its in-sample variance finite but meaningless). There's also zero neurological or genetic data confirming the existence of g, making it a purely imaginary construct that only exists in the minds of psychologists. That this is the best they can come up with just speaks even worse of the field as a whole.

Re:MSU Vice President of research & innovation Steve Hsu has... by DuskStar on Monday June 22, 2020 @01:25PM (#23602955)
> Heritability does not imply the existence of a genetic mechanism. Twin studies have a number of flaws that further demonstrate my points about standards of methodology that geneticists wouldn't touch with a ten foot pole. Even not taking into account stuff like equal environment assumption, identical twins simply do not have the same DNA. Yes, even identical twins, look it up. There are also a number of things that are heritable without involving a genetic mechanism at all, such as the various microbiomes or epigenetics. And of course there are things like transmission of de novo mutations over up to two generations (e.g. affecting germinal cells of one's fetus).

...I think you're missing how twin studies work, here. Yes, they can't distinguish between epigenetic and genetic traits, but they DON'T require twins to have identical genetics, and DON'T require equal environments. Specifically, they work by

1. Measuring the correlation in identical twin IQs (or other traits)

2. Measuring the correlation in fraternal twin IQs (or other traits)

3. Comparing the correlations. If identical twins have a 50% correlation and fraternal twins 25%, well...

Identical twins not having identical genetics makes it harder to find genetic influence, but doesn't make it impossible. (it just means that the twins share 99.9...% of varying DNA instead of 100%, which is still far above fraternal twins and enough to establish a difference) And the reason you use twins is to control for varying environments. (And for that matter both identical and fraternal twins will share germinal cells/microbiomes/etc, too)

Yes, it's fuzzier than a physics experiment. But it's certainly not worse than "this gene causes a 0.1% increase in the chance of breast cancer".

Re:MSU Vice President of research & innovation Steve Hsu has... by skosuri on Saturday June 20, 2020 @02:23AM (#23582168)
If you are actually sincere, take the time to read this: http://ewanbirney.com/2019/10/race-genetics-and-pseudoscienc...
Re:MSU Vice President of research & innovation Steve Hsu has... by ciarannolan on Saturday June 20, 2020 @01:33PM (#23585185)
I am "actually sincere" and I did take the time to read that.

TLDR:

A very long, rambling blog post that essentially says "yeah there's probably genetic differences related to IQ among groups of humans but...", then proceeds to dismiss anyone who disagrees with them as part of "a vocal fringe of race pseudoscience".

Re:MSU Vice President of research & innovation Steve Hsu has... by skosuri on Saturday June 20, 2020 @02:38PM (#23585765)
Lol; what I thought. Do you just make up things you put in quotes? That’s not what those characters are meant for.

The “long, rambling blog post” outlines a number of pieces of evidence that say it’s highly unlikely there are meaningful differences, and because of population stratification, stuff like GWAS (that Hsu uses) can’t give us a lot of info. Look, if you have questions or doubts about these criticisms, you can ask and people will take those questions seriously. Ewan Birney is a world class geneticist that leads Europe’s leading institute on the subject. He’s writing a blog post for lay people to help them understand the complexities. What you wrote in response just shows how blinded you are in just discounting other arguments.

Look, I don’t know who you are; you can look me up (https://www.octant.bio/biosri).. but maybe when most experts in a field are saying one thing, and one person on the fringe is saying another... you at least owe it to yourself to understand the dominant viewpoint. It’s easy to find my email, or Ewan’s or many other researchers that take a lot of time answering questions. Quantitative genetics is complicated and in that blog post he’s trying to break down the major ways people get this stuff wrong; including Hsu. But in general, most on this site thus far have a viewpoint that is set in stone, hold their nose, and think an entire community is really stupid and politically correct. Anyways, if you are not a layperson I apologize and as a researcher, you can always just publish your findings, and the community and I can learn from them.

Re:MSU Vice President of research & innovation Steve Hsu has... by tomp on Saturday June 20, 2020 @08:27PM (#23588633)
Hey, I saw you posted a few comments saying that it's highly unlikely that there are "meaningful" differences. So even if we leave the "obvious" differences between populations like skin color aside as not "meaningful", do you have any sources that explain (or if you want to yourself) the "obvious" or "well-known" differences between groups, such as Dutch height, Ashkenazi intelligence or Kalenjin running endurance? Sincere question, I'm genuinely curious as I can't think of any other obvious explanation than "genetic group differences" (and it's kind-of hard to Google for).
Re:MSU Vice President of research & innovation Steve Hsu has... by skosuri on Saturday June 20, 2020 @09:56PM (#23589081)
First. I’d say read the blog post I linked to above. To answer your question on the “well-known” differences; they definitely are real and genetic. It’s just that IQ and most measures of intelligence look nothing like those traits. The common variants that are tagged in GWAS for IQ are also not different from a variant distribution standpoint. Again, the blog lays this out. There are thousands of papers on this topic. Good books to read are David Reich’s book (which does enter this more touchy territory) and ‘She has her mothers laugh’ for a more general reading. Ewan’s blog in general is also really good as well. In terms of papers, there are a bunch of recent ones that are quite good. Will need more time to post but they tend to be covered by Ewan.
Re:MSU Vice President of research & innovation Steve Hsu has... by tomp on Tuesday June 23, 2020 @06:36AM (#23612117)
I read the article: http://ewanbirney.com/2019/10/race-genetics-and-pseudoscienc...

I don't quite understand what the point is. The main crux of it seems to be "argument by obfuscation", a lot of deferral to (scientific) authority and a lot of "sophiscated" words (hoping that the reader will also defer to authority).

The author seems to argue something like: we can't know this because (1) genetics are complicated, (2) we don't have enough data, (3) the data that we do have is hard to understand, and (4) even if we could draw conclusions from the data, we couldn't generalize them on other populations. This seems like a coop-out - I'm 100% that in the future, we'll have more data, and we'll have better tools to understand the data, so whatever conclusions we make, will be that much more certain.

Also, part of the author's argument seems to be "there wasn't significant evolutionary pressure on these traits". So what? Obviously there are differences in traits that have no evolutionary pressure (e.g. why do SE Asians have slanted eyes? was there some evolutionary pressure for that?), there's lack of differences even when there is evolutionary pressure (e.g. female sex selection seems to prefer height, yet non-Dutch Europeans and SE Asians are smaller; in addition, the latter haven't developed ability to digest milk after infancy, despite presumably the same evolutionary pressures); finally, there's traits that do conform to evolutionary pressure, including all of the ones I highlight above (tall Dutch, smart Jews, fast Kenyans). Either different populations face different evolutionary pressures so have developed different traits, or all populations face the same evolutionary pressures and some have failed to develop some traits. In any case, it's more than possible that there are population-wide differences even in heritable IQ, and we just can't confirm or deny that yet (we already know there are "sub-population-wide" differences, e.g. Ashkenazi Jews).

Part of the authors argument is also that because most studies are done on Europeans, they can't be generalized. Again, irrelevant. The examples I pointed out are "within population" - Kalenjin run faster than other Kenyans, Dutch are taller than other Europeans, Ashkenazi Jews are smarter than other Europeans and other Jews.

The author also argues that it's somehow invalid to exclude recent "immigrants" when studying a population. Why? Obviously if I study Ashkenazi Jew IQ, I'd exclude other (non-Jewish) Europeans, as well as non-Ashkenazi Jews (e.g. Sephardi) and those who are not genetically Ashkenazi Jews (e.g. people who recently converted to the religion). I can't see how that would be "unscientific".

> it is virtually inconceivable that the primary determinant of racial categories – that is skin colour – is strongly associated with the genetic architecture that relates to intelligence.

The author either lacks imagination, or his/her imagination is stifled by political correctness. We know this claim is false, as "race" is associated with height (Europeans are on average taller than SE Asians - and I looked up Japan specifically, a first-world country, so don't "nutrition" me), which is also a complex trait.

To conclude, there's only 3 possible answers to the question of "Why are Ashkenazi Jews smarter?": (1) They aren't. (2) It's (mostly) genetic. and (3) It's mostly not genetic (nutrition, culture, etc.). I don't see the author's incoherent rambling (you can see that I've become more and more disappointed by the article as I read it) arguing in either of these directions, in particular I don't see any argument to dismiss (2), except a few weak sauce non-arguments that I criticize above. I also don't see anything against Steve Hsu's argument, which is basically that IQ & height are at least partially genetic, we'll soon be able to predict them from genes, then we'll be able to use it in embryo selection.

Therefore, I'm going to summarily dismiss your argument and disregard other sources you suggest, unless you can come up with a short counterargument to (2) above that doesn't fail in the silly ways as this article does.

Edit: This isn't to say that I'm not open to other explanations... Maybe Kalenjin just are training better. Maybe Ashkenazi just are rearing their children better. Maybe the Dutch are eating the right food that makes them grow taller. But I find that unlikely (their secrets would have gotten out already) and in any case, I'm mainly just saying that we can't dismiss either explanation.

Re:MSU Vice President of research & innovation Steve Hsu has... by skosuri on Wednesday June 24, 2020 @12:52AM (#23624239)
Besides the insults, there are a couple of real scientific questions. Despite my better judgement... :

1. 'evolutionary pressure on traits' -> so what? It's important, because we are usually much better powered at detecting purifying selection in traits.

2. As for the slanted eyes: why do you think there was no pressure? Assortative mating is a thing.. it's very prevalent in visible traits and really makes this hard.

3. Preference for height... umm ok - this is weird. There are plenty of known examples for purifying height selection in either direction. However, these are rare variants and not the major driver of height variation within bigger populations. Height is actually a good example for looking at variation within populations: https://elifesciences.org/articles/39702 Basically it says that population stratification is a major problem for things that are easily measurable like height (unlike intelligence, which is much harder to measure and way more likely to have known social covariates).

4. In general, you aren't even understanding the simplified responses that the authors are trying to explain, but also can't seem to grasp there might be more that you are not understanding – and just ascribe things to lack of imagination or political correctness of the authors.. who are world class geneticists. You can lead a camel to water...

Re:MSU Vice President of research & innovation Steve Hsu has... by tomp on Wednesday July 01, 2020 @03:37AM (#23698857)
Sorry I haven’t replied, I didn’t have the time to read the article yet, and your reply isn’t very substantial...

If I understa correctly, your point is that there are things that are easy to see (matters for sexual selection) and/or measure (matters for science) and IQ is neither... but those aren’t the only causes of genetic differences? You also have survival pressure (presumably running faster would be here) and just random drift as well (presumably that’s how slanted eyes and blond hair first developed). You didn’t answer which of the options I listed regarding Jewish IQ is correct. Alternatively, why are humans smarter than chimps?

Also, regarding your (4), scientists are notorious for submitting to pressure from the wokeness mob, and in touchy subjects like genetics in particular, so it’s not an unreasonable assumption on my part.

Re:MSU Vice President of research & innovation Steve Hsu has... by ciarannolan on Saturday June 20, 2020 @02:58PM (#23585969)
>Do you just make up things you put in quotes? That’s not what those characters are meant for.

The two "quotes" in my comment:

>>[...] that essentially says "yeah there's probably genetic differences related to IQ among groups of humans but..." [...]

English isn't my first language, but it's crystal clear that this isn't a direct quote but is a summary of my thoughts on their argument.

>>[...] "a vocal fringe of race pseudoscience".

If you read your own link you would know that this is a direct quote.

But please, continue to educate me on punctuation.

>What you wrote in response just shows how blinded you are in just discounting other arguments.

That's rich coming from someone who started their comment with "Lol; what I thought." And someone who started their first reply to my original comment by questioning my sincerity.

>But in general, most on this site thus far have a viewpoint that is set in stone, hold their nose, and think an entire community is really stupid and politically correct.

Where could they possibly get that idea? (https://news.ycombinator.com/item?id=23585765)

Finally, what was the point of trying to flex your credentials? You spent an entire comment reinforcing the idea that academics like you "have a viewpoint that is set in stone, hold their nose, and think an entire community is really stupid and politically correct."

Re:MSU Vice President of research & innovation Steve Hsu has... by skosuri on Saturday June 20, 2020 @03:42PM (#23586541)
So again, you have no substantive questions/comments on the criticisms brought up in the blog post other than to call it long and rambling. Nothing on how common variants associated with intelligence don't segregate substantially across populations. Nothing on the poor accounting for population stratification; nothing on lack of replication on populations outside the regressions; nothing on effect sizes; nothing on IQ test score changes over time that are too short for genetic changes...

anyways, the flex is because you now know who I am and what I research. If you have serious questions about the above topics... the same topics for why most people don't take Hsu seriously, feel free to email them to me. In the meantime, I am laughing because you were anything but sincere.

Re:MSU Vice President of research & innovation Steve Hsu has... by ciarannolan on Saturday June 20, 2020 @03:51PM (#23586640)
>So again, you have no substantive questions/comments on the criticisms brought up in the blog post other than to call it long and rambling.

And you have no reply to my criticisms of your sarcastic, ad hominem filled reply.

>>You spent an entire comment reinforcing the idea that academics like you "have a viewpoint that is set in stone, hold their nose, and think an entire community is really stupid and politically correct."

Look in the mirror once in a while.

Re:MSU Vice President of research & innovation Steve Hsu has... by skosuri on Saturday June 20, 2020 @04:17PM (#23586904)
Sorry, didn't actually see any scientific criticisms or questions in any of your replies; just your "sincere" request for more information, which I gave you. You just don't seem interested in learning more.
Re:MSU Vice President of research & innovation Steve Hsu has... by ciarannolan on Saturday June 20, 2020 @04:29PM (#23587006)
There weren't any. I am criticizing the nasty tone you took from the outset in assuming I am not sincere, which is wrong. I'm not a biologist, so I obviously could have missed something in that blog post. But, I shared my impression of it nonetheless, since this is a board for discussion.

I know what it's like to achieve an extremely high level of expertise in a narrow field; I've done it myself. It forever blinds you to the way a novice sees the same field.

If you've ever wondered why there's increasing skepticism of academia, you should read your comments again. They are nasty, sarcastic, and dismissive.

I'll paste this one more time because I think you need to seriously consider that you are exhibiting exactly the behavior that you're accusing others of:

>[...] have a viewpoint that is set in stone, hold their nose, and think an entire community is really stupid and politically correct.

Re:MSU Vice President of research & innovation Steve Hsu has... by skosuri on Saturday June 20, 2020 @05:07PM (#23587311)
I wasn't being sarcastic. I was saying that if you are actually sincere, and I was hopeful you were, I said to to take the time to read the blog post (knowing it was long); they go through great lengths to explain to people not in the field what the issues with work like Hsu's. It wasn't meant to be mean. You dismissed it, failed to try to understand it, and were incredibly inconsiderate calling it long and rambling. This is when my tone changed, and in my opinion deservedly so. It indicated to me you were being insincere about spending the time to learn more. Considering how many racist trolls I deal with that do this kind of thing often, your dismissiveness made me think you weren't actually interested in learning more. I gave my information, pointed to the things that you were misunderstanding in that blog, and offered if you have scientific questions, I am happy to take more time to try to answer them. That offer still stands.
Re:MSU Vice President of research & innovation Steve Hsu has... by ramraj07 on Saturday June 20, 2020 @12:09AM (#23581570)
There is a simple solution to this: no one other than card-carrying scientists working exclusively on the evolution of intelligence should blurt out their opinions on this topic. This question should be clubbed with other questions which are always easy traps for any scientist into believing their opinion is valuable just because they "know how to do science". Better just stick to what field of science you are good and shut up about socially charged questions of no relevance to you.

In this guy's, case, it is arguable if he's in the field. He seems to have already made up his mind on it, he's just out to find proof of his hypothesis. This is in principle fine, but he should acknowledge that by choosing to study this field, he excludes himself from any position where he's in charge of teaching or training people - he seems to sincerely believe that some of them are dumber by their nature, that seeems like a massive conflict of interest to me.