Information Processing

It's only bold because there is a large number of people (often state founded, often controlling at least the lower levels of education, often with major influence on media) who strongly feel otherwise -- with no numbers to back them up, of course. --- If you actually do want to get to the bottom of it, I can heartily recommend Ian Deary's "Intelligence: A Short Introduction" as a place to start. http://www.amazon.com/Intelligence-A-Very-Short-Introduction/dp/019289… (Who knows, if you dig around a bit you might even be able to find a PDF, perhaps at Emil Kirkegaard's website.) You can also choose to spend an hour on this talk by Stephen Hsu who does research on IQ and genetics: https://www.youtube.com/watch?v=62jZENi1ed8 You can read some of Plomin's stuff -- he is very much into genetics: http://www.nature.com/mp/journal/v20/n1/full/mp2014105a.html There's a paper I can't find at the moment that looks at almost 4000 non-related people, give them an IQ test each + scan their DNA for typical SNPs with a DNA chip. That gives us almost 8,000,000 pairs where we know how similar their IQ score and DNA was. That gives us almost 8,000,000 points we can plot into a 2D coordinate system (and we can of course also throw statistics at them). It turns out that there is a clear correlation between measured IQ and measured SNPs (i.e., the point cloud is denser along a line). In other words, there is a statistical relation between DNA and IQ. This is a result that is…

1. Do you really want a PhD? If you believe so, then read "Don't become a scientist! Philip Greenspun edition" at https://infoproc.blogspot.com/2008/05/dont-become-scientist-philip-gre… 2. Years ago Greenspun graphed level of education vs pay scale. Salaries maxed out at a masters but plunged to high-school levels beyond that. See the graph below "Not So Very Serious Stuff" at: https://philip.greenspun.com/careers/ In summary, best get a master's and pass on the Ph.D. Bonus: you'll gain an extra 3-7 years of your young life (doing what you want). Also look at "Grad school versus prison, quantified". Wish I'd known all this when I was younger. https://philip.greenspun.com/blog/2018/12/09/grad-school-versus-prison…

Relayed by Nick Metropolis: Fermi and von Neumann overlapped. They collaborated on problems of Taylor instabilities and they wrote a report. When Fermi went back to Chicago after that work he called in his very close collaborator, namely Herbert Anderson, a young Ph.D. student at Columbia, a collaboration that began from Fermi's very first days at Columbia and lasted up until the very last moment. Herb was an experimental physicist. (If you want to know about Fermi in great detail, you would do well to interview Herbert Anderson.) But, at any rate, when Fermi got back he called in Herb Anderson to his office and he said, "You know, Herb, how much faster I am in thinking than you are. That is how much faster von Neumann is compared to me." That and a few other bits in http://infoproc.blogspot.com/2012/03/differences-are-enormous.html

> or they have something they want to hide. Indeed. I recommend this interview with Jeffrey Sachs, who headed Lancet's commision to investigate Covid-19. https://infoproc.blogspot.com/2022/10/jeffrey-sachs-lessons-from-covid… You may find your eyebrows a couple inches higher after listening :-)

An illustrative link: https://infoproc.blogspot.com/2017/03/everything-is-heritable.html (A short post with a nice chart, contains a further link to a proper paper.) Regarding the pessimism of genetic pre-determination and this congenital disadvantage: I think the real progress of society is how it learns to make lives of different people productive and satisfying. E.g. nerdy types had it hard in, say, 19th century Europe, but now they have a great and important niche in engineering. Same happened to many more predispositions of people that now lead to something interesting instead of pure misery. I think this process is not going to stop.

I'd recommend these as an interesting starting point ([2] has videos): [1] http://infoproc.blogspot.com/2017/02/on-military-balance-of-power-in-w… [2] http://infoproc.blogspot.com/2016/09/balance-of-power-in-western-pacif…

BTW the author of this article, Stephen Hsu, has a blog which frequently has interesting comments on such things (and AI, and much else): http://infoproc.blogspot.com/

The OP makes an interesting point, but I think Steve Hsu's post below provide a more nuanced discussion of Oppenheimer and of the book mentioned in the linked article: http://infoproc.blogspot.com/2013/07/strange-gadget-robert-oppenheimer… Elsewhere, Hsu also links to this review by Freeman Dyson of the same book: http://www.nybooks.com/articles/archives/2013/aug/15/oppenheimer-shape… in which, rather than to lack of confidence, Dyson attributes Oppenheimer's "failure to be a great scientist" to: a lack of Sitzfleisch. Sitzfleisch is a German word with no equivalent in English. The literal translation is “Sitflesh.” It means the ability to sit still and work quietly. He could never sit still long enough to do a difficult calculation.

The author, Steven Hsu, is a physics professor at the university of Oregon. He has a rather informative blog [1] about physics, and lately genetics. In particular he has recently been doing a lot of work with BGI, currently the largest human genome sequencing company in the world. [1] http://infoproc.blogspot.com

>1. In the top band there are many more males than females. Is 5% of boys scoring in the top band, compared to 3% of girls, really "many more males"? (Yes, I know that means that 60% more boys than girls scored in the top band. But the percentages are what matter here.) EDIT: From a comment: >Also, in the bottom band, there are many more males than females. >It seems to be yet another confirmation of the greater male variability hypothesis I feel foolish for not mentioning this. I think 18% of boys in the bottom band versus 14% of girls is "many more males" only in the sense that the 5% versus 3% figure above is, but taken together, yes, this is more evidence of male scores being the same on average but having a wider distribution. >or is it considered a fact by now? It sure wasn't in 2005! (For others' benefit, the commenter is alluding to what happened to Larry Summers as Harvard president. <https://en.wikipedia.org/wiki/Lawrence_Summers#Differences_between_the… EDIT 2: The blogger himself addressed it back in 2005! <https://infoproc.blogspot.com/2005/02/summers-lynching.html>

The paper says: > The question is whether the bimodality is symptomatic of a poverty trap, namely whether poor people do casual jobs and hold nearly no productive assets because they do not have the talent to do anything else or whether the fact that they are poor prevents them from acquiring the assets needed to climb the occupational ladder. The cause/effect inversion, of talent/ability => asset acquisition, is always worrying. We might benefit, as a society, if we confront the IQ problem honestly and head-on, in order to have a more complete/accurate policy picture especially because it seems reasonable that wealth transfers should be optimised to enable higher productivity. Consider: "The Bell Curve Review:IQ Best Indicates Poverty"(Benjamin, 2018), in which: > I find that IQ is more correlated with poverty than socio-economic-status in every regression, regardless of what is included in the regression, by a factor of 3. Also: Hsu(2015) https://infoproc.blogspot.com/2015/04/income-weath-and-iq.html > Income mobility is strongly affected by IQ. In fact, IQ is a much stronger predictor variable than race for escaping the bottom quintile of income.

> Doesn't this just mean that IQ Scores don't measure what they're supposed to? The most cautious defensible position, in my opinion, is that "we can crudely measure a quantity that is related to the folk notion of intelligence"[1]. IQ tests are good, not great, at measuring something which 1) correlates to performance on just about any task that we regard as "mental", and 2) which we are thus far unable to divide any further – it appears to be impossible to split this "something" into a bundle of uncorrelated disparate abilities[2]. This "something" is called general intelligence, or g. However, all IQ tests are trainable and have different emphases; there is no bulletproof g test. > My understanding is that IQ scores measure how smart you COULD be, all things considered. The g factor is not a potential. It is defined by statistically extracting a factor common to all (or all minus epsilon) measures of intelligence, which factor is then found – empirically – to predict future performance. It does appear to measure a kind of intellectual potential[3], which is one reason psychologists are interested in it. So, does this notion of general intelligence contradict the findings linked in this post? Logically, not at all; there is no reason a priori why the general factor should not be malleable. The question is if the general factor is in fact altered. Not an easy question to answer, due to the fact that g, being a statistical construct, cannot directly be measured, but must…

Steve Hsu: http://infoproc.blogspot.com Example post title: "Preimplantation genetic diagnosis and screening using next generation sequencing" Razib Khan: http://www.unz.com/gnxp/ Example post title: "Most Horse Lineages Are in Heaven" Gregory Cochran: http://westhunt.wordpress.com/ Example post: "It is easier to develop a phobia about snakes than electricity or carbon monoxide, probably because we have built in neurological mechanism that confer that propensity. Likely most animals have a similar propensity to develop a fear of fire: or it might come automatically. If there was such a fear-of-fire mechanism, we have lost it: and dogs have as well. If this is correct, one could learn about this hypothetical mechanism by comparing dogs and wolves." --- These are heavily biased towards population genetics and quantitative genetics. Would appreciate suggestions for blogs of smart people writing about molecular biology.

With respect, you sound like you're under the very common misconception that top universities are interested in educating the most academically impressive undergrads, and that it's a fight between parents to see who can trick the school into thinking their child is among the best students. In fact, universities are driven primarily to increase their influence on the world at large. When hiring professors, this can indeed mean academic acumen is valued, since impressive scholarship shapes the world and brings the university prestige. At the same time, grant-getting abilities are very important, since you can use money to buy influence in lots of ways. When selecting undergrads, academic ability has some instrumental value but very little intrinsic value to the university. This is because academic-superstar undergrads will leave the university before they exert influence with their scholarship. The universities does wants to mold the minds of future powerful people, and that usually means looking for students who are fairly (though not necessarily outrageously) smart. But to first order, you can think of the undergraduate program at top universities as a pure money-making venture. The first strategy used is near-perfect price discrimination to maximize immediate returns, i.e. "send us your parent's W2s and we'll tell you how much you'll pay". (Tuition is extremely high but most students get substantial financial aid.) The second strategy is to groom an ideal donor pool…

But does it really matter that the students are not "American-born"? If you're trying to talk about the impact of American universities on America's innovative capacity and America's economy, then yes, it matters how many students and graduates are American-born -- because only an extreme few of those immigrant students are getting Green Cards. It also matters to consider American-born students rather than all citizens and prospective citizens because we're trying to get an indication of what people think about the American labor market. The best way to measure that is to look at people who are confined to the American labor market. If we had a "points-based" immigration system like Canada, say, I would be asking how many graduates are American citizens. But we don't. Most international students I met were planning on either going "back home" or rising through further academic degrees (ie: grad-school) in order to continue their visas long enough to at least try for a Green Card and/or American citizenship. The economic impact of the latter is limited to academia.... which gets us back to the very problem I've been trying to articulate: that academia (in the sense of teaching and research, not just university incomes or spending) has been killed off as a major economic engine in America. http://infoproc.blogspot.co.il/2008/05/dont-become-scientist-philip-gr… -- here is our problem, described twice. American STEM innovation/research is a bad career path. …

In the blogpost "Roe devised her own high-end intelligence tests as follows: she obtained difficult problems in verbal, spatial and mathematical reasoning from the Educational Testing Service, which administers the SAT, but also performs bespoke testing research for, e.g., the US military. Using these problems, she created three tests (V, S and M), which were administered to the 64 scientists, and also to a cohort of PhD students at Columbia Teacher's College. The PhD students also took standard IQ tests and the results were used to norm the high-end VSM tests using an SD = 15. Most IQ tests are not good indicators of true high level ability (e.g., beyond +3 SD or so)." Seems a decent try at it to me. But a better indicator that IQ is predictive of success is http://infoproc.blogspot.com/2009/01/horsepower-matters-psychometrics-… Can psychometrics separate the top .1 percent from the top 1 percent in ability? Yes: SAT-M quartile within top 1 percent predicts future scientific success, even when the testing is done at age 13. The top quartile clearly outperforms the lower quartiles. These results strongly refute the "IQ above 120 doesn't matter" claim, at least in fields like science and engineering; everyone in this sample is above 120 and the top quartile are at the 1 in 10,000 level. The data comes from the Study of Mathematically Precocious Youth (SMPY), a planned 50-year longitudinal study of intellectual talent. Ability Differences Among People Who Have…

another one for the list http://infoproc.blogspot.com

This is no surprise to those who follow psychometric research. Also, what seems to be true (from preliminary studies) is that the gene variants that were under strong selection (reached fixation) over the last 10k years are different in different clusters. That is, the way that modern people in each cluster differ, due to natural selection, from their own ancestors 10k years ago is not the same in each cluster -- we have been, at least at the genetic level, experiencing divergent evolution. In fact, recent research suggests that 7% or more of all our genes are mutant versions that replaced earlier variants through natural selection over the last tens of thousands of years. There was little gene flow between continental clusters ("races") during that period, so there is circumstantial evidence for group differences beyond the already established ones (superficial appearance, disease resistance). http://infoproc.blogspot.com/2008/01/no-scientific-basis-for-race.html psychology.uwo.ca/faculty/rushtonpdfs/PPPL1.pdf http://www.opinionjournal.com/extra/?id=110007391

> They did to a degree Where? I’m fairly facile with causal inference; this is the crudest observational “study” without even qualitative heuristics to make the comparisons apples-to-apples. > and expand on why that particular quote offends you? Generic low tolerance for high V, low M academics: https://infoproc.blogspot.com/2011/06/high-v-low-m.html?m=1 This is the paper he links to, where you can find references to Derrida: https://unisapressjournals.co.za/index.php/Phronimon/article/view/1323… Anyway, if this is the kind of nonsense that convinces you, I don’t intent to continue this discussion. We have very different notions of what qualifies as evidence.

A rebuttal: https://infoproc.blogspot.com/2013/04/myths-sisyphus-and-g.html?m=1

> that brains are fundamentally survival machines, and agnostic to truth. It isn't true that brains are agnostic to the truth. Having an accurate model of the bits of reality you'll encounter is quite important for survival (a superstitious tribesperson won't believe that they can throw fireballs at incoming lions). > Knowing, or behaving-as-if-knowing a "truth" when that "truth" is being negatively selected for, leads to a disadvantage, leads to no one actively speaking up for it, leads to only ever hearing the "pro" side. Something we are sadly finding out for ourselves these days... with professors being removed from their positions for giving a platform to other academics with unpopular views: https://infoproc.blogspot.com/2020/06/resignation.html?m=1

This anecdote is really something https://infoproc.blogspot.com/2022/10/sat-score-distributions-in-michi…

See also this study of preferences in SMPY (high average IQ) cohort: https://infoproc.blogspot.com/2014/11/gender-differences-in-preference…

“ Not sure the blogpost supports Gould being innumerate or not understanding PCA.. also, an aside, I doubt the author Hsu would agree with your take here, considering his hero Feynman had an underwhelming IQ relative to his success.” Actually, here’s Hsu on Gould again: https://infoproc.blogspot.com/2011/06/high-v-low-m.html?m=1 The author does indeed tackle Feynman’s IQ here, and agrees with me (actually Hsu has informed a large part of my worldview): https://www.psychologytoday.com/us/blog/finding-the-next-einstein/2011… “ BTW what do you mean by IQ measurement is underpinned by PCA (principal component analysis? or the requisite math?)?” Principal component analysis. More here, again from Hsu: https://www.cog-genomics.org/static/pdf/ggoogle.pdf

> Stephen Jay Gould was actually fairly innumerate. It’s clear he didn’t understand even the rather simple math behind the PCA, which underpins IQ measurement. You might find this instructive: https://infoproc.blogspot.com/2011/06/high-v-low-m.html?m=1 Not sure the blogpost supports Gould being innumerate or not understanding PCA.. also, an aside, I doubt the author Hsu would agree with your take here, considering his hero Feynman had an underwhelming IQ relative to his success. BTW what do you mean by IQ measurement is underpinned by PCA (principal component analysis? or the requisite math?)?

Decent starting point: https://infoproc.blogspot.com/2020/06/twitter-attacks-and-defense-of.h…

Personally I don't think any of this is particularly related to "dangerous questions," but you do you I guess. In which he downplays racial bias in police shootings: https://infoproc.blogspot.com/2020/06/re-post-joe-cesario-on-police-de… In which he says lack of segregation is the reason for racial underperformance: https://infoproc.blogspot.com/2020/03/claude-steele-on-challenges-of-m… Interview with an open white supremacist, check: https://en.wikipedia.org/wiki/Stefan_Molyneux

Anyone looking for context. https://infoproc.blogspot.com/2020/06/twitter-attacks-and-defense-of.h… Does anyone have more information for the opposing side? Seems pretty terrible that people can take things you said over the years out of context and destroy your reputation.

This is a good discussion of the accounting of biofuels, with a guy called Tim Searchinger (there is a transcript if you don't want to listen): https://infoproc.blogspot.com/2019/11/manifold-podcast-23-tim-searchin… One takeaway is that you should think of farming trees as fuel as a competitor for land use. If it displaces growing food crops, then they will be grown elsewhere, possibly in newly cleared rainforest. Even if none of your wood comes from there. Edit -- I may have mis-remembered slightly... the food competition story would more obviously apply to corn ethanol, which could otherwise be turned into beef.

Physicist/geneticist Steve Hsu and linguist/neuroscientist Cory Washington have an extremely good podcast called Manifold. This episode in particular is a good introduction to the show: https://infoproc.blogspot.com/2020/04/vineer-bhansali-physics-tail-ris…