I'm not sure that the lack of a tuskegee link means much. If authority figures do a lot of bad stuff to a group, that group is going to, pretty quick, learn not to trust authority figures. Knowing about specific historical events seems not especially important to that. It's not as if Tuskegee was the only event of this kind. If my parents, friends, family, neighbors all say don't trust X, I probably won't trust X! That's a lot of people! I don't need to know one of the many reasons I shouldn't trust X is because of Y.
I respect Heckman but he is a good example of someone who grabbed hold of an idea early on and held on tenaciously in the face of alternative explanations. He has a paper out earlier this year showing that Denmark has no higher educational mobility than the United States despite vastly great socioeconomic equality and lower poverty. He spends a great deal of time explaining the various ways that family effects might influence educational mobility, without once mentioning genes or heritability - because genetic influence cuts against his narrative of the supremacy of early environmental influence. I always find it a little frustrating.
Cooper's take on Greenwald is pretty bad. He writes "Greenwald is treating the revelation of new information as proof that people were lying, when in fact these journalists were just reporting what was available to them at the time. He pulled this same childishly dishonest trick with the publication of the D.C. medical examiner's report on Capitol Police officer Brian Sicknick. 'Watch how easily and often and aggressively and readily they just spread lies,' he writes of a CNN segment with Erin Burnett and Don Lemon from the day of the assault."
But it is perfectly clear from this quote that Greenwald accused them of aggressively and readily spreading lies, not of "consciously lying." In other words, running with an convenient narrative that they couldn't be bothered to interrogate critically. The fact that Cooper can't make this distinction reflects very poorly on his writing and critical thinking skills. It is possible to aggressively spread lies through indifference to the truth without concsiously lying.
4. The write up doesn't convince me. I think intergenerational transmission of general medical hesitancy is possible, even if the original source is forgotten.
That said, I do think Tuskegee is a convenient way to rationalize different reactions to the same behavior by blacks and by deplorables.
I think the real winner from the Violent Language list is "picnic". Apparently some terrible people used to picnic during horrible acts of racial violence, so now we should all say "outdoor eating". I guess we're lucky the rule isn't that decent people should no longer eat outdoors, which would be an equally valid conclusion to draw from the example.
(I'm not trying to make a point or an argument, I'm genuinely curious here)
Can someone explain to me how this can be reconciled with the huge disparities in IQ over time in countries like Ireland and Greece as they became wealthier.
Also, if this data is meaningful, in what mechanism can the Flynn effect be impacting IQ if its not measurable as environmental?
On the AI media classification - it almost looks like the left-right political axis got mapped into a lower left-upper right axis in that graph. WSJ is still misplaced, but everything else looks more plausible.
On the IQ thing - do adoption agencies do a really good job of screening out adoptive families that live in places with high soil lead, or remnants of lead paint in the walls?
2. There was something of the sort in Spinrad's _Bug Jack Baron_-- I think the conspiracy was killing black children for (the rejuvenating effects of?) their glands.
4. I believe without evidence that greater medical neglect and incompetence would be enough to explain black people tending not to trust the vaccines without needing to bring in older history. I'm not sure if there's been research.
For that matter, I don't know whether white people who've had bad experiences with the medical system are less likely to trust the vaccines.
As to why there is no Journal of Datasets, well, dataset papers exist and do get published in fields like machine learning and computer vision. In biology and medicine it seems less common.
BUT, the main reason is not a technicality of whether a journal exists or whether you get a line for your CV. The publishability is a red herring and not the ultimate reason.
The issue is that scientists genuinely give you more "informal reputation" for coming up with interesting results, hypotheses and solutions to problems. Collecting data is boring grunt work that isn't really all that impressive from a subjective point of view. There are exceptions like the ImageNet dataset with did help Fei-Fei Li become really famous and successful, but that's computer vision again.
It's similar with all the complaints about publication bias, the file drawer effect etc. and the idea that a Journal of Negative Results will improve upon this state of affairs. However, the issue is not with whether there is a journal or not. The issue is that experts and scientists will learn your name and help your career if you find some interesting result that makes them go wow. Just because there's a journal for negative results won't mean that it actually gets you true recognition. And in academia informal connections are everything.
Under point 15, the following is a poor conclusion.
"Plausibly SpaceX will still have to follow the rules of its host country, the US, so the big loser here will be China "
China will not be a loser. Tesla has significant assets within China, with the CCCP not distinguishing corporate fictions of Tesla and SpaceX. Additionally, China demonstrated anti-satellite capabilities in 2007, which lends itself to slightly more force more than a simply fist shake. https://en.wikipedia.org/wiki/2007_Chinese_anti-satellite_missile_test
Starlink as anti-censorship technology will only work in peripheral nations.
There are journals for datasets (e.g. Journal of Open Psychology Data), but they are not used much in most fields for two reasons: 1) Scientific culture of valuing exciting (often nonreplicable) discoveries over high-quality data; 2) New-ish journals are almost never popular because it takes a long time to get an impact factor and overturn the inertial prestige held by the older journals, thus almost no one submits to them.
The article linked about the Tuskegee issue was frustrating to read, because it never seemed to attempt to explain why, if not Tuskegee, would Black Americans be more *hesitant* to get a COVID vaccine. There was a lot of stuff in there about unequal access and the problems that Black Americans who *want* to be vaccinated face in achieving that goal, but that's a different issue. I suppose what the article could be implying is that if you've been dealing your whole life with a system that refuses or hesitates to treat you when you want to be treated, when it starts beating down your door begging you to let them treat you for something else, you get suspicious. But I would have liked to have seen some quotes or data that actually indicated that.
It could also be true that the article wasn't ever really trying to answer the question "why don't Black people trust the COVID vaccine", but instead was saying "if you're looking at overall vaccination rates by race and noticing that Black people have lower rates than Whites, don't conclude that it's because they don't want to be vaccinated because Tuskegee", but again, it never said that explicitly. Also, I got a chuckle when it described "religious reasons" as a "modern" objection to vaccines, and once again I think that certain White people are continually confused when Black people have conservative opinions/values even though they are the most reliable Democrat voting block.
RE:Geothermal - I can say that there isn’t any real fear of wind and solar within the oil and gas industry. But geothermal elicits a “yea that’s a thing that could actually work”. Part of that might be that building geothermal capacity looks mechanically similar to E&P (exploration and production) because they both involve sticking things into the earth.
Ultimately, geothermal will run into the same problem fasting does - there’s not really a product anyone can get rich making so the proponents aren’t all that financially incentivized.
"Semiconductor manufacturers have 'about as much built-in voodoo and superstition as a major-league baseball team'"
I will observe that this book was written in 1985. About the semiconductor industry of the early 1980s. Fab process control has gotten a lot better since then.
I can't see the full article, but a key component of this idea relates to the female's use of chemical contraception pills which skew female preference.
The idea being that she'll choose overly similar males as company when she's on the pill with the idea of female spending more time around brothers and fathers and uncles when they are already pregnant. While females who are not on the pill will be more attracted to men who are different to them.
There are many such stories linking this to divorce rates and many stories of women who go off the pill to get pregnant with their partners, only to suddenly become repulsed by their smell and unwilling to have sex with their partner.
Nearly every woman I've talked to about this will know women in their social circle to whom this has happened. Making it clear to me it is not a rare event.
The contraception pill works to make the woman's body act as though it were pregnant. Leading to 'incorrect' selection of males who are similar to them as mates.
So I'd postulate what we may well be seeing as 'random' is in fact a mix of women not taking the pill selecting men with different immune system and women taking the pill selecting men with similar immune systems...leading to no 'overall trend'.
#2: Aldous Huxley mentions Adrenochrome in the fourth paragraph of The Doors of Perception, which was very popular (It's still the #1 search result for books about psychedelics). That's probably where most of those post-1954 fiction writers heard about it.
Adrenochrome is not an imaginary drug, although most of its uses are. The headline of the Daily Beast article is incorrect, and so is the implication that Hunter S Thomson made it up 20 years after Huxley.
Huxley implies adrenochrome is a powerful endogenous psychedelic and explains the then-current but now-disproven theory that adrenochrome contributes to schizophrenia.
See also this paper on adrenochrome inhibiting ATP production and probably not being very safe:
"Perfusion of the heart with 50 mg/L adrenochrome induced a marked decline in contractile force within 5 min and this was associated with a rapid decline in the myocardial ATP/AMP ratio...Adrenochrome at concentrations of 20 mg/L or higher was found to inhibit the oxidative phosphorylation activities of heart mitochondria under in vitro conditions. "
3 is complete bullshit. It should read: "While the Aztecs were sacrificing prisoners to the gods, their neighbors in Tlaxcala were a confederation of four despotic kingdoms that also sacrificed prisoners to the gods, because Aztec culture and Tlaxcalan culture were almost identical."
27 (real estate) seems like further evidence for my "contracts make us dumb" hypothesis. The more long-term contracts people sign, the more that locks us into doing things that no longer make sense if you're looking at objective reality, but seemed like a good idea at the time the contract was signed.
30. There are some interesting ideas in this article, but the Voxsplainer massively oversells its promise - and it is not internally consistent.
The total energy flow out of the earth is about 30 TW. Total human energy use today is 15 TW.
Geothermal energy is not uniformly distributed, but you can use that assumption for some rough estimates. About 2/3 - 3/4 of this heat comes up in the oceans (the Earth's surface is 71% ocean). Land-based geothermal cannot cover total human energy use. The US covers about 1/50th of Earth's surface, so even if we are able to harvest all of the geothermal energy, it would be about 0.6 TW. The US uses about 3 TW.
That doesn't stop Vox from claiming "5,157 gigawatts of electric capacity" (i.e. 5.2 TW) from geothermal in the US is possible. For this to even be physically possible, the US would have to be 10 times more volcanic than usual - which it's not.
Current total geothermal energy production globally is about 15 GW, so 0.1% of global energy use. [I wonder if Vox assumes that people don't keep track of TW vs GW vs MW.] This is concentrated around 'hot spots' - unusually volcanically active areas. New technologies could increase this significantly.
The technology Vox describes is basically using fracking technology (pumping pressurized water through rock) to access more geothermal energy.
This would start as a way to expand the area around a hot spot where geothermal is viable. After the technology is perfected there, the promise is that it could be used "almost anywhere". This is not quite right - it should say "almost everywhere". If we want geothermal to compete at close to the scale of fossil fuels, we need to tap a significant fraction of all the heat flowing up through Earth's crust.
Vox promises that this is not fracking and will not cause any seismic problems, as long as it stays away from seismically active areas. This is inconsistent - I'm not sure where they're planning on finding volcanically active, but not seismically active areas.
There is an even bigger problem for "almost everywhere" geothermal. Rock conducts heat really poorly, unless it contains fluids - either water of magma. There is magma near hot spots, and water in aquifers close to the surface. Small amounts of water can be pumped down, but not enough to saturate all the way down to the mantle. You will extract heat from the rock that is in contact with the pumped water, but that rock will reheat from lower rocks extremely slowly. This should be thought of as mining the heat of the crust. It is not renewable on the timescales it can be extracted.
Using technology developed for fracking to help geothermal is a good idea. I think it's plausible that it could increase geothermal energy production by a factor of 2-10. But claiming that it could compete at the scale of fossil fuels or that it "solves energy" is completely ridiculous.
Further context on #37: "phase of the moon" is a classic idiom in tech for inputs to a process that are fundamentally mysterious or random-seeming. (http://www.catb.org/jargon/html/P/phase-of-the-moon.html) Naturally the link includes two more examples of issues that were actually linked to the phase of the moon.
Re #17. In medicine, data are often collected under very a strict protocol to preserve privacy (see HIPPA). This makes it challenging to post datasets publicly. However, there are at least a few higher tier journals, such as PLoSOne, which insist that data be made public unless authors can provide good reason not to. The journals typically suggest one of several repositories. I've posted numerous datasets to the Inter-University Consortium for Political and Social Research (ICPSR) repository for papers I've published in PLoSOne, JAMA, JAMA Open, and BMJ Open.
"Students trying to prove the hypothesis were less than half as likely to notice the gorilla than those investigating aimlessly. I’m agnostic as to whether this proves something important about science or is just funny."
They're students doing homework. They tested what they were asked to test and then went on with their lives. Doesn't tell you anything about real scientists.
> In 1922, someone wore a straw hat after official stop-wearing-straw-hats date September 15, leading to the week-long Straw Hat Riot in New York and several hospitalizations.
After which the culprits were punished by magistrate Peter Hatting. Very appropriate.
Well, I just filled out the form for the arranged marriage. It asked me for my age, income, a recent picture, and gave a miniature Big 5 personality quiz. Wish me luck, I guess.
"Not sure if this counts as “the left is eating itself”, but the (unofficial) Brandeis Suggested Language List now recommends students not use the term “trigger warning” because it is “violent language”."
Well, you're linking to the Prevention, Advocacy & Resource Center website of a private university, so I don't really see how that would count as "left" politics. Can you explain to me your reasoning here, Scott Alexander?
Re: 17 That explanation doesn't make sense. Data sets are still published material. If someone publishes a derivative of your work using your dataset without crediting you on the paper (IMO is a co-author is a minimum professional courtesy) they are clearly committing academic fraud of some sort.
Re: Media Bias, aka #19, what the heck is that chart? The colors and bubble size is just based on their own pre-assigned metric? How is that the choice you make, when the color and bubble size are the most visually striking parts of the series. The color shade should be how biased they really are in your analysis, and the bubble size should be some sort of viewership/readership metric.
From the Wikipedia link on straw hat riot: "The riot itself began on September 13, 1922, two days before the supposed unspoken date, when a group of youths decided to get an early jump on the tradition." So the description "someone wore a straw hat after official stop-wearing-straw-hats date September 15, leading to the week-long Straw Hat Riot in New York " doesn't seem quite right.
3. Tlaxcala is fascinating. There was a interesting piece a while back about how Cortes is seen as a villainous figure in most of Mexico - with the exception of Tlaxcala.
It also points to how the typical way we're taught about the Spanish conquest is misleading. Cortes didn't just show up with 500 men and roll over an empire - he allied with the empire's enemies, and led a coalition that toppled said empire. To a lesser extent, something like that happened with Pizarro and the Incas - a lot of his local allies were folks who had been on the losing end of a civil war that happened after the Emperor and his heir died from disease.
15. I wouldn't go that far with Starlink, but SpaceX genuinely is a treasure, and if they can get Starship working at the price-points they're aiming for it could have a revolutionary impact on space industry and utilization. Imagine affordable suborbital tourist flights, much cheaper access to space, and so forth. It's great stuff - I just want them to start making their fuel from green hydrogen and air-captured CO2 sooner so it won't produce so much emissions.
30. I feel like they should go for broke on geothermal, see if increased funding and investment (plus some actual contracts) can get super-hot-rock geothermal to work. Maybe you could sell it to Congress as a way to reuse some of the infrastructure from derelict coal plants, since both of them would need to generate steam to provide electrical power.
One fascinating thing you could do if you had ample geothermal power plus plentiful fresh water (or we got better at desalinating seawater) if use it to produce e-fuels for export. That's not economical yet compared to regular fuels, but if you have to phase out regular fuels and really want to get some more mileage out of your gas and oil power systems you could do it (although I think in practice such will mostly be aimed at ships and airplanes as the market).
This all reminds me of Eli Dourado pointing out that if we really didn't care about Old Faithful or Yellowstone, we could turn the whole Yellowstone area into the Mother of All Geothermal Power Plants and provide electricity for half the country.
> By the way, this is the same Tegmark who in 1998 developed a leading theory for what the universe is and why it exists at all.
Leading with whom? All I can find is people damning it with faint praise. The "reception" section on the linked Wikipedia page is one guy calling it "provocative." The physicist quotes that he uses to advertise his book on Amazon seem lukewarm; every physicist writing a review is carefully mincing their words so as to say something positive while distancing themselves from even the perception they like the theory. "Whether you believe it or not..." "You don't have to necessarily agree...", "You may recoil from his thesis, but..."
I'm interested to hear if it made an actual splash with serious scholars in either physical or philosophical fields. My impression is that it didn't; it made a splash in the science pages of some major outlets during his book tour (for what is definitely an engaging book, even if it's trying to motivate an ill-defined theory!) and then everyone more or less moved on, because there wasn't all that much *there*.
The reason it didn't make more of a splash is that the theory is far been more sensational than substantive. It's not bad physics, because it's not physics at all - it's bad philosophy instead. It adds a modern aesthetic to mathematical monism, but doesn't bring any really new ideas to the table. Every mathematically possible universe exists and we're in one, and they *are* math rather than being *described by* math? "Worse Leibniz" pretty much captures it; the idea was better put, and more completely fleshed out, in the frickin' Monadology.
The most frustrating part, to me, is his signature argument in favor - that Occam's Razor prefers it, since the theory has "no free variables" and no perception can argue against it, since it predicts all of those perceptions. But Occam's Razor cuts the opposite way: a theory is good if it doesn't *have to* propose gigantic numbers of entities, like vast arrays of possible worlds or gremlins behind the scenes, in order to account for observations. The original phrasing is literally (translated directly) "entities should not be multiplied beyond necessity." I have no idea how the heck Tegmark thought *that principle* could support a view which basically says, "multiply out every permutationally possible mathematical entity, they all exist." (I know several more refined versions of the Razor have been proposed - but even the best one won't support his argument in the way he needs it to. A mathematical ontology is not, at its core, any simpler than any other.)
#2 Mentions Dune, and adrenochrome harvested from children seems like something Baron Harkonnen would use, but the article actually references "Destination: Void, from a different book series by Frank Herbert.
Re: 19, has anyone here _ever_ had a semantically understandable category fall out of embeddings without feeling like you're trying to force it? I've done a ton of work in that area and whenever I try and understand what the program is doing in real life terms it just seems like nonsesnse.
It looks like (4) studies the link between knowing Tuskegee and refusing treatment in other diseases before 2020, and also finds there isn’t a difference in actually getting treatment. While there is such difference in Covid vaccines. And in the interviews they note Tuskegee is occasionally mentioned, but not as much as others, but I’m not sure those other reasons apply to COVID vax, as they make more sense in a context of active treatment
On Tuskegee, this article (https://academic.oup.com/qje/article-abstract/133/1/407/4060075) published in a top econ journal finds disclosure of Tuskegee in 1972 leads to "increases in medical mistrust and mortality and decreases in both outpatient and inpatient physician interactions for older black men" by comparing changes in the health of older black men closer or farther away from Macon County (county of Tuskegee). It uses difference-in-differences-in-differences and D-i-D style techniques have come under increased scrutiny in recent years. With that caveat, if there was an effect on black adults in 1972, that would likely persist to today through social norms in the black community on medical distrust and going to the doctor even if black people today do not know about Tuskegee. Though someone should look at re-doing this paper's strategy with vaccine hesitancy as the outcome and better D-i-D methods.
Regarding #25. The paper says, "odds ratio = 4.8, P = 0.034, N = 33, Fisher’s exact test". Don't use it. Fisher's exact test has low power.
Lydersen, S., Fagerland, M. W., & Laake, P. (2009). Recommended tests for association in 2×2 tables. Statistics in Medicine, 28 (7), 1159–1175. http://doi.org/10.1002/sim.3531
(This comment is written by someone who would probably not have seen the gorilla)
On straw hats, doesn’t that link say the riots started two days before straw hat day? That feels very human, ‘No! You can smash up my hat if I was wearing it on the wrong day, but it isn’t the wrong day yet, you bastard, let’s fight!’
@37: My favourite sentence in the Wikipedia article for Alchemy is in the section about its transition into chemistry:
"Boyle would note the place in which the experiment was carried out, the wind characteristics, the position of the Sun and Moon, and the barometer reading, all just in case they proved to be relevant."
Up until now I thought it's both cute and inspiring, a combination of a scientist's discipline and a 17th century intuition about the world. Well, the passage about semiconductor yields made me realize I was unfair to Boyle, and that considering the positions of celestial bodies makes sense even today.
I once attended a scientific meeting where one group had gathered a large amount of brain scan data and urged us to poke around in it and see what we could find. Others at the meeting urged us to not explore the data but only test specific hypotheses on it, because if we looked we would be tainted and would only test hypotheses on the data we already knew the data would support. I was inclined to agree with the first group, but I didn't know whether that's just because that's my style and I'm not a careful enough scientist. Anyway, that's what the gorilla made me think of.
That Brandeis list veers between "okay, I see the point here" and "if ever you hear or see me use this term, shoot me and bury me in a ditch":
"Gender exclusive language:
You guys, Ladies and Gentlemen
Gender inclusive language:
Y'all, folks or folx, friends, loved ones, people, everyone
These examples either lump all people under masculine language or within the gender binary (man or woman), which doesn’t include everyone."
I wonder where the use of "lads" (which we used to refer to each other in the collective sense in my all-girl school) falls on this? And if ever I use "folx", you may be sure and certain that that last remaining dim flicker of reason and sanity has burned itself out and I am now a drooling imbecile.
@19: If it's any consolation, the media bias study seems to be a small side-project of his main thing right now, which is creating an artificial physicist - a symbolic regression algorithm where experimental data comes in and equations come out.
"Study shows Donald Trump is grumpy when he does not get enough sleep. Being an important enough figure that scientists study you personally sounds exhausting."
I await with breathless anticipation the follow-up studies "When caught out in the rain without an umbrella, Donald Trump gets wet" and "If he has not yet eaten that day, Donald Trump gets hungry".
But how can we be sure this result replicates unless we get data from figures on the other side of the political divide? Has anyone studied whether or not Joe Biden gets more cheerful the less sleep he gets?
Are people really that hard-up enough for something to publish that they need this kind of "grass is green" obviousness, or was it a joke? If it was a sincerely meant paper, I think some "kick them out of the lab into the sunlight and fresh air" is badly needed.
7: Although the thing about "trigger warning" is in the section on "violent language", the actual reasons given for maybe preferring "content note" over "trigger warning" aren't about the implication of violence at all. (I think there's at least one other good reason for usually using something other than "trigger warning": there are reasons why people might want advance notice of potentially-upsetting things besides outright PTSD-like risk of being "triggered".)
>Imagine a future in which Earth’s dominant ISP is constructed, maintained, administered, and governed by Martians, who operate it for free as a public-relations campaign.
Please, can we colonise a saner place like the Moon, the asteroid belt or the Venerean atmosphere? Mars is one of the worse choices because it's got the downsides of a planet (need chemical rockets to take off due to gravity/atmosphere, extra delta-V from gravity well) without the upside (atmosphere thick enough that you don't need a pressure hull or radiation shielding).
Re #16: This chart has been bugging me since it came out. It seems too cute by half. Does the vaccine really reduce mortality by the same amount across age groups?
Semaglutide: "Yet GLP-1 had serious problems. The first was that it was degraded in the blood by an enzyme within a few minutes. This meant that it wasn’t useful as a drug in its native form because it would have to be injected far too often. This problem was solved by using another version of the hormone identified in the saliva of the gila monster lizard..."
I feel as though there is an untold story behind that last sentence.
#17 Could you segment or subset the data. I really think we should stop incentivizing papers based on unpublished data, much of it gathered with public funding. The replication problem is crippling public confidence in much bio-medical and social science research.
#21. hatis problems are historically deep. They can be over explained based on its history. E.g. the French forced the independent country to pay them an indemnity in the 19th Century. The Spanish did nothing like that to the DR.
#16 Not sure of the relevance, since it seems to me that most parents ordinarily want to expose their kids to massively less risk than themselves in general.
Re: War Against The Weak, even academic historians (very anti-eugenics, and very likely to classify things as eugenics) found Edwin Black hyperbolic and sensationalist, especially when it came to drawing Nazi connnections.
Also re: sterilising the blind, separating views on the desirability of someone passing on their genes from views on their capabilities as a parent is difficult, and historians tend to ignore the latter despite all the evidence that it was a strong factor in what is categorised as "eugenics".
#31 My daughter was prescribed semiglutide under its Type II diabetes label of Ozempic., but it was primarily for weight loss. She lost 100 lbs. She perceived the mode of action to be causing feelings of repletion to occur upon the consumption of very small amounts of food and for the feeling to last much longer than usual, e.g. 6-8 hrs instead of 4-6 hrs.
Thus encouraged, once the Wegovy formulation was FDA approved, I got my doctor to prescribe it. I have taken the first two 0.25 mg dose. The ramp up to a full therapeutic dose is 16 weeks (02.5, 0.5, 1.0, 1.7, 2.4 mg). No effects or side effects yet.
Medicare Part D will not cover the medicine, or any other weight loss medicine either. It is very expensive. $1325/month. if it works, it is worth it. if not, well weren't going to Italy this year anyway.
>By the way, this is the same Tegmark who in 1998 developed a leading theory for what the universe is and why it exists at all. I feel like going from “discover fundamental nature of the universe” to “attempt to investigate media bias, but it has glaring flaws” is a slightly-too-on-the-nose metaphor for the past 25 years of science as a whole.
This is also the same Tegmark who is President of the Foundational Questions Institue, science director of the Future of Life Institute, and has been a major figure in the advocacy for AI Safety (he wrote Life 3.0).
I'm surprised Scott ommits this. This paper on AI bias seems to be almost a kind of passtime. That's the "metaphor for the past 25 years of science as a whole" - there are many accomplishments from remarkable polymaths we don't even know about - though you could just link his Wikipedia or MIT page.
"Now a Haiti econ blog (appropriately named Vodou Economics) proposes a much better explanation - the data are wrong, and the DR was always ahead of Haiti."
It's not so much that the data are "wrong" so much as they are for the 2011 ICP rather than the 2017 one. The 2017 one shows a much larger gap between the Dominican Republic and Haiti (tenfold as of 2017). The growth statistics are the same, which is arguably what should matter more here. The Maddison people (who, it should be said, are more often than not total morons) discarded some Braithwaite (1968) PPPs for being unreliable -but that shouldn't matter here.
Re: #21 Any one visiting Haiti & the Dominican Republic would have a hard time believing they are on the same island. Despite having similar histories (former colony, revolution, foreign occupation, dictatorship, eventual democracy) the two nations could not be more different.
Having direct considerable time in both I've found that the contrast is the result of what each people see as their heritage.
Dominicans see themselves as “European” with social/political/economic/cultural values and ambitions inherited from Spain.
Haitians see themselves as "African” with social/political/economic/cultural values and ambitions inherited from West Africa.
19: Looking for words or phrases with high predictive power with regards to politics actually merely measures political jargon, not bias, although there is obviously a link to bias, as bias is encoded in some of the jargon (or how it is used).
Of course, it much less of a weird mess. Why do those Spanish greenhouses have these weird shapes? No reallocation of farming land (Dutch farmers traded a lot of land in the past to get nice rectangular plots)?
35: Of course, Germany has states with rather strong cultural differences between them. For example, Prussia was famous for it's military prowess. Prussia was dissolved and many ethnic Germans migrated to other parts of Germany. But it's perfectly plausible for Mecklenburg-Western Pomerania, Brandenburg and Saxony to have a different culture to the other states.
I disagree with the commenters saying that datasets don’t make good papers in biology. Well-curated, accessible datasets with some analysis make great papers, get published in leading journals, and get cited hundreds of times by people re-analyzing the data. Based on my personal experiences I firmly believe the real reason researchers are bad at data sharing is because they’re afraid that other researchers will re-analyze the data and find their mistakes (or their fraud).
Devon's piece on Prospera is excellent, though every time I see documents by the National Lawyers Guild (a critique of the ZEDEs linked under "Other Resources" as a legitimate recognition of more critical perspectives), I recall their 2003 report on North Korea, which to the best of my knowledge they have not repudiated. This is from a summary by Christopher Black, one of the delegates, still seeing the great revolutionary dream in DPRK in 2003,
"In 2003 I had, along with some American lawyers, members of the National Lawyers Guild, the good fortune to be able to travel to North Korea, that is the Democratic Peoples Republic of Korea, in order to experience first hand that nation, its socialist system and its people. The joint report issued on our return was titled “The Grand Deception Revealed.” That title was chosen because we discovered that the negative western propaganda myth about North Korea is a grand deception designed to blind the peoples of the world to the accomplishments of the Korean people in the north who have successfully created their own circumstances, their own independent socio-economic system, based on socialist principles, free of the domination of the western powers."
"As in Cuba and other one party socialist societies, North Korea has a system of direct democracy in which elections are held for local peoples committees, district and provincial committees and to the Supreme People’s Assembly. The absence of other parties is not considered a failing, as the entire society is socialist. The question of multiple parties did not even seem understandable to those we spoke to. The delegation questioned whether within that system, there is in fact more participatory democracy than in the American federal system or the parliamentary system in which democracy ceases to operate once the elections are over. It is more circular, with local committees sending up to the next level requests, complaints and so on and so on up to the national level with discussion, at least in theory at these levels and then feedback to the local level until an agreement is reached based on resources available and circumstances.
However, the issue is not whether we agree with DPRK’s system or feel that our democracy is better or more just. . . . Certainly we cannot say that only one political system is successful and generates a participatory and healthy society. We hope that future delegations can learn more about political dialogue within the DPRK system and share the pros and cons of our system without blame or judgment."
Aetna requires prior authorization for Wegovy, the brand name for semaglutide. I tried setting up a Teladoc appointment to get a prescription, and the doctor said she’s not allowed to prescribe “lifestyle medications” and anyway semaglutide is for diabetes.
I don’t interact with the medical industry enough to know if this just typical friction or not, but so far I’m inclined to write this medicine off as vaporware for now.
37: This doesn't surprise me at all. I used to work for a chemical company that built a plant in South Korea, meant to be a copy of a North American plant. Same blueprints and everything, but for at least 6-8 years the chemists and engineers, from Korea or the US, couldn't get it to produce the same purities as the US plant. So far as I know they never figured it out. Apparently somewhat random outcomes like that are fairly common with large scale chemical production, so apparently there are many strange little influences on the process.
Are you telling me that the under-age-15 hat-smashing boys were "spanked ignominiously" and just sent home? No criminal record to prevent future employment? No psych eval and prescription medication? No movement to a stricter alternative school?
I don't think they probably learned any lesson if the authorities didn't wreck their subsequent decades.
It is interesting to me that they are trying to establish this *new* philosophy of human development on the basis of capitalism. They have patrick collison and peter thiel among others on board as, i guess, "thinkers" and patrons, to define progress in their vision. It's no doubt these people are heavily biased towards start up esque systems and will not want to build a future without them, at least, but it's more than likely they will be building that future on the very basis of them. I wonder if they really believe this will turn out different than the 20th century, if they really believe half the problems of our age were not in fact caused by those very systems. In my view skepticism, bureaucracy, and stagnation, is an inescapable consequence of a mature capitalist system (though a young system doesn't have those problems as much). I don't know why they're trying to build the future in this image but it's not going to work.
On insulin and weightloss. I have recently been prescribed metformin for whatever "pre-diabetes" is properly called.
What I have noticed is that I feel warmer - I don't notice being cold or being in a draft as much.
Hypothesis that it may be raising either core body temperature, or (more likely) peripheral circulation, which would make my skin, fingers, etc warmer. That warmth would actually mean that I have more heat loss. That would mean that I'm burning more calories to maintain core temperature - which would be an obvious weightloss mechanism, and I know there is some (mixed) evidence for metformin for weightloss, and also explain why it is mixed - if it works through making it easier to tolerate lower temperatures, then you only get the weightloss if you actually experience lower temperatures.
Anyway, proposed approach: monitor room temperatures for people taking possible weightloss drugs, and see if a drop in room temperature correlates with more weightloss.
Unsure whether it's right to frame East-West migration in Germany before the Berlin wall as "fleeing". I realize the term comes from the abstract of the article (which I have skimmed over and have not yet made up my mind about) so this is not on you.
#36 I agree with much of the criticsim of Henrich's WEIRD argument. I have made similar comments in other fora.
The short version is that Henrich identifies canons of the Medieval Catholic church forbidding cousin marriage as being a key development in the foundation capitalism in Western Europe.
My response is that those rules no longer impacted English society after the Reformation. England is indisputably the home of the industrial revolution. But cousin marriage was common. E.g. Charles Darwin.
It is a key plot point in Jane Austen's Mansfield Park, but is no where mentioned as an objection to the marriage of Edmund and Fanny. In Trollope's Can You Forgive Her, Alice Vavasor is infatuated with her cousin George. The objection to George is not that he is her cousin, it is that he is a jack@$$.
3.) They're somewhat exaggerating the differences in government. Tlaxcala had a monarch and the Aztecs had noble and common political institutions, at least if popular memoriy is to be believed. More interesting is the geopolitical situation. Tlaxcala would be like if an even more brutal version of Rome had risen to power except a state based next door to Rome itself had managed to stay independent and stave off conquest by extreme militarization, long term acceptance of economic deprivation, and alliances with dissident factions in the Aztec empire.
This is a huge context to Cortez's conquest that gets missed. This is why Monctezuma didn't just immediately try and kill Cortez. There was a wider political situation he had to defuse or face a larger revolt/invasion. It's also why driving Cortez out of the city didn't work: Cortez ran back to Tlaxcala and Monctezuma couldn't pursue him without getting past the giant Tlaxcalan-Rebel army. This gave Cortez the time to regroup and engage in more open warfare. (Also, technically Tlaxcala wasn't conquered: they were direct vassals of the Spanish Crown. It was only on independence they were incorporated into Mexico.)
4.) This is overly literal. I have no doubt it's rare. But certain communities have developed intergenerational distrust of the government, often with good reason. This is just a particularly emblematic example that's come to the attention of the chattering classes.
12.) It's worth nothing that the Constitution was a barrier to this kind of thing. Yes, the Supreme Court eventually allowed it. But it took decades to get there, decades that were used in other places to do it even more widely. The most thorough eugenics experiments took place in the territories where the Constitution didn't apply. Likewise, it was often worse in places with less robust conceptions of rights. Rights may forestall good and necessary changes. But they also forestall a lot of evil ones.
15.) It'll probably be the same as how radio and television already work. Lots of little pirate signals in places with significant censorship. They'll be tracked down by the regime in erratic but furious enforcement. In places with freedom of speech the big, legal networks will dominate because they can outspend/out-quality little pirate operations. More importantly, I expect us to get Radio Free Europe: Internet Edition at some point. If we can turn on the uncensored, untraceable internet in North Korea or Manchuria that would undoubtedly be to our advantage.
I'm also curious to see how it affects brands of authoritarianism like Putin. Putin doesn't really do direct censorship: you can go and Google anything you want about him. He even has real elections. He just slants everything as far to his side as he can. Government regulators lean on broadcasters and other organizations. Fraudulent prosecutions go after the wealthy who oppose him. Opposition journalists are attacked. Etc. This might make his form of authoritarianism more durable than the old totalitarian kind. Giving Russians free access to the internet wouldn't change as much as it would in China.
35.) Not to mention the the different regional effects of Nazism. East Berlin/Germany had more Poles, Catholics, and Jews, all of whom didn't fare too well under the regime. Or that being occupied by the Soviets and being occupied by the Americans is a rather different experience. Berlin was conquered by the Soviets wholesale but Americans took their sector over on July 4th 1945. The period from 1945 to 1949 was really turbulent and critical. I think it's very fair to see west vs east Germany as a comparison of two economic systems, especially when comparing specific industries. Food distribution, for example. It's less fair to blindly attribute all differences.
For example, I've long pointed out that the East German female workforce participation rate was higher than the west's in Imperial Germany. So while Communism no doubt did something to get women into the workforce the differential was already somewhat there. Conversely, west Germany has been richer than the east... basically forever. So while it's fair to point out Communism almost certainly did have downward pressure effects simply pointing to raw GDP is too simplistic.
At this stage it's straight up unscientific to be unwilling to even consider the possibility that biology explains much of the difference between Haiti and DR. The difference in outcomes between these countries is only surprising if you completely ignore the outcome differences between every sub-saharan african country and every other mestizo-majority country, and the socio-economic outcomes of these populations within white majority countries.
"34: It’s generally-believed that IQ is mostly-environmental early on in life, but mostly-genetic after adulthood."
What am I missing? I always thought that IQ was an attempt to measure native intelligence, not functional intelligence -- that is, not acquired abilities or knowledge. So it's hardly surprising that "genetics matter vastly more than environment and environment more-or-less doesn’t matter".
And because IQ is more or less impervious to environmental influences (except for prenatal and childhood malnourishment and diseases), the more interesting question to me is what educational practices are best at equipping the rising generation with knowledge and skills?
The trigger warning note sounded pretty interesting and I was going to share it around - except it turned out to be a reasonable note on the linked site. It says to use ‘content note’ instead which seems more professional frankly, and euphemisms like trigger warning - which seems obviously rooted in gun imagery - sounds pretty crass in a society with (real or apparent) endemic gun-problems.
>"IQ of biological parents explained about 42% of variance; IQ of adoptive parents explained only 1%. 'Variance explained' is a non-intuitive statistic with a lot of confusing properties - but in this case the obvious conclusion that genetics matter vastly more than environment and environment more-or-less doesn’t matter at all is basically right."
That "obvious conclusion" is the opposite of what the paper said. Look at the far-right column of table 3. Heritability explained 42% of variance in ICAR-16 , but environmental variation explained 50% of the variance in ICAR-16, *higher* than the heritability.
(The middle three columns account for the fact that ancestry and environment are correlated [me and my parents all grew up in Toronto etc], and gene-environment non-additivity [a gene might give me +1 IQ point if I live by a lake but +2 IQ points if I live by the ocean etc].)
I think #34 significantly understates the problems of variance compression in adoption studies. The most straightforward version of this problem is that adoptive families have low variance in IQ, with few very low-IQ families, as a consequence of the fact that adoption agencies are picky. Meanwhile adoptive families vary a lot, but with a strong skew toward low IQ. This means that you end up with low shares of total IQ variance explained by variance in adoptive parent IQ, but you can adjust for this by reweighting variances to reflect the variation in parental IQ in the full population.
The more serious version of this problem is that for parents who are on the margin of acceptability for adoption agencies, you'll have a negative correlation between IQ and other positive traits (like conscientiousness) even if the correlation between those traits is positive in the general population. If a low IQ parent has been accepted as an adoptive parent, they must have been able to check the boxes of middle-classness, stability, strong references, etc., in spite of their low IQ. The most likely reason for this is being above average for adoptive parents in non-IQ aspects of parent quality. On the other hand, a higher-IQ parent can get accepted as an adoptive parent even with some deficits (relative to most adoptive parents) in other qualities.
The same thing is true of biological parents--high-IQ biological parents who put their kids up for adoption are probably less conscientious, more prone to mental illness, etc, than are either other high-IQ people or even lower IQ biological parents who put their kids up for adoption.
But this is much more likely to bias estimated relationships with adoptive parents. The genetic causal mechanism is likely that high IQ genes lead to high IQ, even if paired with low conscientiousness genes. The environmental causal mechanism is probably that high IQ parents create better childhood environments, which leads to higher child IQ. The connection between conscientious parents and a good childhood environment is plausibly as strong or stronger than the connection between high-IQ parents and good childhood environment. As a result, a negative correlation between IQ and conscientiousness induced by selection into adoption is likely to negatively bias the correlation between adoptive parent IQ and child IQ substantially.
Background The association of physical activity with the development and progression of coronary artery calcium (CAC) scores has not been studied. This study aimed to evaluate the prospective association between physical activity and CAC scores in apparently healthy adults.
Methods Prospective cohort study of men and women free of overt cardiovascular disease who underwent comprehensive health screening examinations between 1 March 2011 and 31 December 2017. Baseline physical activity was measured using the International Physical Activity Questionnaire Short Form (IPAQ-SF) and categorised into three groups (inactive, moderately active and health-enhancing physically active (HEPA)). The primary outcome was the difference in the 5-year change in CAC scores by physical activity category at baseline.
...Conclusion We found a positive, graded association between physical activity and the prevalence and the progression of CAC, regardless of baseline CAC scores
Results
The mean (SD) age of study participants was 42.0 (6.1) years (table 1). The proportions of participants who were inactive, moderately active and HEPA were 46.8%, 38.0% and 15.2%, respectively. Participants with higher physical activity levels were older, less likely to be current smokers, and had lower levels of total cholesterol, LDL-cholesterol and triglycerides, higher levels of HDL-cholesterol, and higher prevalence of hypertension and presence of CAC than participants with lower physical activity levels.
Participants with CAC >0 at baseline were older, were more likely to be male and current smokers, and had higher levels of traditional cardiovascular risk factors.
The cardiovascular benefits of physical activity are unquestionable. ...Regular physical activity reduces the risk of many adverse health outcomes, including mortality, CVD, diabetes, hypertension, obesity and dyslipidaemia.
...High levels of physical activity, however, may be associated with a higher risk of coronary artery calcification. In a meta-analysis, higher physical activity levels were associated with a higher prevalence of CAC (pooled odds ratio (OR) 1.84; 95% CI 1.41 to 2.93)
...A possible mechanism underlying this association is that physical activity may increase coronary atherosclerosis. Potential pathways include mechanical stress and vessel wall injury of coronary arteries, physiological responses during exercise, such as increased blood pressure, increased parathyroid hormone levels and changes in coronary haemodynamics and inflammation. In addition, other factors, such as diet, vitamins and minerals, may change with physical activity.
The second possibility is that physical activity may increase CAC scores without increasing CVD risk. The standard Agatston CAC scores are calculated as a combination of calcium density and the volume of plaque burden. Higher calcium density, which suggests more stable, calcified plaque, produces a higher CAC score, however, it is associated with lower CVD risk.
...In general, progression of CAC scores is associated with a higher risk of CVD among individuals with CAC=0 and those with CAC >0.28 Moreover, the absence of detectable CAC is a strong negative predictor of CVD, whereas the presence of any CAC, even at very low levels, is associated with an increased risk of CVD. However, considering the undeniable protective effect of physical activity on CVD, the positive relationship between physical activity with CAC progression should be interpreted with caution as the complex interplay between physical activity, CAC progression and subsequent CVD risk remains largely unknown."
Conclusions seem to be "exercise is in general good for you but it could be bad for your heart. Or we could be measuring the wrong kind of plaque. Or we could be totally wrong about hardening of the arteries being a bad thing".
1. Hat Riot wikiepedia - "Many of those taken to court following arrests related to the hat-snatching frenzy opted to be fined rather than serve time in jail. The longest recorded time one of the teens was sent to jail was three days served by an A. Silverman, who was sentenced by Magistrate Peter A. *Hatting* during night court."
I feel like there's a ton of useful truth here, but it would really benefit from re-interpretation and summary of the type Mr.Alexander is uniquely good at.
The AI movie thing was fun, but it's pretty clear there were images of from some of the films or their posters in the training dataset. This is most obvious for die hard with the exploding building, and to a lesser extent space jam and point break.
Bioengineer here with a bit of perspective on the data sharing thing (17).
In a few sub-fields, this has been recognized as a problem and major journals require authors to put their data on big NIH-hosted databases. See https://www.ncbi.nlm.nih.gov/guide/genes-expression/ for examples related to gene expression and genetics.
At the same time, there's an undercurrent of discontent in certain fields of biology (I'm thinking mostly about systems biologists and developmental biologists) over the papers that are *just* datasets. We pretty regularly get papers that are, like, "A high-resolution cell atlas of a developing mouse hindbrain" or "comprehensive time-resolved whole-body single-cell gene expression of growing nematodes". Usually they're very impressive, technically, and produce a metric ton of data that can be accessed using a slick (but still kind of clunky) new custom visualization interface... and nobody knows what to do with it.
There's kind of an "okay, what did we learn?" moment that comes with a lot of those papers that's starting to be too familiar. It turns out that in the modern age, the hard part of learning about biology isn't collecting tons of data -- it's getting any useful conclusions out of that data. Like, imagine you have that high-resolution atlas of a developing mouse hindbrain. What do you *do* with that? It's easy to imagine lots of things you can check with that dataset *in a vague noncommittal way*, but actually concretely coming up with a specific hypothesis to check with that sort of thing is hard. So that's why putting giant datasets online isn't as prestigious as writing a paper that synthesizes something out of that dataset.
(I'm guilty of writing one of those papers, albeit a small one on a small subject. Back in my undergrad days I was on a paper that analyzed gene expression over the course of a bacteriophage infection. Collecting the gene expression data wasn't trivial, but in a way it was ridiculously straightforward. Figuring out what it meant was much, much harder, and I don't think we got all that much out of it.)
Incidentally, I don't have a good citation for this, but a while back I read that "science" (I think chemistry and biochemistry, more specifically) went through a similar crisis back in... I want to say the 50s and 60s? For a stretch, most of the "scientific research" that was going on was basically measuring parameters -- stuff like the rates of reactions -- to finer and finer precision. Eventually some big-name scientists got fed up with all the measurement papers and published some high-profile editorials to the effect of "science isn't about measuring numbers, it's about figuring out how stuff actually works!" and pretty quickly people stopped measuring parameters.
Personally, having written a bunch of models of biocircuits that rely heavily on the field's knowledge of underlying rate parameter values that the field definitely doesn't know, I think we could have done with a *bit* less of a course-correction on this. =P
The thing about the gorilla is totally consistent with cog sci literature on problem solving and the way specific goals constrain attention. For more, you can read my essay on cognitive load theory: https://cognitiveloadtheory.wordpress.com/
Why isn't there money to be made for hedge funds to go around and completely buy up the commercial mortgage securities and modify terms? Or for banks to offer different initial terms?
I feel something else must be going on, eg, maybe there are regulations for which book value matters more than actual expectations?
Since the comment thread where marxbro and I are having a fascinating discussion around cultural mores is becoming the length of the Equator, I'm going to drag this one up here so we can have a fresh starting point.
marxbro1917:
"I'm simply making the point that you are doing performative right-wing outrage. You say that you would dislike any politician doing it, but your own posts show that you don't really have any clear poltical analysis of why you like or dislike certain words. Likely you are just repeating what you heard elsewhere."
I, a parrot, repeating what I heard elsewhere and certainly not having any linguistic tastes of my own:
Squawk! 🦜
From "The Pilgrim's Regress", C.S. Lewis:
(T)he jailor addressed the prisoners and said: ‘You see he is trying to argue. Now tell me, someone, what is argument?’ There was a confused murmur.
‘Come, come,’ said the jailor. ‘You must know your catechisms by now. You, there’ (and he pointed to a prisoner little older than a boy whose name was Master Parrot), ‘what is argument?’
‘Argument,’ said Master Parrot, ‘is the attempted rationalisation of the arguer’s desires.’
...‘Good. Now just one more. What is the answer to an argument turning on the belief that two and two make four?’
‘The answer is, “You say that because you are a mathematician.”’
‘You are a very good boy,’ said the jailor. ‘And when I come back I shall bring you something nice. And now for you,’ he added, giving John a kick and opening the grating.
... ‘This psittacosis is a very obstinate disorder,’ said Reason. And she turned to mount the black horse.
There really are people who are up against logistical difficulties, need more information, and/or would like to talk with a human being about the logistics of getting vaccinated.
I was fascinated by your find on arranged marriages. I am really curious if they differ from how Indians do it. If that subject interests you, watch the Netflix show Indian Matchmaking. It is pretty real, I thought. Fascinating show.
I've become skeptical about the WEIRD book's argument ever since I learned that traditional Hinduism was every bit as uncompromising in banning close-kin marriages as medieval Catholicism, and enforced this ban for a lot longer.
Marriages between descendants of the same 7th-generation ancestor on the father's side and between descendants of the same 5th-generation ancestor on the mother's side are banned as far back as the Gautama Dharmasutra (600 BCE-200 BCE). This is reaffirmed in all subsequent ancient and medieval Hindu legal texts (dharmashastras).
(Note that the famous Manusmriti only forbids marriage between third-generation relatives on the maternal side--but the traditional commentaries explain that the 5th-generation standard overrides it.)
The texts explain that this rule applies to all four varnas, unlike the better-known ban on marrying within one's gotra (paternal lineage), which applies only to the twice-born castes.
And, at least according to a random Indian guy I talked to on Twitter, this anti-endogamy rule is actually observed in practice, and makes finding a suitable marriage partner difficult.
On #17 publishing data, I think this is partially true, but doesn't explain why researchers don't publish their code, which does not entail foregone future papers. I think it's because they don't have good code habits, and mainly because this gets you neither citations nor respect (though respect might be trending upwards slowly).
"I link here in penance: Trump’s False Lafayette Square Exoneration. "
One would think you've learned a lesson after the last passage of the Woke ripping your nuts off..... but noooooooo. One. More. Time. Never explain. Never apologize.
Phoney War seems just incomprehensible to me. If this was fiction, holding an idiot ball would be an understatement.
> German military commander Alfred Jodl said that "if we did not collapse already in the year 1939 that was due only to the fact that during the Polish campaign, the approximately 110 French and British divisions in the West were held completely inactive against the 23 German divisions."
How do you... not attack an enemy when most of its army isn't present?
Also, this:
> Leopold Amery suggested to Kingsley Wood that the Black Forest be bombed with incendiaries to burn its ammunition dumps, Wood—the Secretary of State for Air—amazed the member of parliament by responding that the forest was "private property" and could not be bombed
The criticism of Henrich and WEIRD was valuable, but itself misses many, many tricks. The elements he introduces as disproofs sometimes have easy explanations. Still, it's always good to ask newly fashionable theories that claim to explain everything to explain at least something rigorously.
Re. 17, prominent bioinformatics journals require publishing datasets and code. Unfortunately, bioinformatics may have the biggest and most-restrictive datasets of any discipline. This is a big reason why I never published anything while I was working in bioformatics: my work often used terabytes of data; my employer wouldn't give me more than about 100 megabytes of ftp space; I couldn't afford to host terabytes of data myself; and that data would be gathered or computed from dozens of sources which all had different privacy and redistribution requirements.
On #28 -- The reddit post might be relevant to some smaller landlords, but it doesn't really reflect decision making from large institutional landlords. I'm an asset manager with a 3+ million square foot portfolio. With office space, part of the issue is that a landlord almost always has to give upfront concessions to a tenant and in the middle of a recession those look way too risky on long term leases. If you can only get $26/sf rent with 3% escalations per year on a ten year lease and have to give $80/sf in construction allowance to a tenant plus 10 months free rent upfront plus pay commissions to your broker and the tenant's broker... why not wait a year when rents and concessions go back to normal? Also, demand for office space doesn't just dip in a recession, it goes to almost zero and tenants put their space on the market too, so you compete against super cheap sub-lease space. (e.g. the market clearing rent doesn't go from $30 to $26 per square foot, it falls to like $8 per square foot). If you just wait a little while you could get $30/sf rent again and only give a $60/sf construction allowance for a ten year lease. Decent office buildings in major metros trade for about a 14-20x multiple on annual income so you holding out on rent for a year or more pays off much better than locking in an $8/sf rent for ten years. Regular commercial landlords don't do 1-2 year leases because it takes a lot to build out the space and you amortize that over the lease term. WeWork and Regus will do short term leases and they capture all of that demand, but it's still light so they aren't gobbling up the remaining office inventory either.
I'm not sure that the lack of a tuskegee link means much. If authority figures do a lot of bad stuff to a group, that group is going to, pretty quick, learn not to trust authority figures. Knowing about specific historical events seems not especially important to that. It's not as if Tuskegee was the only event of this kind. If my parents, friends, family, neighbors all say don't trust X, I probably won't trust X! That's a lot of people! I don't need to know one of the many reasons I shouldn't trust X is because of Y.
I respect Heckman but he is a good example of someone who grabbed hold of an idea early on and held on tenaciously in the face of alternative explanations. He has a paper out earlier this year showing that Denmark has no higher educational mobility than the United States despite vastly great socioeconomic equality and lower poverty. He spends a great deal of time explaining the various ways that family effects might influence educational mobility, without once mentioning genes or heritability - because genetic influence cuts against his narrative of the supremacy of early environmental influence. I always find it a little frustrating.
Cooper's take on Greenwald is pretty bad. He writes "Greenwald is treating the revelation of new information as proof that people were lying, when in fact these journalists were just reporting what was available to them at the time. He pulled this same childishly dishonest trick with the publication of the D.C. medical examiner's report on Capitol Police officer Brian Sicknick. 'Watch how easily and often and aggressively and readily they just spread lies,' he writes of a CNN segment with Erin Burnett and Don Lemon from the day of the assault."
But it is perfectly clear from this quote that Greenwald accused them of aggressively and readily spreading lies, not of "consciously lying." In other words, running with an convenient narrative that they couldn't be bothered to interrogate critically. The fact that Cooper can't make this distinction reflects very poorly on his writing and critical thinking skills. It is possible to aggressively spread lies through indifference to the truth without concsiously lying.
How dangerous is Semaglutide? Is it a good candidate for use as a nootropic?
I guess it doesn't matter so much where the rain in Spain falls if you've got sprinklers.
PSA: Michaël Trazzi created a version of the "rationalist dating/friending" site reciprocity.io for Twitter. It's called "Twinder".
- pick mutuals to chat with
- they DM you if there's a match
http://twinder.me/
https://twitter.com/MichaelTrazzi/status/1438533888540766210
4. The write up doesn't convince me. I think intergenerational transmission of general medical hesitancy is possible, even if the original source is forgotten.
That said, I do think Tuskegee is a convenient way to rationalize different reactions to the same behavior by blacks and by deplorables.
I think the real winner from the Violent Language list is "picnic". Apparently some terrible people used to picnic during horrible acts of racial violence, so now we should all say "outdoor eating". I guess we're lucky the rule isn't that decent people should no longer eat outdoors, which would be an equally valid conclusion to draw from the example.
re: 34
(I'm not trying to make a point or an argument, I'm genuinely curious here)
Can someone explain to me how this can be reconciled with the huge disparities in IQ over time in countries like Ireland and Greece as they became wealthier.
Also, if this data is meaningful, in what mechanism can the Flynn effect be impacting IQ if its not measurable as environmental?
I believe it should be "Sitzkrieg" and not "Sitzkreig". Except if there a second-level pun going on (with "Sitzkreis"), which seems unlikely to me.
On the AI media classification - it almost looks like the left-right political axis got mapped into a lower left-upper right axis in that graph. WSJ is still misplaced, but everything else looks more plausible.
On the IQ thing - do adoption agencies do a really good job of screening out adoptive families that live in places with high soil lead, or remnants of lead paint in the walls?
2. There was something of the sort in Spinrad's _Bug Jack Baron_-- I think the conspiracy was killing black children for (the rejuvenating effects of?) their glands.
4. I believe without evidence that greater medical neglect and incompetence would be enough to explain black people tending not to trust the vaccines without needing to bring in older history. I'm not sure if there's been research.
For that matter, I don't know whether white people who've had bad experiences with the medical system are less likely to trust the vaccines.
Re: 33. My brother keeps talking to me about that greenhouse plain. It looks quite uncanny on google maps. Does anyone know what is being grown there?
As to why there is no Journal of Datasets, well, dataset papers exist and do get published in fields like machine learning and computer vision. In biology and medicine it seems less common.
BUT, the main reason is not a technicality of whether a journal exists or whether you get a line for your CV. The publishability is a red herring and not the ultimate reason.
The issue is that scientists genuinely give you more "informal reputation" for coming up with interesting results, hypotheses and solutions to problems. Collecting data is boring grunt work that isn't really all that impressive from a subjective point of view. There are exceptions like the ImageNet dataset with did help Fei-Fei Li become really famous and successful, but that's computer vision again.
It's similar with all the complaints about publication bias, the file drawer effect etc. and the idea that a Journal of Negative Results will improve upon this state of affairs. However, the issue is not with whether there is a journal or not. The issue is that experts and scientists will learn your name and help your career if you find some interesting result that makes them go wow. Just because there's a journal for negative results won't mean that it actually gets you true recognition. And in academia informal connections are everything.
Under point 15, the following is a poor conclusion.
"Plausibly SpaceX will still have to follow the rules of its host country, the US, so the big loser here will be China "
China will not be a loser. Tesla has significant assets within China, with the CCCP not distinguishing corporate fictions of Tesla and SpaceX. Additionally, China demonstrated anti-satellite capabilities in 2007, which lends itself to slightly more force more than a simply fist shake. https://en.wikipedia.org/wiki/2007_Chinese_anti-satellite_missile_test
Starlink as anti-censorship technology will only work in peripheral nations.
There are journals for datasets (e.g. Journal of Open Psychology Data), but they are not used much in most fields for two reasons: 1) Scientific culture of valuing exciting (often nonreplicable) discoveries over high-quality data; 2) New-ish journals are almost never popular because it takes a long time to get an impact factor and overturn the inertial prestige held by the older journals, thus almost no one submits to them.
The article linked about the Tuskegee issue was frustrating to read, because it never seemed to attempt to explain why, if not Tuskegee, would Black Americans be more *hesitant* to get a COVID vaccine. There was a lot of stuff in there about unequal access and the problems that Black Americans who *want* to be vaccinated face in achieving that goal, but that's a different issue. I suppose what the article could be implying is that if you've been dealing your whole life with a system that refuses or hesitates to treat you when you want to be treated, when it starts beating down your door begging you to let them treat you for something else, you get suspicious. But I would have liked to have seen some quotes or data that actually indicated that.
It could also be true that the article wasn't ever really trying to answer the question "why don't Black people trust the COVID vaccine", but instead was saying "if you're looking at overall vaccination rates by race and noticing that Black people have lower rates than Whites, don't conclude that it's because they don't want to be vaccinated because Tuskegee", but again, it never said that explicitly. Also, I got a chuckle when it described "religious reasons" as a "modern" objection to vaccines, and once again I think that certain White people are continually confused when Black people have conservative opinions/values even though they are the most reliable Democrat voting block.
RE:Geothermal - I can say that there isn’t any real fear of wind and solar within the oil and gas industry. But geothermal elicits a “yea that’s a thing that could actually work”. Part of that might be that building geothermal capacity looks mechanically similar to E&P (exploration and production) because they both involve sticking things into the earth.
Ultimately, geothermal will run into the same problem fasting does - there’s not really a product anyone can get rich making so the proponents aren’t all that financially incentivized.
"Semiconductor manufacturers have 'about as much built-in voodoo and superstition as a major-league baseball team'"
I will observe that this book was written in 1985. About the semiconductor industry of the early 1980s. Fab process control has gotten a lot better since then.
The Tsimane' aren't hunter-gatherers. They are mostly horticulturalists.
28. HMC and marriage.
I can't see the full article, but a key component of this idea relates to the female's use of chemical contraception pills which skew female preference.
The idea being that she'll choose overly similar males as company when she's on the pill with the idea of female spending more time around brothers and fathers and uncles when they are already pregnant. While females who are not on the pill will be more attracted to men who are different to them.
There are many such stories linking this to divorce rates and many stories of women who go off the pill to get pregnant with their partners, only to suddenly become repulsed by their smell and unwilling to have sex with their partner.
Nearly every woman I've talked to about this will know women in their social circle to whom this has happened. Making it clear to me it is not a rare event.
The contraception pill works to make the woman's body act as though it were pregnant. Leading to 'incorrect' selection of males who are similar to them as mates.
So I'd postulate what we may well be seeing as 'random' is in fact a mix of women not taking the pill selecting men with different immune system and women taking the pill selecting men with similar immune systems...leading to no 'overall trend'.
#2: Aldous Huxley mentions Adrenochrome in the fourth paragraph of The Doors of Perception, which was very popular (It's still the #1 search result for books about psychedelics). That's probably where most of those post-1954 fiction writers heard about it.
Adrenochrome is not an imaginary drug, although most of its uses are. The headline of the Daily Beast article is incorrect, and so is the implication that Hunter S Thomson made it up 20 years after Huxley.
Here is a link to the full pdf of The Doors of Perception, where adrenochrome is mentioned at the top of page 3: https://maps.org/images/pdf/books/HuxleyA1954TheDoorsOfPerception.pdf
Huxley implies adrenochrome is a powerful endogenous psychedelic and explains the then-current but now-disproven theory that adrenochrome contributes to schizophrenia.
See also this paper on adrenochrome inhibiting ATP production and probably not being very safe:
https://pubmed.ncbi.nlm.nih.gov/3635424/
"Perfusion of the heart with 50 mg/L adrenochrome induced a marked decline in contractile force within 5 min and this was associated with a rapid decline in the myocardial ATP/AMP ratio...Adrenochrome at concentrations of 20 mg/L or higher was found to inhibit the oxidative phosphorylation activities of heart mitochondria under in vitro conditions. "
3 is complete bullshit. It should read: "While the Aztecs were sacrificing prisoners to the gods, their neighbors in Tlaxcala were a confederation of four despotic kingdoms that also sacrificed prisoners to the gods, because Aztec culture and Tlaxcalan culture were almost identical."
27 (real estate) seems like further evidence for my "contracts make us dumb" hypothesis. The more long-term contracts people sign, the more that locks us into doing things that no longer make sense if you're looking at objective reality, but seemed like a good idea at the time the contract was signed.
30. There are some interesting ideas in this article, but the Voxsplainer massively oversells its promise - and it is not internally consistent.
The total energy flow out of the earth is about 30 TW. Total human energy use today is 15 TW.
Geothermal energy is not uniformly distributed, but you can use that assumption for some rough estimates. About 2/3 - 3/4 of this heat comes up in the oceans (the Earth's surface is 71% ocean). Land-based geothermal cannot cover total human energy use. The US covers about 1/50th of Earth's surface, so even if we are able to harvest all of the geothermal energy, it would be about 0.6 TW. The US uses about 3 TW.
That doesn't stop Vox from claiming "5,157 gigawatts of electric capacity" (i.e. 5.2 TW) from geothermal in the US is possible. For this to even be physically possible, the US would have to be 10 times more volcanic than usual - which it's not.
Current total geothermal energy production globally is about 15 GW, so 0.1% of global energy use. [I wonder if Vox assumes that people don't keep track of TW vs GW vs MW.] This is concentrated around 'hot spots' - unusually volcanically active areas. New technologies could increase this significantly.
The technology Vox describes is basically using fracking technology (pumping pressurized water through rock) to access more geothermal energy.
This would start as a way to expand the area around a hot spot where geothermal is viable. After the technology is perfected there, the promise is that it could be used "almost anywhere". This is not quite right - it should say "almost everywhere". If we want geothermal to compete at close to the scale of fossil fuels, we need to tap a significant fraction of all the heat flowing up through Earth's crust.
Vox promises that this is not fracking and will not cause any seismic problems, as long as it stays away from seismically active areas. This is inconsistent - I'm not sure where they're planning on finding volcanically active, but not seismically active areas.
There is an even bigger problem for "almost everywhere" geothermal. Rock conducts heat really poorly, unless it contains fluids - either water of magma. There is magma near hot spots, and water in aquifers close to the surface. Small amounts of water can be pumped down, but not enough to saturate all the way down to the mantle. You will extract heat from the rock that is in contact with the pumped water, but that rock will reheat from lower rocks extremely slowly. This should be thought of as mining the heat of the crust. It is not renewable on the timescales it can be extracted.
Using technology developed for fracking to help geothermal is a good idea. I think it's plausible that it could increase geothermal energy production by a factor of 2-10. But claiming that it could compete at the scale of fossil fuels or that it "solves energy" is completely ridiculous.
Further context on #37: "phase of the moon" is a classic idiom in tech for inputs to a process that are fundamentally mysterious or random-seeming. (http://www.catb.org/jargon/html/P/phase-of-the-moon.html) Naturally the link includes two more examples of issues that were actually linked to the phase of the moon.
Re #17. In medicine, data are often collected under very a strict protocol to preserve privacy (see HIPPA). This makes it challenging to post datasets publicly. However, there are at least a few higher tier journals, such as PLoSOne, which insist that data be made public unless authors can provide good reason not to. The journals typically suggest one of several repositories. I've posted numerous datasets to the Inter-University Consortium for Political and Social Research (ICPSR) repository for papers I've published in PLoSOne, JAMA, JAMA Open, and BMJ Open.
Typo on #35: difference -> different
"Students trying to prove the hypothesis were less than half as likely to notice the gorilla than those investigating aimlessly. I’m agnostic as to whether this proves something important about science or is just funny."
They're students doing homework. They tested what they were asked to test and then went on with their lives. Doesn't tell you anything about real scientists.
> In 1922, someone wore a straw hat after official stop-wearing-straw-hats date September 15, leading to the week-long Straw Hat Riot in New York and several hospitalizations.
After which the culprits were punished by magistrate Peter Hatting. Very appropriate.
Well, I just filled out the form for the arranged marriage. It asked me for my age, income, a recent picture, and gave a miniature Big 5 personality quiz. Wish me luck, I guess.
On the subject of high-context memetics: https://twitter.com/lukechampine/status/1367547150494752770
"Not sure if this counts as “the left is eating itself”, but the (unofficial) Brandeis Suggested Language List now recommends students not use the term “trigger warning” because it is “violent language”."
Well, you're linking to the Prevention, Advocacy & Resource Center website of a private university, so I don't really see how that would count as "left" politics. Can you explain to me your reasoning here, Scott Alexander?
Re: 17 That explanation doesn't make sense. Data sets are still published material. If someone publishes a derivative of your work using your dataset without crediting you on the paper (IMO is a co-author is a minimum professional courtesy) they are clearly committing academic fraud of some sort.
Re: Media Bias, aka #19, what the heck is that chart? The colors and bubble size is just based on their own pre-assigned metric? How is that the choice you make, when the color and bubble size are the most visually striking parts of the series. The color shade should be how biased they really are in your analysis, and the bubble size should be some sort of viewership/readership metric.
Then you can position them just like you would on a language relationships chart. Such as: https://matadornetwork.com/abroad/mapped-crazy-relationships-among-european-languauges/
Jeez, that's a pretty devastating evisceration of the WEIRD book.
What is PCM?
From the Wikipedia link on straw hat riot: "The riot itself began on September 13, 1922, two days before the supposed unspoken date, when a group of youths decided to get an early jump on the tradition." So the description "someone wore a straw hat after official stop-wearing-straw-hats date September 15, leading to the week-long Straw Hat Riot in New York " doesn't seem quite right.
3. Tlaxcala is fascinating. There was a interesting piece a while back about how Cortes is seen as a villainous figure in most of Mexico - with the exception of Tlaxcala.
It also points to how the typical way we're taught about the Spanish conquest is misleading. Cortes didn't just show up with 500 men and roll over an empire - he allied with the empire's enemies, and led a coalition that toppled said empire. To a lesser extent, something like that happened with Pizarro and the Incas - a lot of his local allies were folks who had been on the losing end of a civil war that happened after the Emperor and his heir died from disease.
15. I wouldn't go that far with Starlink, but SpaceX genuinely is a treasure, and if they can get Starship working at the price-points they're aiming for it could have a revolutionary impact on space industry and utilization. Imagine affordable suborbital tourist flights, much cheaper access to space, and so forth. It's great stuff - I just want them to start making their fuel from green hydrogen and air-captured CO2 sooner so it won't produce so much emissions.
30. I feel like they should go for broke on geothermal, see if increased funding and investment (plus some actual contracts) can get super-hot-rock geothermal to work. Maybe you could sell it to Congress as a way to reuse some of the infrastructure from derelict coal plants, since both of them would need to generate steam to provide electrical power.
One fascinating thing you could do if you had ample geothermal power plus plentiful fresh water (or we got better at desalinating seawater) if use it to produce e-fuels for export. That's not economical yet compared to regular fuels, but if you have to phase out regular fuels and really want to get some more mileage out of your gas and oil power systems you could do it (although I think in practice such will mostly be aimed at ships and airplanes as the market).
This all reminds me of Eli Dourado pointing out that if we really didn't care about Old Faithful or Yellowstone, we could turn the whole Yellowstone area into the Mother of All Geothermal Power Plants and provide electricity for half the country.
> By the way, this is the same Tegmark who in 1998 developed a leading theory for what the universe is and why it exists at all.
Leading with whom? All I can find is people damning it with faint praise. The "reception" section on the linked Wikipedia page is one guy calling it "provocative." The physicist quotes that he uses to advertise his book on Amazon seem lukewarm; every physicist writing a review is carefully mincing their words so as to say something positive while distancing themselves from even the perception they like the theory. "Whether you believe it or not..." "You don't have to necessarily agree...", "You may recoil from his thesis, but..."
I'm interested to hear if it made an actual splash with serious scholars in either physical or philosophical fields. My impression is that it didn't; it made a splash in the science pages of some major outlets during his book tour (for what is definitely an engaging book, even if it's trying to motivate an ill-defined theory!) and then everyone more or less moved on, because there wasn't all that much *there*.
The reason it didn't make more of a splash is that the theory is far been more sensational than substantive. It's not bad physics, because it's not physics at all - it's bad philosophy instead. It adds a modern aesthetic to mathematical monism, but doesn't bring any really new ideas to the table. Every mathematically possible universe exists and we're in one, and they *are* math rather than being *described by* math? "Worse Leibniz" pretty much captures it; the idea was better put, and more completely fleshed out, in the frickin' Monadology.
The most frustrating part, to me, is his signature argument in favor - that Occam's Razor prefers it, since the theory has "no free variables" and no perception can argue against it, since it predicts all of those perceptions. But Occam's Razor cuts the opposite way: a theory is good if it doesn't *have to* propose gigantic numbers of entities, like vast arrays of possible worlds or gremlins behind the scenes, in order to account for observations. The original phrasing is literally (translated directly) "entities should not be multiplied beyond necessity." I have no idea how the heck Tegmark thought *that principle* could support a view which basically says, "multiply out every permutationally possible mathematical entity, they all exist." (I know several more refined versions of the Razor have been proposed - but even the best one won't support his argument in the way he needs it to. A mathematical ontology is not, at its core, any simpler than any other.)
#2 Mentions Dune, and adrenochrome harvested from children seems like something Baron Harkonnen would use, but the article actually references "Destination: Void, from a different book series by Frank Herbert.
Re: 19, has anyone here _ever_ had a semantically understandable category fall out of embeddings without feeling like you're trying to force it? I've done a ton of work in that area and whenever I try and understand what the program is doing in real life terms it just seems like nonsesnse.
It looks like (4) studies the link between knowing Tuskegee and refusing treatment in other diseases before 2020, and also finds there isn’t a difference in actually getting treatment. While there is such difference in Covid vaccines. And in the interviews they note Tuskegee is occasionally mentioned, but not as much as others, but I’m not sure those other reasons apply to COVID vax, as they make more sense in a context of active treatment
James Hawes's The Shortest History of Germany is a (biased) book about how "East Elbia" was always a different thing than the Rhine zone.
On Tuskegee, this article (https://academic.oup.com/qje/article-abstract/133/1/407/4060075) published in a top econ journal finds disclosure of Tuskegee in 1972 leads to "increases in medical mistrust and mortality and decreases in both outpatient and inpatient physician interactions for older black men" by comparing changes in the health of older black men closer or farther away from Macon County (county of Tuskegee). It uses difference-in-differences-in-differences and D-i-D style techniques have come under increased scrutiny in recent years. With that caveat, if there was an effect on black adults in 1972, that would likely persist to today through social norms in the black community on medical distrust and going to the doctor even if black people today do not know about Tuskegee. Though someone should look at re-doing this paper's strategy with vaccine hesitancy as the outcome and better D-i-D methods.
> I can’t tell whether “two components” naturally fell out of the data, or they decided it by fiat
with PCA it gives you a list of component basis vectors sorted by contribution / importance, I imagine they just took top 2 like you usually do
Regarding #25. The paper says, "odds ratio = 4.8, P = 0.034, N = 33, Fisher’s exact test". Don't use it. Fisher's exact test has low power.
Lydersen, S., Fagerland, M. W., & Laake, P. (2009). Recommended tests for association in 2×2 tables. Statistics in Medicine, 28 (7), 1159–1175. http://doi.org/10.1002/sim.3531
(This comment is written by someone who would probably not have seen the gorilla)
On straw hats, doesn’t that link say the riots started two days before straw hat day? That feels very human, ‘No! You can smash up my hat if I was wearing it on the wrong day, but it isn’t the wrong day yet, you bastard, let’s fight!’
Leading theory my ass. (Metaphysical speculation) != (physics). We have guild rules about that.
@37: My favourite sentence in the Wikipedia article for Alchemy is in the section about its transition into chemistry:
"Boyle would note the place in which the experiment was carried out, the wind characteristics, the position of the Sun and Moon, and the barometer reading, all just in case they proved to be relevant."
Up until now I thought it's both cute and inspiring, a combination of a scientist's discipline and a 17th century intuition about the world. Well, the passage about semiconductor yields made me realize I was unfair to Boyle, and that considering the positions of celestial bodies makes sense even today.
I once attended a scientific meeting where one group had gathered a large amount of brain scan data and urged us to poke around in it and see what we could find. Others at the meeting urged us to not explore the data but only test specific hypotheses on it, because if we looked we would be tainted and would only test hypotheses on the data we already knew the data would support. I was inclined to agree with the first group, but I didn't know whether that's just because that's my style and I'm not a careful enough scientist. Anyway, that's what the gorilla made me think of.
That Brandeis list veers between "okay, I see the point here" and "if ever you hear or see me use this term, shoot me and bury me in a ditch":
"Gender exclusive language:
You guys, Ladies and Gentlemen
Gender inclusive language:
Y'all, folks or folx, friends, loved ones, people, everyone
These examples either lump all people under masculine language or within the gender binary (man or woman), which doesn’t include everyone."
I wonder where the use of "lads" (which we used to refer to each other in the collective sense in my all-girl school) falls on this? And if ever I use "folx", you may be sure and certain that that last remaining dim flicker of reason and sanity has burned itself out and I am now a drooling imbecile.
Oh, I'm sorry - is "imbecile" an ableist term?
@19: If it's any consolation, the media bias study seems to be a small side-project of his main thing right now, which is creating an artificial physicist - a symbolic regression algorithm where experimental data comes in and equations come out.
"Study shows Donald Trump is grumpy when he does not get enough sleep. Being an important enough figure that scientists study you personally sounds exhausting."
I await with breathless anticipation the follow-up studies "When caught out in the rain without an umbrella, Donald Trump gets wet" and "If he has not yet eaten that day, Donald Trump gets hungry".
But how can we be sure this result replicates unless we get data from figures on the other side of the political divide? Has anyone studied whether or not Joe Biden gets more cheerful the less sleep he gets?
Are people really that hard-up enough for something to publish that they need this kind of "grass is green" obviousness, or was it a joke? If it was a sincerely meant paper, I think some "kick them out of the lab into the sunlight and fresh air" is badly needed.
7: Although the thing about "trigger warning" is in the section on "violent language", the actual reasons given for maybe preferring "content note" over "trigger warning" aren't about the implication of violence at all. (I think there's at least one other good reason for usually using something other than "trigger warning": there are reasons why people might want advance notice of potentially-upsetting things besides outright PTSD-like risk of being "triggered".)
>Imagine a future in which Earth’s dominant ISP is constructed, maintained, administered, and governed by Martians, who operate it for free as a public-relations campaign.
Please, can we colonise a saner place like the Moon, the asteroid belt or the Venerean atmosphere? Mars is one of the worse choices because it's got the downsides of a planet (need chemical rockets to take off due to gravity/atmosphere, extra delta-V from gravity well) without the upside (atmosphere thick enough that you don't need a pressure hull or radiation shielding).
Re #16: This chart has been bugging me since it came out. It seems too cute by half. Does the vaccine really reduce mortality by the same amount across age groups?
Semaglutide: "Yet GLP-1 had serious problems. The first was that it was degraded in the blood by an enzyme within a few minutes. This meant that it wasn’t useful as a drug in its native form because it would have to be injected far too often. This problem was solved by using another version of the hormone identified in the saliva of the gila monster lizard..."
I feel as though there is an untold story behind that last sentence.
#17 Could you segment or subset the data. I really think we should stop incentivizing papers based on unpublished data, much of it gathered with public funding. The replication problem is crippling public confidence in much bio-medical and social science research.
#21. hatis problems are historically deep. They can be over explained based on its history. E.g. the French forced the independent country to pay them an indemnity in the 19th Century. The Spanish did nothing like that to the DR.
#16 Not sure of the relevance, since it seems to me that most parents ordinarily want to expose their kids to massively less risk than themselves in general.
Re: War Against The Weak, even academic historians (very anti-eugenics, and very likely to classify things as eugenics) found Edwin Black hyperbolic and sensationalist, especially when it came to drawing Nazi connnections.
Also re: sterilising the blind, separating views on the desirability of someone passing on their genes from views on their capabilities as a parent is difficult, and historians tend to ignore the latter despite all the evidence that it was a strong factor in what is categorised as "eugenics".
#31 My daughter was prescribed semiglutide under its Type II diabetes label of Ozempic., but it was primarily for weight loss. She lost 100 lbs. She perceived the mode of action to be causing feelings of repletion to occur upon the consumption of very small amounts of food and for the feeling to last much longer than usual, e.g. 6-8 hrs instead of 4-6 hrs.
Thus encouraged, once the Wegovy formulation was FDA approved, I got my doctor to prescribe it. I have taken the first two 0.25 mg dose. The ramp up to a full therapeutic dose is 16 weeks (02.5, 0.5, 1.0, 1.7, 2.4 mg). No effects or side effects yet.
Medicare Part D will not cover the medicine, or any other weight loss medicine either. It is very expensive. $1325/month. if it works, it is worth it. if not, well weren't going to Italy this year anyway.
>By the way, this is the same Tegmark who in 1998 developed a leading theory for what the universe is and why it exists at all. I feel like going from “discover fundamental nature of the universe” to “attempt to investigate media bias, but it has glaring flaws” is a slightly-too-on-the-nose metaphor for the past 25 years of science as a whole.
This is also the same Tegmark who is President of the Foundational Questions Institue, science director of the Future of Life Institute, and has been a major figure in the advocacy for AI Safety (he wrote Life 3.0).
I'm surprised Scott ommits this. This paper on AI bias seems to be almost a kind of passtime. That's the "metaphor for the past 25 years of science as a whole" - there are many accomplishments from remarkable polymaths we don't even know about - though you could just link his Wikipedia or MIT page.
"Now a Haiti econ blog (appropriately named Vodou Economics) proposes a much better explanation - the data are wrong, and the DR was always ahead of Haiti."
It's not so much that the data are "wrong" so much as they are for the 2011 ICP rather than the 2017 one. The 2017 one shows a much larger gap between the Dominican Republic and Haiti (tenfold as of 2017). The growth statistics are the same, which is arguably what should matter more here. The Maddison people (who, it should be said, are more often than not total morons) discarded some Braithwaite (1968) PPPs for being unreliable -but that shouldn't matter here.
Re: #21 Any one visiting Haiti & the Dominican Republic would have a hard time believing they are on the same island. Despite having similar histories (former colony, revolution, foreign occupation, dictatorship, eventual democracy) the two nations could not be more different.
Having direct considerable time in both I've found that the contrast is the result of what each people see as their heritage.
Dominicans see themselves as “European” with social/political/economic/cultural values and ambitions inherited from Spain.
Haitians see themselves as "African” with social/political/economic/cultural values and ambitions inherited from West Africa.
19: Looking for words or phrases with high predictive power with regards to politics actually merely measures political jargon, not bias, although there is obviously a link to bias, as bias is encoded in some of the jargon (or how it is used).
33: You can take somewhat similar pictures in the Westland region of Holland: http://i.imgur.com/8PNGO.jpg
Of course, it much less of a weird mess. Why do those Spanish greenhouses have these weird shapes? No reallocation of farming land (Dutch farmers traded a lot of land in the past to get nice rectangular plots)?
35: Of course, Germany has states with rather strong cultural differences between them. For example, Prussia was famous for it's military prowess. Prussia was dissolved and many ethnic Germans migrated to other parts of Germany. But it's perfectly plausible for Mecklenburg-Western Pomerania, Brandenburg and Saxony to have a different culture to the other states.
I disagree with the commenters saying that datasets don’t make good papers in biology. Well-curated, accessible datasets with some analysis make great papers, get published in leading journals, and get cited hundreds of times by people re-analyzing the data. Based on my personal experiences I firmly believe the real reason researchers are bad at data sharing is because they’re afraid that other researchers will re-analyze the data and find their mistakes (or their fraud).
Devon's piece on Prospera is excellent, though every time I see documents by the National Lawyers Guild (a critique of the ZEDEs linked under "Other Resources" as a legitimate recognition of more critical perspectives), I recall their 2003 report on North Korea, which to the best of my knowledge they have not repudiated. This is from a summary by Christopher Black, one of the delegates, still seeing the great revolutionary dream in DPRK in 2003,
"In 2003 I had, along with some American lawyers, members of the National Lawyers Guild, the good fortune to be able to travel to North Korea, that is the Democratic Peoples Republic of Korea, in order to experience first hand that nation, its socialist system and its people. The joint report issued on our return was titled “The Grand Deception Revealed.” That title was chosen because we discovered that the negative western propaganda myth about North Korea is a grand deception designed to blind the peoples of the world to the accomplishments of the Korean people in the north who have successfully created their own circumstances, their own independent socio-economic system, based on socialist principles, free of the domination of the western powers."
https://journal-neo.org/2017/03/13/north-korea-the-grand-deception-revealed/
A sample from the report itself,
"As in Cuba and other one party socialist societies, North Korea has a system of direct democracy in which elections are held for local peoples committees, district and provincial committees and to the Supreme People’s Assembly. The absence of other parties is not considered a failing, as the entire society is socialist. The question of multiple parties did not even seem understandable to those we spoke to. The delegation questioned whether within that system, there is in fact more participatory democracy than in the American federal system or the parliamentary system in which democracy ceases to operate once the elections are over. It is more circular, with local committees sending up to the next level requests, complaints and so on and so on up to the national level with discussion, at least in theory at these levels and then feedback to the local level until an agreement is reached based on resources available and circumstances.
However, the issue is not whether we agree with DPRK’s system or feel that our democracy is better or more just. . . . Certainly we cannot say that only one political system is successful and generates a participatory and healthy society. We hope that future delegations can learn more about political dialogue within the DPRK system and share the pros and cons of our system without blame or judgment."
http://www.soviet-empire.com/ussr/viewtopic.php?t=43901
So ZEDEs are suspiciously undemocratic from the National Lawyers Guild perspective but the "direct democracy" of the DPRK is just fine?
Aetna requires prior authorization for Wegovy, the brand name for semaglutide. I tried setting up a Teladoc appointment to get a prescription, and the doctor said she’s not allowed to prescribe “lifestyle medications” and anyway semaglutide is for diabetes.
I don’t interact with the medical industry enough to know if this just typical friction or not, but so far I’m inclined to write this medicine off as vaporware for now.
37: This doesn't surprise me at all. I used to work for a chemical company that built a plant in South Korea, meant to be a copy of a North American plant. Same blueprints and everything, but for at least 6-8 years the chemists and engineers, from Korea or the US, couldn't get it to produce the same purities as the US plant. So far as I know they never figured it out. Apparently somewhat random outcomes like that are fairly common with large scale chemical production, so apparently there are many strange little influences on the process.
Are you telling me that the under-age-15 hat-smashing boys were "spanked ignominiously" and just sent home? No criminal record to prevent future employment? No psych eval and prescription medication? No movement to a stricter alternative school?
I don't think they probably learned any lesson if the authorities didn't wreck their subsequent decades.
As funny as the straw hat story is, the summary is inaccurate :)
re 14, the roots of progress:
It is interesting to me that they are trying to establish this *new* philosophy of human development on the basis of capitalism. They have patrick collison and peter thiel among others on board as, i guess, "thinkers" and patrons, to define progress in their vision. It's no doubt these people are heavily biased towards start up esque systems and will not want to build a future without them, at least, but it's more than likely they will be building that future on the very basis of them. I wonder if they really believe this will turn out different than the 20th century, if they really believe half the problems of our age were not in fact caused by those very systems. In my view skepticism, bureaucracy, and stagnation, is an inescapable consequence of a mature capitalist system (though a young system doesn't have those problems as much). I don't know why they're trying to build the future in this image but it's not going to work.
On insulin and weightloss. I have recently been prescribed metformin for whatever "pre-diabetes" is properly called.
What I have noticed is that I feel warmer - I don't notice being cold or being in a draft as much.
Hypothesis that it may be raising either core body temperature, or (more likely) peripheral circulation, which would make my skin, fingers, etc warmer. That warmth would actually mean that I have more heat loss. That would mean that I'm burning more calories to maintain core temperature - which would be an obvious weightloss mechanism, and I know there is some (mixed) evidence for metformin for weightloss, and also explain why it is mixed - if it works through making it easier to tolerate lower temperatures, then you only get the weightloss if you actually experience lower temperatures.
Anyway, proposed approach: monitor room temperatures for people taking possible weightloss drugs, and see if a drop in room temperature correlates with more weightloss.
Unsure whether it's right to frame East-West migration in Germany before the Berlin wall as "fleeing". I realize the term comes from the abstract of the article (which I have skimmed over and have not yet made up my mind about) so this is not on you.
#36 I agree with much of the criticsim of Henrich's WEIRD argument. I have made similar comments in other fora.
The short version is that Henrich identifies canons of the Medieval Catholic church forbidding cousin marriage as being a key development in the foundation capitalism in Western Europe.
My response is that those rules no longer impacted English society after the Reformation. England is indisputably the home of the industrial revolution. But cousin marriage was common. E.g. Charles Darwin.
It is a key plot point in Jane Austen's Mansfield Park, but is no where mentioned as an objection to the marriage of Edmund and Fanny. In Trollope's Can You Forgive Her, Alice Vavasor is infatuated with her cousin George. The objection to George is not that he is her cousin, it is that he is a jack@$$.
3.) They're somewhat exaggerating the differences in government. Tlaxcala had a monarch and the Aztecs had noble and common political institutions, at least if popular memoriy is to be believed. More interesting is the geopolitical situation. Tlaxcala would be like if an even more brutal version of Rome had risen to power except a state based next door to Rome itself had managed to stay independent and stave off conquest by extreme militarization, long term acceptance of economic deprivation, and alliances with dissident factions in the Aztec empire.
This is a huge context to Cortez's conquest that gets missed. This is why Monctezuma didn't just immediately try and kill Cortez. There was a wider political situation he had to defuse or face a larger revolt/invasion. It's also why driving Cortez out of the city didn't work: Cortez ran back to Tlaxcala and Monctezuma couldn't pursue him without getting past the giant Tlaxcalan-Rebel army. This gave Cortez the time to regroup and engage in more open warfare. (Also, technically Tlaxcala wasn't conquered: they were direct vassals of the Spanish Crown. It was only on independence they were incorporated into Mexico.)
4.) This is overly literal. I have no doubt it's rare. But certain communities have developed intergenerational distrust of the government, often with good reason. This is just a particularly emblematic example that's come to the attention of the chattering classes.
12.) It's worth nothing that the Constitution was a barrier to this kind of thing. Yes, the Supreme Court eventually allowed it. But it took decades to get there, decades that were used in other places to do it even more widely. The most thorough eugenics experiments took place in the territories where the Constitution didn't apply. Likewise, it was often worse in places with less robust conceptions of rights. Rights may forestall good and necessary changes. But they also forestall a lot of evil ones.
15.) It'll probably be the same as how radio and television already work. Lots of little pirate signals in places with significant censorship. They'll be tracked down by the regime in erratic but furious enforcement. In places with freedom of speech the big, legal networks will dominate because they can outspend/out-quality little pirate operations. More importantly, I expect us to get Radio Free Europe: Internet Edition at some point. If we can turn on the uncensored, untraceable internet in North Korea or Manchuria that would undoubtedly be to our advantage.
I'm also curious to see how it affects brands of authoritarianism like Putin. Putin doesn't really do direct censorship: you can go and Google anything you want about him. He even has real elections. He just slants everything as far to his side as he can. Government regulators lean on broadcasters and other organizations. Fraudulent prosecutions go after the wealthy who oppose him. Opposition journalists are attacked. Etc. This might make his form of authoritarianism more durable than the old totalitarian kind. Giving Russians free access to the internet wouldn't change as much as it would in China.
35.) Not to mention the the different regional effects of Nazism. East Berlin/Germany had more Poles, Catholics, and Jews, all of whom didn't fare too well under the regime. Or that being occupied by the Soviets and being occupied by the Americans is a rather different experience. Berlin was conquered by the Soviets wholesale but Americans took their sector over on July 4th 1945. The period from 1945 to 1949 was really turbulent and critical. I think it's very fair to see west vs east Germany as a comparison of two economic systems, especially when comparing specific industries. Food distribution, for example. It's less fair to blindly attribute all differences.
For example, I've long pointed out that the East German female workforce participation rate was higher than the west's in Imperial Germany. So while Communism no doubt did something to get women into the workforce the differential was already somewhat there. Conversely, west Germany has been richer than the east... basically forever. So while it's fair to point out Communism almost certainly did have downward pressure effects simply pointing to raw GDP is too simplistic.
7. If only it were satire.
36. I see three hyperlinks to the one Amazon review. Wondering if you meant to link others?
Fwiw, this PolicyTensor critique is also worth the read.
https://policytensor.substack.com/p/the-churchs-crusade-against-cousin
Me: Astrology is dumb and annoying.
Also me: Fuck yeah my compass coordinate is on Apollo! Apollo gang!
At this stage it's straight up unscientific to be unwilling to even consider the possibility that biology explains much of the difference between Haiti and DR. The difference in outcomes between these countries is only surprising if you completely ignore the outcome differences between every sub-saharan african country and every other mestizo-majority country, and the socio-economic outcomes of these populations within white majority countries.
Re 20: "Once you remove elderly Hispanic self-identified Attack Helicopters the effect largely disappears" is one of the better sentences.
"34: It’s generally-believed that IQ is mostly-environmental early on in life, but mostly-genetic after adulthood."
What am I missing? I always thought that IQ was an attempt to measure native intelligence, not functional intelligence -- that is, not acquired abilities or knowledge. So it's hardly surprising that "genetics matter vastly more than environment and environment more-or-less doesn’t matter".
And because IQ is more or less impervious to environmental influences (except for prenatal and childhood malnourishment and diseases), the more interesting question to me is what educational practices are best at equipping the rising generation with knowledge and skills?
The trigger warning note sounded pretty interesting and I was going to share it around - except it turned out to be a reasonable note on the linked site. It says to use ‘content note’ instead which seems more professional frankly, and euphemisms like trigger warning - which seems obviously rooted in gun imagery - sounds pretty crass in a society with (real or apparent) endemic gun-problems.
>"IQ of biological parents explained about 42% of variance; IQ of adoptive parents explained only 1%. 'Variance explained' is a non-intuitive statistic with a lot of confusing properties - but in this case the obvious conclusion that genetics matter vastly more than environment and environment more-or-less doesn’t matter at all is basically right."
That "obvious conclusion" is the opposite of what the paper said. Look at the far-right column of table 3. Heritability explained 42% of variance in ICAR-16 , but environmental variation explained 50% of the variance in ICAR-16, *higher* than the heritability.
(The middle three columns account for the fact that ancestry and environment are correlated [me and my parents all grew up in Toronto etc], and gene-environment non-additivity [a gene might give me +1 IQ point if I live by a lake but +2 IQ points if I live by the ocean etc].)
I think #34 significantly understates the problems of variance compression in adoption studies. The most straightforward version of this problem is that adoptive families have low variance in IQ, with few very low-IQ families, as a consequence of the fact that adoption agencies are picky. Meanwhile adoptive families vary a lot, but with a strong skew toward low IQ. This means that you end up with low shares of total IQ variance explained by variance in adoptive parent IQ, but you can adjust for this by reweighting variances to reflect the variation in parental IQ in the full population.
The more serious version of this problem is that for parents who are on the margin of acceptability for adoption agencies, you'll have a negative correlation between IQ and other positive traits (like conscientiousness) even if the correlation between those traits is positive in the general population. If a low IQ parent has been accepted as an adoptive parent, they must have been able to check the boxes of middle-classness, stability, strong references, etc., in spite of their low IQ. The most likely reason for this is being above average for adoptive parents in non-IQ aspects of parent quality. On the other hand, a higher-IQ parent can get accepted as an adoptive parent even with some deficits (relative to most adoptive parents) in other qualities.
The same thing is true of biological parents--high-IQ biological parents who put their kids up for adoption are probably less conscientious, more prone to mental illness, etc, than are either other high-IQ people or even lower IQ biological parents who put their kids up for adoption.
But this is much more likely to bias estimated relationships with adoptive parents. The genetic causal mechanism is likely that high IQ genes lead to high IQ, even if paired with low conscientiousness genes. The environmental causal mechanism is probably that high IQ parents create better childhood environments, which leads to higher child IQ. The connection between conscientious parents and a good childhood environment is plausibly as strong or stronger than the connection between high-IQ parents and good childhood environment. As a result, a negative correlation between IQ and conscientiousness induced by selection into adoption is likely to negatively bias the correlation between adoptive parent IQ and child IQ substantially.
In today's "what the hell, medical science, make up your minds!" news, a new study says that exercise may cause heart attacks.
Or not. Conversely, it may demonstrate that the thing they've been measuring to indicate likelihood of heart attack was the wrong thing all along.
So all those stories about early promoters of jogging dropping dead could be explained by this, or not.
https://heart.bmj.com/content/early/2021/08/27/heartjnl-2021-319346
"Abstract
Background The association of physical activity with the development and progression of coronary artery calcium (CAC) scores has not been studied. This study aimed to evaluate the prospective association between physical activity and CAC scores in apparently healthy adults.
Methods Prospective cohort study of men and women free of overt cardiovascular disease who underwent comprehensive health screening examinations between 1 March 2011 and 31 December 2017. Baseline physical activity was measured using the International Physical Activity Questionnaire Short Form (IPAQ-SF) and categorised into three groups (inactive, moderately active and health-enhancing physically active (HEPA)). The primary outcome was the difference in the 5-year change in CAC scores by physical activity category at baseline.
...Conclusion We found a positive, graded association between physical activity and the prevalence and the progression of CAC, regardless of baseline CAC scores
Results
The mean (SD) age of study participants was 42.0 (6.1) years (table 1). The proportions of participants who were inactive, moderately active and HEPA were 46.8%, 38.0% and 15.2%, respectively. Participants with higher physical activity levels were older, less likely to be current smokers, and had lower levels of total cholesterol, LDL-cholesterol and triglycerides, higher levels of HDL-cholesterol, and higher prevalence of hypertension and presence of CAC than participants with lower physical activity levels.
Participants with CAC >0 at baseline were older, were more likely to be male and current smokers, and had higher levels of traditional cardiovascular risk factors.
The cardiovascular benefits of physical activity are unquestionable. ...Regular physical activity reduces the risk of many adverse health outcomes, including mortality, CVD, diabetes, hypertension, obesity and dyslipidaemia.
...High levels of physical activity, however, may be associated with a higher risk of coronary artery calcification. In a meta-analysis, higher physical activity levels were associated with a higher prevalence of CAC (pooled odds ratio (OR) 1.84; 95% CI 1.41 to 2.93)
...A possible mechanism underlying this association is that physical activity may increase coronary atherosclerosis. Potential pathways include mechanical stress and vessel wall injury of coronary arteries, physiological responses during exercise, such as increased blood pressure, increased parathyroid hormone levels and changes in coronary haemodynamics and inflammation. In addition, other factors, such as diet, vitamins and minerals, may change with physical activity.
The second possibility is that physical activity may increase CAC scores without increasing CVD risk. The standard Agatston CAC scores are calculated as a combination of calcium density and the volume of plaque burden. Higher calcium density, which suggests more stable, calcified plaque, produces a higher CAC score, however, it is associated with lower CVD risk.
...In general, progression of CAC scores is associated with a higher risk of CVD among individuals with CAC=0 and those with CAC >0.28 Moreover, the absence of detectable CAC is a strong negative predictor of CVD, whereas the presence of any CAC, even at very low levels, is associated with an increased risk of CVD. However, considering the undeniable protective effect of physical activity on CVD, the positive relationship between physical activity with CAC progression should be interpreted with caution as the complex interplay between physical activity, CAC progression and subsequent CVD risk remains largely unknown."
Conclusions seem to be "exercise is in general good for you but it could be bad for your heart. Or we could be measuring the wrong kind of plaque. Or we could be totally wrong about hardening of the arteries being a bad thing".
1. Hat Riot wikiepedia - "Many of those taken to court following arrests related to the hat-snatching frenzy opted to be fined rather than serve time in jail. The longest recorded time one of the teens was sent to jail was three days served by an A. Silverman, who was sentenced by Magistrate Peter A. *Hatting* during night court."
20. typo "the populations were already pretty *difference*"
I feel like there's a ton of useful truth here, but it would really benefit from re-interpretation and summary of the type Mr.Alexander is uniquely good at.
https://www.datapacrat.com/Opinion/Reciprocality/r1/index.html
The AI movie thing was fun, but it's pretty clear there were images of from some of the films or their posters in the training dataset. This is most obvious for die hard with the exploding building, and to a lesser extent space jam and point break.
Bioengineer here with a bit of perspective on the data sharing thing (17).
In a few sub-fields, this has been recognized as a problem and major journals require authors to put their data on big NIH-hosted databases. See https://www.ncbi.nlm.nih.gov/guide/genes-expression/ for examples related to gene expression and genetics.
At the same time, there's an undercurrent of discontent in certain fields of biology (I'm thinking mostly about systems biologists and developmental biologists) over the papers that are *just* datasets. We pretty regularly get papers that are, like, "A high-resolution cell atlas of a developing mouse hindbrain" or "comprehensive time-resolved whole-body single-cell gene expression of growing nematodes". Usually they're very impressive, technically, and produce a metric ton of data that can be accessed using a slick (but still kind of clunky) new custom visualization interface... and nobody knows what to do with it.
There's kind of an "okay, what did we learn?" moment that comes with a lot of those papers that's starting to be too familiar. It turns out that in the modern age, the hard part of learning about biology isn't collecting tons of data -- it's getting any useful conclusions out of that data. Like, imagine you have that high-resolution atlas of a developing mouse hindbrain. What do you *do* with that? It's easy to imagine lots of things you can check with that dataset *in a vague noncommittal way*, but actually concretely coming up with a specific hypothesis to check with that sort of thing is hard. So that's why putting giant datasets online isn't as prestigious as writing a paper that synthesizes something out of that dataset.
(I'm guilty of writing one of those papers, albeit a small one on a small subject. Back in my undergrad days I was on a paper that analyzed gene expression over the course of a bacteriophage infection. Collecting the gene expression data wasn't trivial, but in a way it was ridiculously straightforward. Figuring out what it meant was much, much harder, and I don't think we got all that much out of it.)
Incidentally, I don't have a good citation for this, but a while back I read that "science" (I think chemistry and biochemistry, more specifically) went through a similar crisis back in... I want to say the 50s and 60s? For a stretch, most of the "scientific research" that was going on was basically measuring parameters -- stuff like the rates of reactions -- to finer and finer precision. Eventually some big-name scientists got fed up with all the measurement papers and published some high-profile editorials to the effect of "science isn't about measuring numbers, it's about figuring out how stuff actually works!" and pretty quickly people stopped measuring parameters.
Personally, having written a bunch of models of biocircuits that rely heavily on the field's knowledge of underlying rate parameter values that the field definitely doesn't know, I think we could have done with a *bit* less of a course-correction on this. =P
The thing about the gorilla is totally consistent with cog sci literature on problem solving and the way specific goals constrain attention. For more, you can read my essay on cognitive load theory: https://cognitiveloadtheory.wordpress.com/
Why isn't there money to be made for hedge funds to go around and completely buy up the commercial mortgage securities and modify terms? Or for banks to offer different initial terms?
I feel something else must be going on, eg, maybe there are regulations for which book value matters more than actual expectations?
Since the comment thread where marxbro and I are having a fascinating discussion around cultural mores is becoming the length of the Equator, I'm going to drag this one up here so we can have a fresh starting point.
marxbro1917:
"I'm simply making the point that you are doing performative right-wing outrage. You say that you would dislike any politician doing it, but your own posts show that you don't really have any clear poltical analysis of why you like or dislike certain words. Likely you are just repeating what you heard elsewhere."
I, a parrot, repeating what I heard elsewhere and certainly not having any linguistic tastes of my own:
Squawk! 🦜
From "The Pilgrim's Regress", C.S. Lewis:
(T)he jailor addressed the prisoners and said: ‘You see he is trying to argue. Now tell me, someone, what is argument?’ There was a confused murmur.
‘Come, come,’ said the jailor. ‘You must know your catechisms by now. You, there’ (and he pointed to a prisoner little older than a boy whose name was Master Parrot), ‘what is argument?’
‘Argument,’ said Master Parrot, ‘is the attempted rationalisation of the arguer’s desires.’
...‘Good. Now just one more. What is the answer to an argument turning on the belief that two and two make four?’
‘The answer is, “You say that because you are a mathematician.”’
‘You are a very good boy,’ said the jailor. ‘And when I come back I shall bring you something nice. And now for you,’ he added, giving John a kick and opening the grating.
... ‘This psittacosis is a very obstinate disorder,’ said Reason. And she turned to mount the black horse.
https://www.washingtonpost.com/health/2021/09/23/covid-unvaccinated-but-willing/
There really are people who are up against logistical difficulties, need more information, and/or would like to talk with a human being about the logistics of getting vaccinated.
I was fascinated by your find on arranged marriages. I am really curious if they differ from how Indians do it. If that subject interests you, watch the Netflix show Indian Matchmaking. It is pretty real, I thought. Fascinating show.
https://youtu.be/zZ_dXjsyP5E
I've become skeptical about the WEIRD book's argument ever since I learned that traditional Hinduism was every bit as uncompromising in banning close-kin marriages as medieval Catholicism, and enforced this ban for a lot longer.
Marriages between descendants of the same 7th-generation ancestor on the father's side and between descendants of the same 5th-generation ancestor on the mother's side are banned as far back as the Gautama Dharmasutra (600 BCE-200 BCE). This is reaffirmed in all subsequent ancient and medieval Hindu legal texts (dharmashastras).
(Note that the famous Manusmriti only forbids marriage between third-generation relatives on the maternal side--but the traditional commentaries explain that the 5th-generation standard overrides it.)
The texts explain that this rule applies to all four varnas, unlike the better-known ban on marrying within one's gotra (paternal lineage), which applies only to the twice-born castes.
And, at least according to a random Indian guy I talked to on Twitter, this anti-endogamy rule is actually observed in practice, and makes finding a suitable marriage partner difficult.
On #17 publishing data, I think this is partially true, but doesn't explain why researchers don't publish their code, which does not entail foregone future papers. I think it's because they don't have good code habits, and mainly because this gets you neither citations nor respect (though respect might be trending upwards slowly).
"I link here in penance: Trump’s False Lafayette Square Exoneration. "
One would think you've learned a lesson after the last passage of the Woke ripping your nuts off..... but noooooooo. One. More. Time. Never explain. Never apologize.
Phoney War seems just incomprehensible to me. If this was fiction, holding an idiot ball would be an understatement.
> German military commander Alfred Jodl said that "if we did not collapse already in the year 1939 that was due only to the fact that during the Polish campaign, the approximately 110 French and British divisions in the West were held completely inactive against the 23 German divisions."
How do you... not attack an enemy when most of its army isn't present?
Also, this:
> Leopold Amery suggested to Kingsley Wood that the Black Forest be bombed with incendiaries to burn its ammunition dumps, Wood—the Secretary of State for Air—amazed the member of parliament by responding that the forest was "private property" and could not be bombed
WTF.
> now recommends students not use the term “trigger warning” because it is “violent language”.
Hopefully someone will soon decide the word "violent" constitutes "violent language".
The criticism of Henrich and WEIRD was valuable, but itself misses many, many tricks. The elements he introduces as disproofs sometimes have easy explanations. Still, it's always good to ask newly fashionable theories that claim to explain everything to explain at least something rigorously.
Re. 17, prominent bioinformatics journals require publishing datasets and code. Unfortunately, bioinformatics may have the biggest and most-restrictive datasets of any discipline. This is a big reason why I never published anything while I was working in bioformatics: my work often used terabytes of data; my employer wouldn't give me more than about 100 megabytes of ftp space; I couldn't afford to host terabytes of data myself; and that data would be gathered or computed from dozens of sources which all had different privacy and redistribution requirements.
On #28 -- The reddit post might be relevant to some smaller landlords, but it doesn't really reflect decision making from large institutional landlords. I'm an asset manager with a 3+ million square foot portfolio. With office space, part of the issue is that a landlord almost always has to give upfront concessions to a tenant and in the middle of a recession those look way too risky on long term leases. If you can only get $26/sf rent with 3% escalations per year on a ten year lease and have to give $80/sf in construction allowance to a tenant plus 10 months free rent upfront plus pay commissions to your broker and the tenant's broker... why not wait a year when rents and concessions go back to normal? Also, demand for office space doesn't just dip in a recession, it goes to almost zero and tenants put their space on the market too, so you compete against super cheap sub-lease space. (e.g. the market clearing rent doesn't go from $30 to $26 per square foot, it falls to like $8 per square foot). If you just wait a little while you could get $30/sf rent again and only give a $60/sf construction allowance for a ten year lease. Decent office buildings in major metros trade for about a 14-20x multiple on annual income so you holding out on rent for a year or more pays off much better than locking in an $8/sf rent for ten years. Regular commercial landlords don't do 1-2 year leases because it takes a lot to build out the space and you amortize that over the lease term. WeWork and Regus will do short term leases and they capture all of that demand, but it's still light so they aren't gobbling up the remaining office inventory either.