I don't understand pedestrian crossing lights. Why do you have to push a button to make them work? As far as I can tell, it doesn't extend your crossing time, which makes me think it's about conserving energy: why make the pedestrian crossing lights work when there are no pedestrians? Except that, when there are no pedestrians, the light stays "Don't Walk". Does that somehow conserve energy?
Pushing a pedestrian crossing button to get the sign to work seems stupid, but I'm going to take Chesterton's advice and assume there must be or have been a good reason for them. But what is it?
Many intersections only rarely have pedestrians, and halting traffic for 30-40 seconds every few minutes to let imaginary pedestrians cross is inefficient. Same reason some intersections with little vehicular cross-traffic will default to the main route staying green until a sensor detects a car on the crossing road, except it's harder to make a reliable automated sensor for a thing that isn't a metric ton of steel.
Well in the Texas town where I live, the pedestrian WALK sign doesn't halt vehicular traffic. It just turns the WALK sign on when the parallel traffic has a green light. So it doesn't protect you from cars turning right on a red light or oncoming traffic turning left. As a result, pedestrians get hit all the time by traffic while crossing with the WALK sign on. I, myself, jaywalk whenever possible, because it is much safer than trusting that a car won't turn into you at a corner.
It's plausible the WALK sign makes the green light in your direction an imperceptible-to-humans fraction of a second longer, but, being human, I can't detect if that's the case.
Most - but definitely not all - drivers in Saint Paul will yield to a pedestrian at a crossing. It gets weird at times. I have to cross a busy street without a light on my way to and from the corner grocery store. If I so much as glance to the other side of the street when I’m at a crosswalk, cars will slow to a stop for me.
I’m fine with waiting for a break in traffic but some hyper polite drivers won’t even respond to a waved arm ‘Go ahead I’m not in any hurry’ gesture.
Pete Davidson canceled his trip into space? I don’t think he realizes how a thing that could make him more interesting to women. Why his romantic life would probably just take right off. No more lonely nights with something like that on his CV!
Zvi in his most recent Covidpost mentioned that he couldn't find a video because it was pulled from YouTube. I found the video by putting the YouTube URL into the Wayback Machine - I didn't actually think the Wayback Machine worked on videos, but apparently it does.
Posted in case anyone here read said Covidpost and is interested in the video, or in case someone here can get it to Zvi/is Zvi (I don't know any way to tell Zvi things other than making a commenter account on one of his blogs), or in case anyone here doesn't know that the Wayback Machine works on YouTube (which is a pretty big deal considering how much stuff YouTube burns).
As a guy who is getting old and no longer in the market, thought I would offer some dating advice since I see guys here sometimes asking for it.
One thing I learned over the course of decades is that women reject you for two main reasons: because you are too in a hurry to get laid or because you are too in a hurry to have a serious relationship. It's easy to mistake a type 1 rejection for a type 2 rejection and vice versa. Maybe that seems obvious, but it didn't seem obvious to me when i was younger so I doubt it seems obvious to every young guy reading this.
An important take-away, I think, is to realize that when a women rejects you it is often for the opposite reason that you imagine. You could learn this from reading Proust, but I'm going to try to keep this shorter than Proust did. (Proust may have been gay, but he understood romantic relationships and sex better than most, straight gay or otherwise.)
If you think a woman rejects you because you aren't attractive enough, that probably isn't the reason. It's more likely because you seem either too interested in getting laid or too interesting in having a long term relationship. Meaning, if you get rejected often, you should change your strategy. If you are trying to get laid, don't. Work on signaling that you are interested in a relationship. OTOH, if you are mainly interested in a relationship, don't. Just try to get laid.
Either way, women are going to figure out what you are really interested in pretty quickly, so don't worry about sending the wrong signals. Do everything you can to counter-signal, because that will send a more balanced signal in the short run. Sending a balanced signal is what most women find attractive.
EDIT: And don't make the mistake of thinking "but THIS WOMAN I am interested in isn't MOST WOMEN". You can't read minds.
I really don't think it is. While men rate women 1-10, women are mjch more pass-fail regarding men. And there's a low bar to passing: don't be smaller than the woman, smell OK. Even those criteria will be waived if you're funny as fuck.
It also helps to be/appear deeply interested it something other than sex or relationship. Something she can relate to, e.g. some hobby you have in common.
Today's post-secondary institutions are expensive and backward. How can we do better?
Idea for accreditation system based on accrediting students rather than institutions:
1. An accreditation company (nonprofit foundation? PBC?) produces standardized tests to measure student knowledge & abilities. Tests are broken down into a set of mini-tests, and each mini-test tests a small amount of knowledge and ability. The company charges money to an institution whose students take a test, or to a person who takes a test independently, and revenues are used to produce more tests (and to prepare defenses against cheating). Students earn diplomas from the company according to some set of rules to be determined. Much like TripleByte, the accreditation company earns reputation by verifying ability correctly, and it can increase prices as its reputation increases (but if it's a nonprofit or PBC, prices should hopefully not rise without limit.) IMO students should take tests on the same topic twice, at least 8 months apart (to verify knowledge retention and discourage cram-based learning), but that's not my call to make.
2. Educational institutions (e.g. MOOCs) teach students, who pay tuition as usual. The institution chooses what (and how) to teach in each of its courses, and at regular intervals, offers a test from the accreditation company. A test will typically be composed of two to twenty mini-tests chosen according to the material that was taught in the course. The institution earns revenue equal to the difference between tuition fees and test costs. The accreditation company tracks which mini-tests each student has passed; if a student moves between institutions, the gaps and overlaps between courses at the two institutions are tracked exactly. And of course, some people (Sal Kahn?) will offer completely free courses.
This type of system bypasses traditional accreditation boards run by encumbent institutions, which have an incentive to avoid accrediting new entrants. Thus it should produce a competitive online market with low prices, while still giving students a meaningful diploma that they can tout to employers.
Now, surely I'm not the first to think of this, so why hasn't this kind of system become popular?
"Now, surely I'm not the first to think of this, so why hasn't this kind of system become popular?"
They do exist, so I think the question is "why aren't they more popular?" and the short answer is that because if the accreditation is to mean anything and not be a diploma mill, you need some way of checking that the course material is good, the students are qualified, and the diplomas or certificates mean something - and that means you end up re-inventing colleges in some form. E.g. how can you be sure QuadrupleBit graduates are as qualified (this is distinct from capable or clever) as IvyWreathed U graduates? One way is to compare coursework and results. But if both institutions have different forms of assessment and coursework? Well, get QB grads to sit a final exam.
That means you need questions for the exam, a curriculum to cover exam topics, and a place to sit the exam where you can be sure that the QB lot are not all cheating and just have Google open on their computer at home to give them the answers.
Congratulations, you have re-invented the exam hall. And if you need a place to host it, the simplest answer is to get QuadrupleBit, the online certifying company, to hire or rent or find someplace to hold that exam. And since they need this place on a permanent basis and for as many final exams as they're running throughout the year, they may as well have a building for their own use.
And eventually 'online only' becomes 'well we have all these offices and now classroom spaces as well' and it's a new college.
How do you develop a reputation as an accreditation company in the first place?
Universities have history, your QuadrupleByte has nothing. At best you're rated as equivalent to that shitty bootcamp someone's running in the city, whose alumni cannot program their way out of a wet paper bag.
You want is an institution that is a Schelling point for intellectual elites, and does some basic filtering to keep that Schelling point stable. If it is known that clever kids finish CS at $EXPENSIVE_COLLEGE, companies will hire from $EXPENSIVE_COLLEGE. The actual quality of courses at $EXPENSIVE_COLLEGE is an afterthought.
Speaking from my own experience as someone who got a Bachelor's in Computer Science, and then went into Software Engineering - conceivably the very sort of person you're trying to reach - I think that the big missing element is projects.
I credit a great deal of my success in industry to the project-oriented curriculum at the institution I attended, and particularly the class which brought together groups of fifteen people for seven weeks for the closest I got to an actual Software Engineering experience in school. They introduced me to ideas and practices around teamwork and source control which I suppose I could recite, but which would be very difficult to articulate in ways that could convince an observer that I had gotten them in a brief period of time. And if perhaps I could, the ability to verify in a test environment that I had these skills would, I think, hinge on my communication skills to an undue degree - while communication skills are important in my field, I do think this would put more weight on them than is warranted.
It's quite possible that schools do a poor job of actually verifying that a given student has learned the things that a given project is meant to teach. Perhaps they are simply going off some strong prior - whether it's the latest in education research, the dead reckoning of their most experienced faculty, or the dean's latest interpretation of their star signs (though hopefully accreditation puts some kind of lower bound on how bad that methodology can be) - that tells them that given a project which roughly fulfills these specifications, the students should have learned these skills, and if that's true of only 90% of the students and even those learn on average 90% of the skills, there's the opportunity to instill it with further overlap and maybe we'll catch it before they graduate, if they missed out on a foundational skill in their first year that a fourth-year project depends on in the same way that it depends on the students being functionally literate in the university's language.
But I don't think there's a way to identify the kind of learning we want to happen from projects in an environment where the assessments and the teaching are so decoupled. Tests can be fine for getting a student to apply specific knowledge in a narrow sandbox, but they can't really capture how well a student works at something - over days, over weeks, consulting resources (because consulting resources is a skill, which may well be more important than what most given resources have to teach; as a Software Engineer much of my job relies not on me knowing what needs to be done before but on being able to find, interpret, and apply the relevant sources).
I think this gets much worse for fields where there is very definitely not a singular concise correct answer at the end of an equation or as the result of a suite of automated tests. Which is a shame, because one big thing we want from knowledge workers is the ability to work somewhat autonomously, on broad problems with ill-defined endpoints, often collaboratively, and in novel situations. The stuff that doesn't fit this description - well, that'll get snapped up by automation sooner than later.
You'd be turning universities into teach-to-the-test cram schools.
Most of what you'd expect to learn at the university level can't be adequately tested in an examination, and for the stuff that _is_, it would be reasonably easy to cram (e.g. there's only, like, twelve possible classes of exam question for Special Relativity so let's just study them all).
Since universities already have accreditation, I would expect them to reject a new system completely.
But when I was going to my university Engineering program, I crammed a lot because my teachers were so bad that I felt that *actually understanding* the course material was out of reach for me. This is completely different than my high-school experience, in which I *never* crammed.
Uncharitable of you to ignore the anti-cramming measure that I proposed. Cramming is a short-term trick; if you have two tests spaced >8 months apart, you are certainly better off learning the material properly.
It is impossible to "learn the material" for the long term if it has no day-to-day practical use. The material would need to be integrated into practical projects executed during that 8 month interval.
Not sure why you think cramming twice is "better" than learning the material properly. I disagree about the impossibility of learning things without day-to-day practical use; I learned such things throughout K-12 without cramming.
Current education/accreditation system is a mess of all kinds of signals. It is difficult to improve, because sometimes it is supposed to suck.
For example, imagine that you create a parallel educational system that is neither better nor worse than the traditional one, but it is 10x cheaper. Would it be popular? No, because using the new system would signal that you are poor, and so are most people you know. Rich people would avoid it; and they would also have an incentive to pretend that the new system is worse.
Or imagine a new educational system that is just as good, only 10x less frustrating for students. Then employers would avoid hiring people from that system, because they would suspect that such employees would have low frustration tolerance and would soon quit after the normal amount of abuse at workplace.
Or a new system where kids learn more easily because somehow all lessons are magically easy to understand? The employers would suspect that the kids are actually less smart then they seem, and will fail when facing a novel situation.
In other words, trying to make education better is like trying to make a marathon shorter -- the people who usually run marathons will reject the idea. Education is inefficient, frustrating, and gives unfair advantage to rich people... which is exactly the point. It prepares you perfectly for your future workplace.
(If you are interested in knowledge for knowledge's sake, then of course, Khan Academy is the way.)
> No, because using the new system would signal that you are poor
I'm pretty sure most employers aren't judging you based on how rich your parents are, with some obvious exceptions (Harvard) in which wealth isn't the only thing being signaled.
> employers would avoid hiring people from that system, because they would suspect that such employees would have low frustration tolerance
As a (very) small-fry CTO myself, I want knowledge and skill and more than I want frustration tolerance. I guess some companies want that, but other companies want other things. (and why would poor people have worse frustration tolerance?)
"(and why would poor people have worse frustration tolerance?)"
It's not that poor people would have worse frustration tolerance, it's that (whether rich or poor) coming out of an education system where learning was effortless, the teachers wee excellent, it was easy to learn, everything was ready for you when you were ready to take the next step, etc. is not good preparation for a workplace where it's "dunno the answer to that, the guy who does know the answer is out for two weeks so you have to wait that long, there is paperwork with the answer on it someplace but you'll have to search for it and nobody knows where to start, and there isn't a clear-cut answer at the end, just a 'good enough' and besides the boss/boss's boss/client is going to change their mind three times about what they want anyway".
The elephant in the room is that if your parents are rich they're probably higher in IQ and/or cultural capital, which makes you higher in IQ and/or cultural capital, both of which are hugely desirable traits in new hires.
(Ah, please take the previous comment with a grain of salt; it was exaggerated for artistic purposes. This comment reflects my beliefs literally.)
There are two ways how wealth impacts knowledge/skills that come to my mind immediately (and there are probably more):
First, rich people can spend more money on education-related expenses, and I think they have more free time on average (no need to keep two jobs; can save some time by spending money on something). So if we imagine two kids with equal intelligence, talent, chatacter traits, and hobbies; learning e.g. computer science in exactly the same classroom with the same teacher using the same curriculum; but one comes from a poor family and the other comes from a rich family, I would still expect different results, given the following:
The rich kid will have a better computer at home, more time to use it, no problem with paying for a course or a tutor if necessary. The poor kid will feel lucky to have a computer at all, will be limited to free resources, and will probably spend some time helping their family (working, taking care of younger siblings). -- Therefore, at the end of the year I would expect the rich kid to know more, on average.
Second, when people think about the quality of school, they usually think about teachers, curriculum, didactic tools, and whatever... but a crucial and often ignored factor is classmates. Kids inspire each other and learn from each other a lot. Or they can prevent each other from learning by disrupting the lessons. Your ability to choose a school with better classmates is often limited by money. With the same curriculum and same quality of teachers, I would expect rich kids to get better outcome at school, simply because they are not surrounded by classmates from dysfunctional families etc.
They differences may seem small, but they add together, and their effects compound. With exactly the same curriculum, I would expect rich kids to get better results, on average, even if we control for intelligence and other traits.
Therefore, from the company perspective, a rich kid seems a better bet, ceteris paribus. (The only disadvantage is that the rich kid will probably expect a higher salary.) Therefore, if you are - or pretend to be - a rich kid, you probably do not want to signal that you attended the school for poor kids... or the cheap school.
Re: frustration tolerance -- you want some basic level of it, like people who won't give up and quit after the first problem.
And speaking as someone who has come out of the 'practical training, non-university, on a ladder of accreditation' route and worked with people who went the university route, there is a difference. Not even so much in practical skills, depending what course they did the university people can be as good and up on those, but the entire shape of the learning experience, the environment, the content.
The practical-oriented really was teaching to the exam, telling you what you needed to know to do the tasks, but nothing extra. It was to get you qualified and out into some kind of paying job as fast as possible. If you wanted extra, you could then go up the ladder to the university. People from the university just had an entirely different experience, and yes they did indeed seem to have a more rounded education, a better understanding of the subject, and just that familiarity with how academia works that is hard to put into words but you know it when you don't have it and are trying to engage on the academic level.
Hey, you seem to know your shit. Could really use an impartial set of eyes on something...would you be willing to read a few pages and give an opinion?
It seems like this sort of info about a student's mastery and retention of material would be taken seriously by grad programs or employers in STEM fields -- have my doubts about other fields though.
The step you're missing is to make big companies accept your accreditation as a valid measure of a prospective employee's worth. Companies are often at least as interested in proof of conscientiousness as in intelligence, and often not at all interested in subject matter mastery.
At higher levels, the top institutions sell networking as much as anything else, and that's a very sticky equilibrium - one of the main benefit of a Harvard grad is that they know other Harvard grads, and that's valuable because all the top places hire Harvard grads, so their network makes them valuable enough for all the top places to want to hire them
Yes, and this is why I suggested that the price of the service would be correlated to its reputation: big companies accepting it = reputation. I guess it would need somebody with deep pockets to help it survive the early low-reputation phase.
Obviously, my proposal is not intended for people who have the means to make it into (and through) Harvard. I went to an ordinary university, and the amount of networking I did there was basically zero. (edit: except the internship program - the job I got out of that lasted 7 years.)
What's the textbook example of how to tactically use tanks in warfare? I'm thinking a WW2 battle or something were tanks saved the day, did lots of the special tanks things that can't be done by artillery or motorized infantry or whatever, and then some colonel wrote the book on it.
I'm hearing a lot from armchair generals on how to not use tanks ("Don't use them against other tanks", "Don't use them in urban areas", "Don't use them without infantry support" etc.), but I don't hear much about how they are supposed to be used, thus my question.
If you want to ague that tanks are obsolete in modern warfare this is your spot as well I guess.
If your enemy has a tank (or worse, a bunch of them) somewhere, and you don't have anti-tank forces nearby, then the tank can destroy whatever stuff you put through that area, unless they are well-hidden or well-fortified. And it's hard to move while being well-hidden or well-fortified, so if the tank got there first you are outta luck. Of course, you are aware of that and will not move into that area, but that means you can't have anything there.
That makes tanks basically movable walls (the size of the wall is the range of the tank of course, not the physical size of the tank). You put them where you don't want the other guy to go through - an important version of that is to cut off a force from their rear lines and force them to surrender.
The reason they are good at that is that they are armored pretty well and can fire on the move, which makes anti-tank platforms far more limited than say anti-artillery platforms (it's easier to sneak up on a self-propelled gun or destroy it after it fires, than to do the same on a tank).
First, a tank is a km-length wall that can move operationally at speeds of tens of km/h. A kilometer of wall is a lot of wall to place, or to move to a more relevant location when needed (and you want to fill in or exploit breaches quickly, which makes moving to relevant locations very important).
Second, its much harder to destroy a tank (that is maneuvering through tank-friendly terrain, not that is driving through hostile city streets) than to breach a wall, because the tank can shoot at you then move into hiding and possibly call for assistance, and a concrete wall can't.
Depends a lot on the period. You can get a good spiel of the WW1 retrospective & pre-WW2 expectations (which held up at least for early WW2) by reading "Achtung - Panzer" by Guderian, but the key doctrinal takeaway I remember would be:
-Use them in large concentrations (i.e: not in piecemeal), with motorized infantry to support and exploit their breakthrough (and absolutely not "supported" by leg infantry, which can't match their speed, and slow them down to get pummeled by artillery fire, as in WW1, or WW2 french doctrine).
-Best way to stop them is another tank (that was before air support got precise enough to pose a threat)
Supposedly De Gaulle reached similar conclusions in "Vers l'armée de métier" (or was it "La France et son armée"?), but I haven't got around to read it yet, so I can't comment.
What you're really asking for is how to do combined-arms warfare. There are very few military problems where the solution is "send tanks, just tanks", but many where tanks are a useful or vital part of the solution.
The classic use case for tanks, and *maybe* justifying a pure-tank force, is exploiting a breakthrough in mobile warfare. Not making the breakthrough itself; you'll almost certainly want artillery and infantry to help with that. But if you've broken through the enemy's defenses, you want to move fast and break things behind the lines before they can offer a coherent response, which will involve meeting engagements with elements of their incoherent response, and that's something tanks are really good at.
And for much the same reason, tanks are good at mounting rapid counterattacks in the face of an enemy breakthrough.
There are some good recent-ish examples (but not pure-tank) in operation Desert Storm, and in the Arab-Israeli wars.
Sure, I get that you want to do combined arms warfare. But what is the role of tanks in combined arms warfare?
Why are they better than e.g. mechanized infantry at exploiting breakthroughs? Are they better armored so that they have an easier time driving past pockets of resistance? (If so, can't we just slap more armor on an APC to achieve something similar?)
What does exploiting a breakthrough actually entail in practice? Do you hunt down enemy artillery and C&C? Do you try to get the enemy logistics (by firing randomly at trucks)? Do you attack the enemy in the rear? Do you just drive as fast as you can against Berlin/Bagdad and hope to create as much chaos as possible on the way? I guess tanks need (mechanized) infantry to create a famous WW2/style pocket, but it's good to have the tanks in the front while doing so?
The Iraq tank battles just seem to be coalition tanks driving through and obliterating technologically inferior Iraqy tanks. Couldn't this have been done by infantry or air power? Valley of Tears looks interesting but the Wikipedia article is hard to parse, I'll look into it.
Mechanized or even motorized infantry can support breakthrough operations, and in some cases (e.g. when the enemy only has leg infantry), can do the whole thing.
But infantry, even mechanized, has to dismount to fight. Otherwise it's not infantry, it's just an inferior form of armor handicapped by having to haul a bunch of useless people around - and no, their shooting assault rifles out of firing ports isn't useful enough to be worth the bother.
And dismounting, fighting even a skirmish at a walking pace, and then remounting, takes time and costs tempo. Sometimes it's necessary, e.g. to clear enemy infantry blocking your advance in close terrain, but if at all possible in a breakthrough operation you want to defeat at least minor blocking forces on the move. For that, you want tanks.
As for what you do in a breakthrough, part of it trying to overrun C3I facilities, logistics nodes (not random supply dumps, but supply depots, railheads, critical road junctions etc), and artillery. I'd put it in that order of importance, but it's debatable. The other part of it is maneuvering to block enemy lines of retreat and reinforcement.
And pretty much all of it, is creating in the enemy's front-line troops and their leaders the firm perception that they don't know what the hell is going on in their rear, but it's really bad and if they don't run away *right now* they'll never have the chance. Then your breakthrough forces can ambush and kill them as they flee.
As for using infantry to "drive through and obliterate technologically inferior tanks", no. Literally no, because infantry doesn't drive, it walks. And if you're thinking they're going to drive through in their technologically superior infantry fighting vehicles, maybe but now the "infantry" isn't doing anything and even technologically inferior tanks are carrying bigger guns with longer range and heavier armor because the aren't carrying around useless infantry. Maybe you've got enough of a technological edge for that, but it's still fighting with a handicap.
If you're talking about infantry advancing against tanks on foot, no, infantry can't advance against fire like that. First, because infantry survives under fire by emulating hobbits - small, nigh-invisible, and living in holes in the ground. And second, because infantry can't fire on the move.
Even a "technologically inferior" tanks is in this context a machine gun on a gyrostabilized mount with a magnifying and if needed night-vision optical sight and nigh-infinite ammunition, with a gunner who is basically immune to suppressive fire, and capable of firing accurately while retreating faster than infantry can advance. Think first-person shooter video game - completely unrealistic at duplicating the experience of a *soldier* in combat, but pretty good for a tank gunner. Oh, and he has a big-ass cannon if he needs it.
The infantryman, is a very not-machine-gun-proof target whose attempt to advance largely voids his concealment and subterranean-ness, and sure maybe he's carrying a high-tech missile that could destroy tanks if he weren't trying to advance, but since he is it's just ballast slowing him down.
If the enemy brought *only* tanks, and parked them too close to a town or treeline or whatnot, you could imagine your infantry sneakily infiltrating into firing positions. But the enemy probably has some infantry of his own, and he's got that deployed to cover all those firing positions you were trying to sneak into. His infantry doesn't have to defeat your infantry, it just has to force it to reveal itself prematurely so that the enemy can put heavy firepower from his tanks - or better yet artillery - onto it.
Wait, that first line makes no sense to me, I thought (a subset of) tanks are explicitly designed to be used against other tanks?
More generally, I think tanks in open terrain beat infantry without tanks; combined arms is generally very important, of course, and the tanks can't be totally unsupported, but outside of cities it's hard to sneak up on a tank. Air superiority is of course the ultimate trump card, tanks don't beat bombers.
Historically, I think Germany's Blitzkreig of France is the ur-example of the value of light tanks and motorized infantry. The biggest advantage of tanks over artillery is their mobility, so if you want examples of things only tanks can do look for maneuver warfare.
It's a misunderstanding/oversimplification of American armored warfare doctrine during WW2, which emphasized tank destroyers (typically in battalion-sized units attached to infantry divisions) as a defensive counter to massed enemy armored offensives. The misunderstanding comes in reading this as saying that *only* TDs should fight tanks. Tanks were seen in this doctrine as being perfectly capable of fighting other tanks. The actual point of the TD emphasis was that the design differences optimized TDs for a defensive response role, while tanks were optimized for breakthrough and exploitation: American TDs of mid-to-late WW2 were somewhat cheaper, a bit faster, and mounted heavier guns than tanks of the same generation, at the cost of substantially lighter armor, making them better at responding to enemy offensives and fighting defensively from cover with infantry and artillery support, but much less survivable on the offensive. This, TDs were the first-line response in support of infantry for enemy armored offensives, freeing up tanks for things they did better than TDs.
When the doctrine was first developed, the intent was for TD to be a lot cheaper than tanks, initially conceived as light infantry-support antitank guns towed or mounted on jeeps or light trucks. This way, you could have TDs everywhere you needed them for a fraction of the cost of tanks. But by mid-war, TD designs got more capable, more tank-like, and correspondingly more expensive so an M4 Sherman tank wound up being only a little more expensive than an M10 tank destroyer.
What would have happened if the Germany had done the invasion of France with 50% less tanks and more motorized infantry instead? Would it have been worse, and if so why?
This is a random place to post this but I asked a question like this before on an open thread and Scott responded. Any help appreciated from anyone with relevant medical industry knowledge.
I am certain I have ADHD. It hugely affects my job performance and I'm constantly worried about getting fired. I'm hoping to get prescribed adderall. I'm wondering what the chances are with my current planned process:
-I have made an appointment with a telehealth psychiatrist through some large online group.
-This site specifically says they themselves do not prescribe drugs like Xanax and adderall, but that if they think it's necessary they can fax your PCP to have them make a prescription.
-I have made an appointment to see someone as a PCP next week, two days before the psychiatry appointment.
-This appointment is my first time meeting them and I said I wanted to talk about ADHD in an office visit
-But they are a Nurse Practitioner, so I have no idea if they're allowed to prescribe anything/more hesitant.
Checkout Ahead (helloahead.com). The PNP I work with prescribed me Adderall. No contact with a PCP (which I don't have). They also manage my anxiety meds.
Some NPs are not allowed to prescribe certain drugs. It should be fine for you to call and clarify with them, asking them if you come with a referral will they be able to prescribe for you. Congrats on taking steps towards treatment.
The authors of Meta-analysis studies should be required to present a table listing the included studies and excluded studies along with the exclusion criteria that each of the excluded studies failed to meet. That is all.
They should also list their reasoning behind their inclusion/exclusion criteria!
Just finished reading two meta-analyses with differing conclusions, and I've come to the conclusion that I'm no closer to the truth than I was before this exercise. I don't know if one or both of the authors are trying to pull a fast one. And without being able to look at the studies included in the meta, I cannot judge the validity of either of the metas.
Wow, that's irritating. They put "Meta" in their title and become arbiters who can't be challenged. (Sort of like freakin' Facebook changing it's name.). What field are these articles in? I've read a dozen or so psychology and neurology meta-analyses recently and they all really spelled out their criteria for judging study quality. Many did not list all the articles they considered, but in many cases that seemed reasonable -- there were thousands. Maybe writers of metas should be required to publish that info in an appendix, though.
It was two meta-analyses of long-term COVID. Actually, they didn't have opposite conclusions—just different conclusions. They spelled out their criteria, but it was what wasn't mentioned that made me unable to compare the two. The first M-A's criteria was only studies published in the English language. The other one didn't mention that criteria, so I'm left wondering if they had a more diverse pool of international studies, and that that may account for the difference in conclusions.
Also the first study had a criteria that each study have a minimum sample size of 30. The other study had higher sample size criteria. But I'm left wondering why the first study was OK with 30. Seems too small to be statistically valid (?). And how many of those studies in the first one had sample sizes less than 100? Aarrggghhhh.
About study size: the smaller the study, the larger the effect size has to be to capture it. If I wanted to find out whether capybaras weighed more than hamsters, and compared 2 groups of 15 randomly selected members of each species, I would definitely find a statistically significant difference in weight. If a small study finds a statistically significant difference, then you can trust the result (unless there is some other problem with the study design -- such as wrong statistical methods done, groups compared were different in important ways beyond the one you were checking for). However, if small study finds no effect that may be because effect doesn't exist or just because study was too small to have enough power to capture the effect.
About getting info about long Covid: Epidemiologist Jetelina, on Substack, just put up a series of posts on the subject. So this is sort of her unofficial meta-analysis. I trust her, mostly. She’s smart and thorough and seems to have no ax to grind.
Ha! I got kicked off her Facebook group. It was early in the pandemic and I was questioning the some of the consensus wisdom of epidemiological theory. I thought I was polite, but I got exiled from that FB group. OTOH, I can be rather outspoken so maybe I deserved it. ;-)
Wow. Well, she does have a bit of a kindergarten teacher quality to her -- sort of pathologically nice and hyperconventional. If you can still stand to have anything further to do with her, her Substack blog seems pretty good to me -- informative, and politics-free.
As for getting kicked off things -- I'm a member of the too-rude-to-remain club too. Spent a year on Twitter, driven crazy by snark and trolls even though I only followed science writers. One day a red-state male troll started dropping turds on a thread about some technical virus thing, so I came back with the term most likely to offend his demographic: cocksucker. Now I'm banned from Twitter and glad of it. Heh.
‘ McDonald's sells "food” that is absolutely impervious to rot and decay. You can buy one of their hamburgers, put in on a shelf in your living room and just leave it there. After a year the burger will still look and smell the same. None of the rodents that you unwittingly share your house with will have deigned to touch it. Nor will any insect, no fly no wasp, nothing. Even bacteria will stay away from McDonald's products. This will give you an idea of the quality of American fast food. KFC specializes in products made from bio-engineered, hormone and antibiotics-fed chickens growing so fast they never learn to walk. The meat from such creatures will probably help accelerate your transition from "cisgender” to anything in the LGBTQ spectrum, whether you want it or not. Starbuck's specializes in something it dares to call coffee but that anyone who really knows and likes coffee will shun.’
I never felt "wow this is quality food" while eating McD, it always tastes like some kind of space colony faux food that's supposed to remind me of what they used to eat back on Earth, but fails at it.
> None of the rodents that you unwittingly share your house with will have deigned to touch it
McDonald's, KFC and especially Starbucks may be terrible, but I'm glad I live in a country where it's not simply taken for granted that every house has rodents.
"Degenerate Western food makes you trans" (presumably in contrast to virile, natural Russian foodstuffs) is a take so cartoonishly anti-Woke I didn't think pravda.ru would actually publish it.
That sort of stuff is aimed at Poland, Czech Republic etc. Russia is the defender of traditional values. It’s an on going thing with Pravda. Trying to appeal to East European NATO countries. I know. It’s all so very strange.
The grammar is pretty good in this one. I’m guessing it was written in English by a human. The machine translated stuff usually makes a hash of idioms. The underlying differences in grammars show through too.
Are regular KKK meetings actually a thing? I was under the impression that the KKK doesn't meaningfully exist any more. This isn't Robert Byrd's day.
The ADL (whose incentives certainly run towards maximising rather than minimising the extent of Klan activity) most recently https://www.adl.org/education/resources/reports/state-of-the-kkk reports the existence of thirty groups claiming to be the Klan, but most of them are just a handful of people and they tend to pop in and out of existence rather rapidly as people lose interest.
Sadly, the actual results seem to only be in the article body, which is behind a paywall. Unless someone here has institutional access to Sagepub or JSTOR and feels like reading it and summarizing for us.
I don't like that study because it focuses on "former VIOLENT U.S. White supremacists" (emphasis mine). I'd rather see the data for the KKK members who don't actually act on their racist beliefs by attacking nonwhites (they probably form the group's majority).
My hypothesis is that, if you're so racist that you're actually willing to go to KKK meetings, you're probably mentally ill.
Seems like a fair comparison group would be other extremists, both left, right, & totally disaffiliated -- like test the people in Anonymous (if only they weren't all anonymous!). It may be that unhappy and desperate people are drawn to extremism, and/or extreme views create desperation. If some of the wilder apocalyptic theories were true, suicide might be a rational choice.
Notably, he says that Alexsandr Dugin ("Putin's brain") wrote a book in 1997 called "Foundations of Geopolitics: the Geopolitical Future of Russia" which is Putin's playbook and can be used to understand and predict Putin's moves.
The book's 40-year plan:
Step 1. Invade Georgia
Step 2. Annex Crimea and control Ukraine
Step 3. Separate Great Britain from Europe (Brexit?)
Step 4. Chaos: sow division in Britain and the US
Step 5. Create "Eurasia" which (based on the map) looks like basically Russia surrounded by "buffer states", with China "divided and in turmoil", and Japan and India as allies of Russia (I note that while Japan voted to condemn Russia's invasion, India was neutral and is now setting up a special payment system to avoid commerce interruptions caused by sanctions. Evidently Russia bagged China as a Russian ally instead of Japan. My impression is that while China isn't completely sold on the invasion yet, it is spiritually siding with Russia and the reason it isn't doing more to help Russia is that it fears "secondary sanctions".)
Steps 3 and 4 involve using the 3 Ds, Deception, Destabilization and Disinformation, to create internal divisions in Britain and the U.S.; internal divisions in the U.S. are to make the U.S. more isolatist and distant from Europe (hence Putin's support for Trump, who in turn pulled out of multiple international treaties). Also the book calls for a "Continental Russia-Islamist alliance [as] the foundation of anti-Atlanticist strategy" (hence Russia's ties to Iran & Syria)
He also has a video about China's master plan: https://www.youtube.com/watch?v=WaAOss6W1u0 - this video begins by telling me that China already beat the U.S. on the metric of "GDP by PPP (purchasing power parity)" in 2014, though note that the per-capita *incomes* of Chinese people by PPP are just over one-fourth the incomes of U.S. people. Which itself is probably part of the plan... to sacrifice income for more GDP and more power on the world stage. China seeks world domination, and on the economic front, they seem to be ahead of schedule.
Oh, now that makes me wonder if some Chinese policy wonk read that book and decided that instead of letting Russia be Ruler of Eurasia, with Japan as an ally and China internally divided and in turmoil, it would be smarter to cosy up to Russia, sell themselves as an ally, and remain a major, non-conflicted, partner if or when Eurasia is a thing that happens.
That would explain (to me) why China is lining up with Russia right now, instead of sitting back and seeing how things play out. Even if Russia manages to shoot itself in both feet with Ukraine, China can still be the "I'm your friend, see how I supported you?" partner and be in a good position to gather up the fragments from the fall-out if Russia instead starts falling apart with internal conflict.
Meanwhile, in this video he predicted a near-term financial crisis which didn't happen (we just got some inflation, and if there's a crisis now I think it'll be triggered by Russia) - I suspect that he doesn't understand macroeconomics well enough (which is not unusual; my impression is that even economists themselves have multiple incompatible models that make different predictions): https://www.youtube.com/watch?v=EYOVoQT2yQg
In his otherwise reasonably accurate 2020 prediction video about vaccines, he characterizes what sounds like it should have been a crisis in 2021 as "Massive wave of bankruptcies, unemployment rise again, and debt bubbles bursting": https://www.youtube.com/watch?v=yahfx_JIihQ ... looks like he overweights the importance of debt and QE https://www.youtube.com/watch?v=nUOVRo_EIrE ... whereas my model is closer to market monetarism: I do not find debt to be important except indirectly, and I expected a "market adjustment" but no crash.
I have a friend who's in the market for a dating coach. He's in the general Astral Codex Ten audience demographics - about 30 years old, tech professional, generally liberal. He has been unable to find a good option with experience working in those demographics. Does anyone have a person to refer him to?
I would advise finding a local one. Dating advice changes quite significantly depending on country or even city. A local coach knows local quirks and also good spots/locations to meet people.
Also check the coaches age. A lot of them are mid 20s or even younger. The game works differently for 30+. A 22 year old won't give you useful advice for your age bracket.
If your friend is into online dating I might give some pointers on optimizing his profile. Been doing a lot of work on this topic for a machine learning project I'm working on.
In general the youtube channel "School of Attraction" has a lot of online dating advice. Also search for "Reddit Tinder Guide". There are multiple good ones.
More than that we need to go into the specifics of the actual profile. I'm pretty new at substack and don't know if there is any kind of personal message. But if you want you can contact me and i can give you/your friend specific advice about the profile(s)
I’m going to assume that date coaching is a thing now. It wasn’t when I was single.
Is your friend very shy? That would make it harder. If that is part of the problem, he could try getting regular exercise. Cardio and weight training relieve anxiety and help with self confidence.
If he is up to it, being able to dance a bit would give him a chance to meet potential partners.
I grew up in Russia (but left to the US as a teenager, on my father's H1B visa). This means I still know a number of people in Russia who are disproportionately techy, and statistically I'd expect some of them to be interested in no longer being in Russia right now. (Some will have left already, some will want to stay no matter what.) Is there a more effective way to look for jobs that might sponsor them than "ask your company if they'll sponsor a visa, ask your friends to ask their companies if they'll sponsor a visa, etc."? Also, consider yourself asked :)
I work for a biggish consulting company and our policy is to sponsor advanced degree holders (MBAs, PhDs, MDs etc.) coming in at the consultant level but not for analysts who usually come in at the undergrad (BS/BA) level. That's for the U.S., not sure what the policy is in our international offices. I've also asked about expanding sponsorship to further down the ladder and it's definitely being considered but I don't think that policy is likely to change at least in the short term.
If you know anyone who might be interested they can reach me for more info at gbz.uraarffrl@tznvy.pbz (rot13)
Note that I asked this again lower down in the advert post, and they (Dave92f1) said that they're willing in principle but haven't done it in practice, and also that they'd consider remote.
Also, if you want to send a resume my way, my company (http://www.cyberoptics.com/) is looking for at least one software person and at least in principle willing to sponsor people; I'm lastname at gmail.
Scott, thank you for helping set me on the path towards effective altruism. Your writing was deeply influential to me in high school and early college, and I think it was a really big part of why I got into EA (where I get a lot of self-esteem from these days). Since I think its relevant, I'm a senior software engineer at a FAANG and I donate around 30% of my pre-tax income, so include some fraction of that in your total impact!
[Context, I'm reading through 'What got you here won't get you there'. It recommends thanking the top 25 folks most influential in your professional life. Scott handily qualifies for me.]
Seconded. I'm in the middle of a career switch from lucrative but soulless software dev to medical bioinformatics that's socially useful, very interesting, very frustrating, and paying next to nothing.
Scott's writing, especially UNSONG, has been one of the main things that pushed me to finally do it and ruin/fix my life.
Has anyone seen studies citing a failure rate of one country invading another country? It seems like this might help forecasters take the outside view of Russia-Ukraine. I can't figure out the right search terms.
I expect a wider range than the 70%-90% for M&A failure. One reason is the difficulty of identifying invasions due to proxy warfare. (Should the Bay of Pigs landing by anti-Castro Cuban exiles be classified as the US invading Cuba by proxy, or an abortive civil war?) Another reason is the difficulty of defining failure, since political goals are harder to evaluate than corporate profits/losses.
The rates probably vary by technological era, as new weapons make offense or defense easier.
I have a theory that Scott pseudonymously wrote an alchemical allegory disguised as a bad Harry Potter fanfic, but nobody got the joke, so he had to write a whole essay explaining it. If true, that is my favourite.
Reading "Sort by Controversial" and the comments thereof made me learn the origin of the phrase "not by one iota", which is now one of my favorite facts. Is this what people learn in Sunday School? Why did no one tell me?
> The First Council of Nicaea in 325 debated the terms homoousios and homoiousios. The word homoousios means "same substance", whereas the word homoiousios means "similar substance". The council affirmed the Father, Son, and Holy Spirit (Godhead) are of the homoousious (same substance). This is the source of the English idiom "differ not by one iota." Note that the words homoousios and homoiousios differ only by one 'i' (or the Greek letter iota). Thus, to say two things differ not one iota, is to say that they are the same substance.
Except, is that etymology true? The online dictionaries I check don't mention it as an origin for that meaning and instead say it's from iota being the smallest letter and therefore almost insignificant. (The do mention that 'jot' derives from this, as iota also is transcribed as jota.) And Wiktionary quotes it as being from the New Testament "until heaven and earth pass away, not an iota, not a dot, will pass from the Law". So that predates Nicaea (I presume?).
So, uh, is this a scissor statement? Discuss at your own risk.
One piece of evidence against this etymology - Hebrew has an expression "On the tip of a yod", which means "decided based on a really tiny difference between two otherwise equal things", which feels like the same expression. Yod is the Hebrew alphabet version of iota and is just a really small letter (אבגדהוזחטי - yod is the little one on the left). Iota is also a pretty small letter. So if "on the tip of a yod" and "by one iota" have the same origin it was probably from the graphics of it rather than some complex greek spelling.
Why isn't there a political movement against the tactics of the IRS?
Libertarianism?
I mean specifically their dirty tactics not their existence or income taxes in general.
I don't understand pedestrian crossing lights. Why do you have to push a button to make them work? As far as I can tell, it doesn't extend your crossing time, which makes me think it's about conserving energy: why make the pedestrian crossing lights work when there are no pedestrians? Except that, when there are no pedestrians, the light stays "Don't Walk". Does that somehow conserve energy?
Pushing a pedestrian crossing button to get the sign to work seems stupid, but I'm going to take Chesterton's advice and assume there must be or have been a good reason for them. But what is it?
Many intersections only rarely have pedestrians, and halting traffic for 30-40 seconds every few minutes to let imaginary pedestrians cross is inefficient. Same reason some intersections with little vehicular cross-traffic will default to the main route staying green until a sensor detects a car on the crossing road, except it's harder to make a reliable automated sensor for a thing that isn't a metric ton of steel.
Well in the Texas town where I live, the pedestrian WALK sign doesn't halt vehicular traffic. It just turns the WALK sign on when the parallel traffic has a green light. So it doesn't protect you from cars turning right on a red light or oncoming traffic turning left. As a result, pedestrians get hit all the time by traffic while crossing with the WALK sign on. I, myself, jaywalk whenever possible, because it is much safer than trusting that a car won't turn into you at a corner.
It's plausible the WALK sign makes the green light in your direction an imperceptible-to-humans fraction of a second longer, but, being human, I can't detect if that's the case.
Most - but definitely not all - drivers in Saint Paul will yield to a pedestrian at a crossing. It gets weird at times. I have to cross a busy street without a light on my way to and from the corner grocery store. If I so much as glance to the other side of the street when I’m at a crosswalk, cars will slow to a stop for me.
I’m fine with waiting for a break in traffic but some hyper polite drivers won’t even respond to a waved arm ‘Go ahead I’m not in any hurry’ gesture.
https://www.minnpost.com/cityscape/2016/03/st-paul-launches-effort-change-citys-driving-culture-enforcing-crosswalk-laws/
Wow, Minnesota niceness. In Boston we either run over pedestrians crossing without a light, or at least almost do to teach them not to fuck with us.
In my little town, I think it activates the audible - and obnoxiously aggressive - “WAIT WAIT, WAIT” for the vision impaired.
Pete Davidson canceled his trip into space? I don’t think he realizes how a thing that could make him more interesting to women. Why his romantic life would probably just take right off. No more lonely nights with something like that on his CV!
Zvi in his most recent Covidpost mentioned that he couldn't find a video because it was pulled from YouTube. I found the video by putting the YouTube URL into the Wayback Machine - I didn't actually think the Wayback Machine worked on videos, but apparently it does.
http://web.archive.org/web/20220106042900/https://www.youtube.com/watch?v=B8knn6U5Igs
Posted in case anyone here read said Covidpost and is interested in the video, or in case someone here can get it to Zvi/is Zvi (I don't know any way to tell Zvi things other than making a commenter account on one of his blogs), or in case anyone here doesn't know that the Wayback Machine works on YouTube (which is a pretty big deal considering how much stuff YouTube burns).
As a guy who is getting old and no longer in the market, thought I would offer some dating advice since I see guys here sometimes asking for it.
One thing I learned over the course of decades is that women reject you for two main reasons: because you are too in a hurry to get laid or because you are too in a hurry to have a serious relationship. It's easy to mistake a type 1 rejection for a type 2 rejection and vice versa. Maybe that seems obvious, but it didn't seem obvious to me when i was younger so I doubt it seems obvious to every young guy reading this.
An important take-away, I think, is to realize that when a women rejects you it is often for the opposite reason that you imagine. You could learn this from reading Proust, but I'm going to try to keep this shorter than Proust did. (Proust may have been gay, but he understood romantic relationships and sex better than most, straight gay or otherwise.)
If you think a woman rejects you because you aren't attractive enough, that probably isn't the reason. It's more likely because you seem either too interested in getting laid or too interesting in having a long term relationship. Meaning, if you get rejected often, you should change your strategy. If you are trying to get laid, don't. Work on signaling that you are interested in a relationship. OTOH, if you are mainly interested in a relationship, don't. Just try to get laid.
Either way, women are going to figure out what you are really interested in pretty quickly, so don't worry about sending the wrong signals. Do everything you can to counter-signal, because that will send a more balanced signal in the short run. Sending a balanced signal is what most women find attractive.
EDIT: And don't make the mistake of thinking "but THIS WOMAN I am interested in isn't MOST WOMEN". You can't read minds.
> If you think a woman rejects you because you aren't attractive enough, that probably isn't the reason.
Uh, yes it is.
I really don't think it is. While men rate women 1-10, women are mjch more pass-fail regarding men. And there's a low bar to passing: don't be smaller than the woman, smell OK. Even those criteria will be waived if you're funny as fuck.
It also helps to be/appear deeply interested it something other than sex or relationship. Something she can relate to, e.g. some hobby you have in common.
Today's post-secondary institutions are expensive and backward. How can we do better?
Idea for accreditation system based on accrediting students rather than institutions:
1. An accreditation company (nonprofit foundation? PBC?) produces standardized tests to measure student knowledge & abilities. Tests are broken down into a set of mini-tests, and each mini-test tests a small amount of knowledge and ability. The company charges money to an institution whose students take a test, or to a person who takes a test independently, and revenues are used to produce more tests (and to prepare defenses against cheating). Students earn diplomas from the company according to some set of rules to be determined. Much like TripleByte, the accreditation company earns reputation by verifying ability correctly, and it can increase prices as its reputation increases (but if it's a nonprofit or PBC, prices should hopefully not rise without limit.) IMO students should take tests on the same topic twice, at least 8 months apart (to verify knowledge retention and discourage cram-based learning), but that's not my call to make.
2. Educational institutions (e.g. MOOCs) teach students, who pay tuition as usual. The institution chooses what (and how) to teach in each of its courses, and at regular intervals, offers a test from the accreditation company. A test will typically be composed of two to twenty mini-tests chosen according to the material that was taught in the course. The institution earns revenue equal to the difference between tuition fees and test costs. The accreditation company tracks which mini-tests each student has passed; if a student moves between institutions, the gaps and overlaps between courses at the two institutions are tracked exactly. And of course, some people (Sal Kahn?) will offer completely free courses.
This type of system bypasses traditional accreditation boards run by encumbent institutions, which have an incentive to avoid accrediting new entrants. Thus it should produce a competitive online market with low prices, while still giving students a meaningful diploma that they can tout to employers.
Now, surely I'm not the first to think of this, so why hasn't this kind of system become popular?
"Now, surely I'm not the first to think of this, so why hasn't this kind of system become popular?"
They do exist, so I think the question is "why aren't they more popular?" and the short answer is that because if the accreditation is to mean anything and not be a diploma mill, you need some way of checking that the course material is good, the students are qualified, and the diplomas or certificates mean something - and that means you end up re-inventing colleges in some form. E.g. how can you be sure QuadrupleBit graduates are as qualified (this is distinct from capable or clever) as IvyWreathed U graduates? One way is to compare coursework and results. But if both institutions have different forms of assessment and coursework? Well, get QB grads to sit a final exam.
That means you need questions for the exam, a curriculum to cover exam topics, and a place to sit the exam where you can be sure that the QB lot are not all cheating and just have Google open on their computer at home to give them the answers.
Congratulations, you have re-invented the exam hall. And if you need a place to host it, the simplest answer is to get QuadrupleBit, the online certifying company, to hire or rent or find someplace to hold that exam. And since they need this place on a permanent basis and for as many final exams as they're running throughout the year, they may as well have a building for their own use.
And eventually 'online only' becomes 'well we have all these offices and now classroom spaces as well' and it's a new college.
How do you develop a reputation as an accreditation company in the first place?
Universities have history, your QuadrupleByte has nothing. At best you're rated as equivalent to that shitty bootcamp someone's running in the city, whose alumni cannot program their way out of a wet paper bag.
You want is an institution that is a Schelling point for intellectual elites, and does some basic filtering to keep that Schelling point stable. If it is known that clever kids finish CS at $EXPENSIVE_COLLEGE, companies will hire from $EXPENSIVE_COLLEGE. The actual quality of courses at $EXPENSIVE_COLLEGE is an afterthought.
Speaking from my own experience as someone who got a Bachelor's in Computer Science, and then went into Software Engineering - conceivably the very sort of person you're trying to reach - I think that the big missing element is projects.
I credit a great deal of my success in industry to the project-oriented curriculum at the institution I attended, and particularly the class which brought together groups of fifteen people for seven weeks for the closest I got to an actual Software Engineering experience in school. They introduced me to ideas and practices around teamwork and source control which I suppose I could recite, but which would be very difficult to articulate in ways that could convince an observer that I had gotten them in a brief period of time. And if perhaps I could, the ability to verify in a test environment that I had these skills would, I think, hinge on my communication skills to an undue degree - while communication skills are important in my field, I do think this would put more weight on them than is warranted.
It's quite possible that schools do a poor job of actually verifying that a given student has learned the things that a given project is meant to teach. Perhaps they are simply going off some strong prior - whether it's the latest in education research, the dead reckoning of their most experienced faculty, or the dean's latest interpretation of their star signs (though hopefully accreditation puts some kind of lower bound on how bad that methodology can be) - that tells them that given a project which roughly fulfills these specifications, the students should have learned these skills, and if that's true of only 90% of the students and even those learn on average 90% of the skills, there's the opportunity to instill it with further overlap and maybe we'll catch it before they graduate, if they missed out on a foundational skill in their first year that a fourth-year project depends on in the same way that it depends on the students being functionally literate in the university's language.
But I don't think there's a way to identify the kind of learning we want to happen from projects in an environment where the assessments and the teaching are so decoupled. Tests can be fine for getting a student to apply specific knowledge in a narrow sandbox, but they can't really capture how well a student works at something - over days, over weeks, consulting resources (because consulting resources is a skill, which may well be more important than what most given resources have to teach; as a Software Engineer much of my job relies not on me knowing what needs to be done before but on being able to find, interpret, and apply the relevant sources).
I think this gets much worse for fields where there is very definitely not a singular concise correct answer at the end of an equation or as the result of a suite of automated tests. Which is a shame, because one big thing we want from knowledge workers is the ability to work somewhat autonomously, on broad problems with ill-defined endpoints, often collaboratively, and in novel situations. The stuff that doesn't fit this description - well, that'll get snapped up by automation sooner than later.
You'd be turning universities into teach-to-the-test cram schools.
Most of what you'd expect to learn at the university level can't be adequately tested in an examination, and for the stuff that _is_, it would be reasonably easy to cram (e.g. there's only, like, twelve possible classes of exam question for Special Relativity so let's just study them all).
Since universities already have accreditation, I would expect them to reject a new system completely.
But when I was going to my university Engineering program, I crammed a lot because my teachers were so bad that I felt that *actually understanding* the course material was out of reach for me. This is completely different than my high-school experience, in which I *never* crammed.
Uncharitable of you to ignore the anti-cramming measure that I proposed. Cramming is a short-term trick; if you have two tests spaced >8 months apart, you are certainly better off learning the material properly.
No, you just cram twice.
It is impossible to "learn the material" for the long term if it has no day-to-day practical use. The material would need to be integrated into practical projects executed during that 8 month interval.
Not sure why you think cramming twice is "better" than learning the material properly. I disagree about the impossibility of learning things without day-to-day practical use; I learned such things throughout K-12 without cramming.
Current education/accreditation system is a mess of all kinds of signals. It is difficult to improve, because sometimes it is supposed to suck.
For example, imagine that you create a parallel educational system that is neither better nor worse than the traditional one, but it is 10x cheaper. Would it be popular? No, because using the new system would signal that you are poor, and so are most people you know. Rich people would avoid it; and they would also have an incentive to pretend that the new system is worse.
Or imagine a new educational system that is just as good, only 10x less frustrating for students. Then employers would avoid hiring people from that system, because they would suspect that such employees would have low frustration tolerance and would soon quit after the normal amount of abuse at workplace.
Or a new system where kids learn more easily because somehow all lessons are magically easy to understand? The employers would suspect that the kids are actually less smart then they seem, and will fail when facing a novel situation.
In other words, trying to make education better is like trying to make a marathon shorter -- the people who usually run marathons will reject the idea. Education is inefficient, frustrating, and gives unfair advantage to rich people... which is exactly the point. It prepares you perfectly for your future workplace.
(If you are interested in knowledge for knowledge's sake, then of course, Khan Academy is the way.)
> No, because using the new system would signal that you are poor
I'm pretty sure most employers aren't judging you based on how rich your parents are, with some obvious exceptions (Harvard) in which wealth isn't the only thing being signaled.
> employers would avoid hiring people from that system, because they would suspect that such employees would have low frustration tolerance
As a (very) small-fry CTO myself, I want knowledge and skill and more than I want frustration tolerance. I guess some companies want that, but other companies want other things. (and why would poor people have worse frustration tolerance?)
"(and why would poor people have worse frustration tolerance?)"
It's not that poor people would have worse frustration tolerance, it's that (whether rich or poor) coming out of an education system where learning was effortless, the teachers wee excellent, it was easy to learn, everything was ready for you when you were ready to take the next step, etc. is not good preparation for a workplace where it's "dunno the answer to that, the guy who does know the answer is out for two weeks so you have to wait that long, there is paperwork with the answer on it someplace but you'll have to search for it and nobody knows where to start, and there isn't a clear-cut answer at the end, just a 'good enough' and besides the boss/boss's boss/client is going to change their mind three times about what they want anyway".
The elephant in the room is that if your parents are rich they're probably higher in IQ and/or cultural capital, which makes you higher in IQ and/or cultural capital, both of which are hugely desirable traits in new hires.
(Ah, please take the previous comment with a grain of salt; it was exaggerated for artistic purposes. This comment reflects my beliefs literally.)
There are two ways how wealth impacts knowledge/skills that come to my mind immediately (and there are probably more):
First, rich people can spend more money on education-related expenses, and I think they have more free time on average (no need to keep two jobs; can save some time by spending money on something). So if we imagine two kids with equal intelligence, talent, chatacter traits, and hobbies; learning e.g. computer science in exactly the same classroom with the same teacher using the same curriculum; but one comes from a poor family and the other comes from a rich family, I would still expect different results, given the following:
The rich kid will have a better computer at home, more time to use it, no problem with paying for a course or a tutor if necessary. The poor kid will feel lucky to have a computer at all, will be limited to free resources, and will probably spend some time helping their family (working, taking care of younger siblings). -- Therefore, at the end of the year I would expect the rich kid to know more, on average.
Second, when people think about the quality of school, they usually think about teachers, curriculum, didactic tools, and whatever... but a crucial and often ignored factor is classmates. Kids inspire each other and learn from each other a lot. Or they can prevent each other from learning by disrupting the lessons. Your ability to choose a school with better classmates is often limited by money. With the same curriculum and same quality of teachers, I would expect rich kids to get better outcome at school, simply because they are not surrounded by classmates from dysfunctional families etc.
They differences may seem small, but they add together, and their effects compound. With exactly the same curriculum, I would expect rich kids to get better results, on average, even if we control for intelligence and other traits.
Therefore, from the company perspective, a rich kid seems a better bet, ceteris paribus. (The only disadvantage is that the rich kid will probably expect a higher salary.) Therefore, if you are - or pretend to be - a rich kid, you probably do not want to signal that you attended the school for poor kids... or the cheap school.
Re: frustration tolerance -- you want some basic level of it, like people who won't give up and quit after the first problem.
And speaking as someone who has come out of the 'practical training, non-university, on a ladder of accreditation' route and worked with people who went the university route, there is a difference. Not even so much in practical skills, depending what course they did the university people can be as good and up on those, but the entire shape of the learning experience, the environment, the content.
The practical-oriented really was teaching to the exam, telling you what you needed to know to do the tasks, but nothing extra. It was to get you qualified and out into some kind of paying job as fast as possible. If you wanted extra, you could then go up the ladder to the university. People from the university just had an entirely different experience, and yes they did indeed seem to have a more rounded education, a better understanding of the subject, and just that familiarity with how academia works that is hard to put into words but you know it when you don't have it and are trying to engage on the academic level.
https://www.qqi.ie/what-we-do/qqi-awards/certifying-qqi-awards-provider
https://www.youtube.com/watch?v=qK15HlhDbo4
Hey, you seem to know your shit. Could really use an impartial set of eyes on something...would you be willing to read a few pages and give an opinion?
It seems like this sort of info about a student's mastery and retention of material would be taken seriously by grad programs or employers in STEM fields -- have my doubts about other fields though.
The step you're missing is to make big companies accept your accreditation as a valid measure of a prospective employee's worth. Companies are often at least as interested in proof of conscientiousness as in intelligence, and often not at all interested in subject matter mastery.
At higher levels, the top institutions sell networking as much as anything else, and that's a very sticky equilibrium - one of the main benefit of a Harvard grad is that they know other Harvard grads, and that's valuable because all the top places hire Harvard grads, so their network makes them valuable enough for all the top places to want to hire them
Yes, and this is why I suggested that the price of the service would be correlated to its reputation: big companies accepting it = reputation. I guess it would need somebody with deep pockets to help it survive the early low-reputation phase.
Obviously, my proposal is not intended for people who have the means to make it into (and through) Harvard. I went to an ordinary university, and the amount of networking I did there was basically zero. (edit: except the internship program - the job I got out of that lasted 7 years.)
What's the textbook example of how to tactically use tanks in warfare? I'm thinking a WW2 battle or something were tanks saved the day, did lots of the special tanks things that can't be done by artillery or motorized infantry or whatever, and then some colonel wrote the book on it.
I'm hearing a lot from armchair generals on how to not use tanks ("Don't use them against other tanks", "Don't use them in urban areas", "Don't use them without infantry support" etc.), but I don't hear much about how they are supposed to be used, thus my question.
If you want to ague that tanks are obsolete in modern warfare this is your spot as well I guess.
If your enemy has a tank (or worse, a bunch of them) somewhere, and you don't have anti-tank forces nearby, then the tank can destroy whatever stuff you put through that area, unless they are well-hidden or well-fortified. And it's hard to move while being well-hidden or well-fortified, so if the tank got there first you are outta luck. Of course, you are aware of that and will not move into that area, but that means you can't have anything there.
That makes tanks basically movable walls (the size of the wall is the range of the tank of course, not the physical size of the tank). You put them where you don't want the other guy to go through - an important version of that is to cut off a force from their rear lines and force them to surrender.
The reason they are good at that is that they are armored pretty well and can fire on the move, which makes anti-tank platforms far more limited than say anti-artillery platforms (it's easier to sneak up on a self-propelled gun or destroy it after it fires, than to do the same on a tank).
So why not simply use fortification then? Concrete slabs can be moved by trucks, isn't that enough to make an area hard to move into?
First, a tank is a km-length wall that can move operationally at speeds of tens of km/h. A kilometer of wall is a lot of wall to place, or to move to a more relevant location when needed (and you want to fill in or exploit breaches quickly, which makes moving to relevant locations very important).
Second, its much harder to destroy a tank (that is maneuvering through tank-friendly terrain, not that is driving through hostile city streets) than to breach a wall, because the tank can shoot at you then move into hiding and possibly call for assistance, and a concrete wall can't.
Depends a lot on the period. You can get a good spiel of the WW1 retrospective & pre-WW2 expectations (which held up at least for early WW2) by reading "Achtung - Panzer" by Guderian, but the key doctrinal takeaway I remember would be:
-Use them in large concentrations (i.e: not in piecemeal), with motorized infantry to support and exploit their breakthrough (and absolutely not "supported" by leg infantry, which can't match their speed, and slow them down to get pummeled by artillery fire, as in WW1, or WW2 french doctrine).
-Best way to stop them is another tank (that was before air support got precise enough to pose a threat)
Supposedly De Gaulle reached similar conclusions in "Vers l'armée de métier" (or was it "La France et son armée"?), but I haven't got around to read it yet, so I can't comment.
What you're really asking for is how to do combined-arms warfare. There are very few military problems where the solution is "send tanks, just tanks", but many where tanks are a useful or vital part of the solution.
The classic use case for tanks, and *maybe* justifying a pure-tank force, is exploiting a breakthrough in mobile warfare. Not making the breakthrough itself; you'll almost certainly want artillery and infantry to help with that. But if you've broken through the enemy's defenses, you want to move fast and break things behind the lines before they can offer a coherent response, which will involve meeting engagements with elements of their incoherent response, and that's something tanks are really good at.
And for much the same reason, tanks are good at mounting rapid counterattacks in the face of an enemy breakthrough.
There are some good recent-ish examples (but not pure-tank) in operation Desert Storm, and in the Arab-Israeli wars.
https://en.wikipedia.org/wiki/Battle_of_Norfolk
https://en.wikipedia.org/wiki/Battle_of_Medina_Ridge
https://en.wikipedia.org/wiki/Battle_of_73_Easting
https://en.wikipedia.org/wiki/Valley_of_Tears
Sure, I get that you want to do combined arms warfare. But what is the role of tanks in combined arms warfare?
Why are they better than e.g. mechanized infantry at exploiting breakthroughs? Are they better armored so that they have an easier time driving past pockets of resistance? (If so, can't we just slap more armor on an APC to achieve something similar?)
What does exploiting a breakthrough actually entail in practice? Do you hunt down enemy artillery and C&C? Do you try to get the enemy logistics (by firing randomly at trucks)? Do you attack the enemy in the rear? Do you just drive as fast as you can against Berlin/Bagdad and hope to create as much chaos as possible on the way? I guess tanks need (mechanized) infantry to create a famous WW2/style pocket, but it's good to have the tanks in the front while doing so?
The Iraq tank battles just seem to be coalition tanks driving through and obliterating technologically inferior Iraqy tanks. Couldn't this have been done by infantry or air power? Valley of Tears looks interesting but the Wikipedia article is hard to parse, I'll look into it.
Mechanized or even motorized infantry can support breakthrough operations, and in some cases (e.g. when the enemy only has leg infantry), can do the whole thing.
But infantry, even mechanized, has to dismount to fight. Otherwise it's not infantry, it's just an inferior form of armor handicapped by having to haul a bunch of useless people around - and no, their shooting assault rifles out of firing ports isn't useful enough to be worth the bother.
And dismounting, fighting even a skirmish at a walking pace, and then remounting, takes time and costs tempo. Sometimes it's necessary, e.g. to clear enemy infantry blocking your advance in close terrain, but if at all possible in a breakthrough operation you want to defeat at least minor blocking forces on the move. For that, you want tanks.
As for what you do in a breakthrough, part of it trying to overrun C3I facilities, logistics nodes (not random supply dumps, but supply depots, railheads, critical road junctions etc), and artillery. I'd put it in that order of importance, but it's debatable. The other part of it is maneuvering to block enemy lines of retreat and reinforcement.
And pretty much all of it, is creating in the enemy's front-line troops and their leaders the firm perception that they don't know what the hell is going on in their rear, but it's really bad and if they don't run away *right now* they'll never have the chance. Then your breakthrough forces can ambush and kill them as they flee.
As for using infantry to "drive through and obliterate technologically inferior tanks", no. Literally no, because infantry doesn't drive, it walks. And if you're thinking they're going to drive through in their technologically superior infantry fighting vehicles, maybe but now the "infantry" isn't doing anything and even technologically inferior tanks are carrying bigger guns with longer range and heavier armor because the aren't carrying around useless infantry. Maybe you've got enough of a technological edge for that, but it's still fighting with a handicap.
If you're talking about infantry advancing against tanks on foot, no, infantry can't advance against fire like that. First, because infantry survives under fire by emulating hobbits - small, nigh-invisible, and living in holes in the ground. And second, because infantry can't fire on the move.
Even a "technologically inferior" tanks is in this context a machine gun on a gyrostabilized mount with a magnifying and if needed night-vision optical sight and nigh-infinite ammunition, with a gunner who is basically immune to suppressive fire, and capable of firing accurately while retreating faster than infantry can advance. Think first-person shooter video game - completely unrealistic at duplicating the experience of a *soldier* in combat, but pretty good for a tank gunner. Oh, and he has a big-ass cannon if he needs it.
The infantryman, is a very not-machine-gun-proof target whose attempt to advance largely voids his concealment and subterranean-ness, and sure maybe he's carrying a high-tech missile that could destroy tanks if he weren't trying to advance, but since he is it's just ballast slowing him down.
If the enemy brought *only* tanks, and parked them too close to a town or treeline or whatnot, you could imagine your infantry sneakily infiltrating into firing positions. But the enemy probably has some infantry of his own, and he's got that deployed to cover all those firing positions you were trying to sneak into. His infantry doesn't have to defeat your infantry, it just has to force it to reveal itself prematurely so that the enemy can put heavy firepower from his tanks - or better yet artillery - onto it.
This clarified a lot. Thanks!
Wait, that first line makes no sense to me, I thought (a subset of) tanks are explicitly designed to be used against other tanks?
More generally, I think tanks in open terrain beat infantry without tanks; combined arms is generally very important, of course, and the tanks can't be totally unsupported, but outside of cities it's hard to sneak up on a tank. Air superiority is of course the ultimate trump card, tanks don't beat bombers.
Historically, I think Germany's Blitzkreig of France is the ur-example of the value of light tanks and motorized infantry. The biggest advantage of tanks over artillery is their mobility, so if you want examples of things only tanks can do look for maneuver warfare.
It's a misunderstanding/oversimplification of American armored warfare doctrine during WW2, which emphasized tank destroyers (typically in battalion-sized units attached to infantry divisions) as a defensive counter to massed enemy armored offensives. The misunderstanding comes in reading this as saying that *only* TDs should fight tanks. Tanks were seen in this doctrine as being perfectly capable of fighting other tanks. The actual point of the TD emphasis was that the design differences optimized TDs for a defensive response role, while tanks were optimized for breakthrough and exploitation: American TDs of mid-to-late WW2 were somewhat cheaper, a bit faster, and mounted heavier guns than tanks of the same generation, at the cost of substantially lighter armor, making them better at responding to enemy offensives and fighting defensively from cover with infantry and artillery support, but much less survivable on the offensive. This, TDs were the first-line response in support of infantry for enemy armored offensives, freeing up tanks for things they did better than TDs.
When the doctrine was first developed, the intent was for TD to be a lot cheaper than tanks, initially conceived as light infantry-support antitank guns towed or mounted on jeeps or light trucks. This way, you could have TDs everywhere you needed them for a fraction of the cost of tanks. But by mid-war, TD designs got more capable, more tank-like, and correspondingly more expensive so an M4 Sherman tank wound up being only a little more expensive than an M10 tank destroyer.
I think tank destroyers technically aren't tanks.
What would have happened if the Germany had done the invasion of France with 50% less tanks and more motorized infantry instead? Would it have been worse, and if so why?
This is a random place to post this but I asked a question like this before on an open thread and Scott responded. Any help appreciated from anyone with relevant medical industry knowledge.
I am certain I have ADHD. It hugely affects my job performance and I'm constantly worried about getting fired. I'm hoping to get prescribed adderall. I'm wondering what the chances are with my current planned process:
-I have made an appointment with a telehealth psychiatrist through some large online group.
-This site specifically says they themselves do not prescribe drugs like Xanax and adderall, but that if they think it's necessary they can fax your PCP to have them make a prescription.
-I have made an appointment to see someone as a PCP next week, two days before the psychiatry appointment.
-This appointment is my first time meeting them and I said I wanted to talk about ADHD in an office visit
-But they are a Nurse Practitioner, so I have no idea if they're allowed to prescribe anything/more hesitant.
Checkout Ahead (helloahead.com). The PNP I work with prescribed me Adderall. No contact with a PCP (which I don't have). They also manage my anxiety meds.
Some NPs are not allowed to prescribe certain drugs. It should be fine for you to call and clarify with them, asking them if you come with a referral will they be able to prescribe for you. Congrats on taking steps towards treatment.
The authors of Meta-analysis studies should be required to present a table listing the included studies and excluded studies along with the exclusion criteria that each of the excluded studies failed to meet. That is all.
Actually, I lied. That is not all.
They should also list their reasoning behind their inclusion/exclusion criteria!
Just finished reading two meta-analyses with differing conclusions, and I've come to the conclusion that I'm no closer to the truth than I was before this exercise. I don't know if one or both of the authors are trying to pull a fast one. And without being able to look at the studies included in the meta, I cannot judge the validity of either of the metas.
Wow, that's irritating. They put "Meta" in their title and become arbiters who can't be challenged. (Sort of like freakin' Facebook changing it's name.). What field are these articles in? I've read a dozen or so psychology and neurology meta-analyses recently and they all really spelled out their criteria for judging study quality. Many did not list all the articles they considered, but in many cases that seemed reasonable -- there were thousands. Maybe writers of metas should be required to publish that info in an appendix, though.
It was two meta-analyses of long-term COVID. Actually, they didn't have opposite conclusions—just different conclusions. They spelled out their criteria, but it was what wasn't mentioned that made me unable to compare the two. The first M-A's criteria was only studies published in the English language. The other one didn't mention that criteria, so I'm left wondering if they had a more diverse pool of international studies, and that that may account for the difference in conclusions.
Also the first study had a criteria that each study have a minimum sample size of 30. The other study had higher sample size criteria. But I'm left wondering why the first study was OK with 30. Seems too small to be statistically valid (?). And how many of those studies in the first one had sample sizes less than 100? Aarrggghhhh.
About study size: the smaller the study, the larger the effect size has to be to capture it. If I wanted to find out whether capybaras weighed more than hamsters, and compared 2 groups of 15 randomly selected members of each species, I would definitely find a statistically significant difference in weight. If a small study finds a statistically significant difference, then you can trust the result (unless there is some other problem with the study design -- such as wrong statistical methods done, groups compared were different in important ways beyond the one you were checking for). However, if small study finds no effect that may be because effect doesn't exist or just because study was too small to have enough power to capture the effect.
About getting info about long Covid: Epidemiologist Jetelina, on Substack, just put up a series of posts on the subject. So this is sort of her unofficial meta-analysis. I trust her, mostly. She’s smart and thorough and seems to have no ax to grind.
Ha! I got kicked off her Facebook group. It was early in the pandemic and I was questioning the some of the consensus wisdom of epidemiological theory. I thought I was polite, but I got exiled from that FB group. OTOH, I can be rather outspoken so maybe I deserved it. ;-)
Wow. Well, she does have a bit of a kindergarten teacher quality to her -- sort of pathologically nice and hyperconventional. If you can still stand to have anything further to do with her, her Substack blog seems pretty good to me -- informative, and politics-free.
As for getting kicked off things -- I'm a member of the too-rude-to-remain club too. Spent a year on Twitter, driven crazy by snark and trolls even though I only followed science writers. One day a red-state male troll started dropping turds on a thread about some technical virus thing, so I came back with the term most likely to offend his demographic: cocksucker. Now I'm banned from Twitter and glad of it. Heh.
Should I be concerned about endocrine disruptors from microwaveable plastic bags of vegetables e.g. steamfresh?
McDonalds leaves Russia?
Pravda web site says good riddance.
‘ McDonald's sells "food” that is absolutely impervious to rot and decay. You can buy one of their hamburgers, put in on a shelf in your living room and just leave it there. After a year the burger will still look and smell the same. None of the rodents that you unwittingly share your house with will have deigned to touch it. Nor will any insect, no fly no wasp, nothing. Even bacteria will stay away from McDonald's products. This will give you an idea of the quality of American fast food. KFC specializes in products made from bio-engineered, hormone and antibiotics-fed chickens growing so fast they never learn to walk. The meat from such creatures will probably help accelerate your transition from "cisgender” to anything in the LGBTQ spectrum, whether you want it or not. Starbuck's specializes in something it dares to call coffee but that anyone who really knows and likes coffee will shun.’
Well, if you skip the LGBT thing, are they wrong?
I never felt "wow this is quality food" while eating McD, it always tastes like some kind of space colony faux food that's supposed to remind me of what they used to eat back on Earth, but fails at it.
No. It’s not exactly fine dining.
> None of the rodents that you unwittingly share your house with will have deigned to touch it
McDonald's, KFC and especially Starbucks may be terrible, but I'm glad I live in a country where it's not simply taken for granted that every house has rodents.
It’s a classic case of ‘those grapes were sour anyway’
https://dailyhive.com/vancouver/mcdonalds-russia-rebrand-uncle-vanya
"Degenerate Western food makes you trans" (presumably in contrast to virile, natural Russian foodstuffs) is a take so cartoonishly anti-Woke I didn't think pravda.ru would actually publish it.
That sort of stuff is aimed at Poland, Czech Republic etc. Russia is the defender of traditional values. It’s an on going thing with Pravda. Trying to appeal to East European NATO countries. I know. It’s all so very strange.
Rodents? This situation in Russia is grimmer than I would have thought.
Probably just the odd capybara in the pantry here and there.
The grammar is pretty good in this one. I’m guessing it was written in English by a human. The machine translated stuff usually makes a hash of idioms. The underlying differences in grammars show through too.
What percentage of people who regularly attend KKK meetings probably have diagnosable mental illnesses?
Are regular KKK meetings actually a thing? I was under the impression that the KKK doesn't meaningfully exist any more. This isn't Robert Byrd's day.
The ADL (whose incentives certainly run towards maximising rather than minimising the extent of Klan activity) most recently https://www.adl.org/education/resources/reports/state-of-the-kkk reports the existence of thirty groups claiming to be the Klan, but most of them are just a handful of people and they tend to pop in and out of existence rather rapidly as people lose interest.
The biggest remaining group seems to be the Loyal White Knights https://www.adl.org/resources/backgrounders/loyal-white-knights-of-the-ku-klux-klan who apparently have about a hundred members. But they seem to be pretty geographically spread out, so I have my doubts whether they have anything you could call a "regular meeting".
Based on the abstract, this study looks like it attempts to answer that question:
https://journals.sagepub.com/doi/10.1177/0002764219831746
Sadly, the actual results seem to only be in the article body, which is behind a paywall. Unless someone here has institutional access to Sagepub or JSTOR and feels like reading it and summarizing for us.
It appears to be open-access on ResearchGate: https://www.researchgate.net/publication/331428143_The_Problem_of_Overgeneralization_The_Case_of_Mental_Health_Problems_and_US_Violent_White_Supremacists
I'm too much of a basket-weaving major to confidently give a reading of this, but here's some cliff notes until someone more qualified comes along:
- Sample size of 44 ex-extremists undergoing long interviews
- 57% reported mental health problems before or during extremist involvement
- 62% attempted or seriously considered suicide
- 73% reported substance abuse issues (64% prior to age 16)
- 59% reported family history of mental health issues
The conclusion says "so far the field says the rate is no higher than average, but we found it's actually pretty common".
I don't like that study because it focuses on "former VIOLENT U.S. White supremacists" (emphasis mine). I'd rather see the data for the KKK members who don't actually act on their racist beliefs by attacking nonwhites (they probably form the group's majority).
My hypothesis is that, if you're so racist that you're actually willing to go to KKK meetings, you're probably mentally ill.
Could work the other way round, too: if you're so mentally ill that the only group that will tolerate you is the KKK
What would those people’s diagnosis be?
I suspect a lot of depression, antisocial personality disorder, and PTSD.
Seems like a fair comparison group would be other extremists, both left, right, & totally disaffiliated -- like test the people in Anonymous (if only they weren't all anonymous!). It may be that unhappy and desperate people are drawn to extremism, and/or extreme views create desperation. If some of the wilder apocalyptic theories were true, suicide might be a rational choice.
https://www.youtube.com/watch?v=rfjZgdQsr6s&ab_channel=RogerJamesHamilton
From 2020-- predicts the invasion of Ukraine. Assumes that dictatorships can play a long game, and win. That part might be wrong.
Notably, he says that Alexsandr Dugin ("Putin's brain") wrote a book in 1997 called "Foundations of Geopolitics: the Geopolitical Future of Russia" which is Putin's playbook and can be used to understand and predict Putin's moves.
The book's 40-year plan:
Step 1. Invade Georgia
Step 2. Annex Crimea and control Ukraine
Step 3. Separate Great Britain from Europe (Brexit?)
Step 4. Chaos: sow division in Britain and the US
Step 5. Create "Eurasia" which (based on the map) looks like basically Russia surrounded by "buffer states", with China "divided and in turmoil", and Japan and India as allies of Russia (I note that while Japan voted to condemn Russia's invasion, India was neutral and is now setting up a special payment system to avoid commerce interruptions caused by sanctions. Evidently Russia bagged China as a Russian ally instead of Japan. My impression is that while China isn't completely sold on the invasion yet, it is spiritually siding with Russia and the reason it isn't doing more to help Russia is that it fears "secondary sanctions".)
Steps 3 and 4 involve using the 3 Ds, Deception, Destabilization and Disinformation, to create internal divisions in Britain and the U.S.; internal divisions in the U.S. are to make the U.S. more isolatist and distant from Europe (hence Putin's support for Trump, who in turn pulled out of multiple international treaties). Also the book calls for a "Continental Russia-Islamist alliance [as] the foundation of anti-Atlanticist strategy" (hence Russia's ties to Iran & Syria)
He also has a video about China's master plan: https://www.youtube.com/watch?v=WaAOss6W1u0 - this video begins by telling me that China already beat the U.S. on the metric of "GDP by PPP (purchasing power parity)" in 2014, though note that the per-capita *incomes* of Chinese people by PPP are just over one-fourth the incomes of U.S. people. Which itself is probably part of the plan... to sacrifice income for more GDP and more power on the world stage. China seeks world domination, and on the economic front, they seem to be ahead of schedule.
Oh, now that makes me wonder if some Chinese policy wonk read that book and decided that instead of letting Russia be Ruler of Eurasia, with Japan as an ally and China internally divided and in turmoil, it would be smarter to cosy up to Russia, sell themselves as an ally, and remain a major, non-conflicted, partner if or when Eurasia is a thing that happens.
That would explain (to me) why China is lining up with Russia right now, instead of sitting back and seeing how things play out. Even if Russia manages to shoot itself in both feet with Ukraine, China can still be the "I'm your friend, see how I supported you?" partner and be in a good position to gather up the fragments from the fall-out if Russia instead starts falling apart with internal conflict.
Meanwhile, in this video he predicted a near-term financial crisis which didn't happen (we just got some inflation, and if there's a crisis now I think it'll be triggered by Russia) - I suspect that he doesn't understand macroeconomics well enough (which is not unusual; my impression is that even economists themselves have multiple incompatible models that make different predictions): https://www.youtube.com/watch?v=EYOVoQT2yQg
In his otherwise reasonably accurate 2020 prediction video about vaccines, he characterizes what sounds like it should have been a crisis in 2021 as "Massive wave of bankruptcies, unemployment rise again, and debt bubbles bursting": https://www.youtube.com/watch?v=yahfx_JIihQ ... looks like he overweights the importance of debt and QE https://www.youtube.com/watch?v=nUOVRo_EIrE ... whereas my model is closer to market monetarism: I do not find debt to be important except indirectly, and I expected a "market adjustment" but no crash.
I have a friend who's in the market for a dating coach. He's in the general Astral Codex Ten audience demographics - about 30 years old, tech professional, generally liberal. He has been unable to find a good option with experience working in those demographics. Does anyone have a person to refer him to?
I would advise finding a local one. Dating advice changes quite significantly depending on country or even city. A local coach knows local quirks and also good spots/locations to meet people.
Also check the coaches age. A lot of them are mid 20s or even younger. The game works differently for 30+. A 22 year old won't give you useful advice for your age bracket.
If your friend is into online dating I might give some pointers on optimizing his profile. Been doing a lot of work on this topic for a machine learning project I'm working on.
What are your recommendations on profiles?
In general the youtube channel "School of Attraction" has a lot of online dating advice. Also search for "Reddit Tinder Guide". There are multiple good ones.
More than that we need to go into the specifics of the actual profile. I'm pretty new at substack and don't know if there is any kind of personal message. But if you want you can contact me and i can give you/your friend specific advice about the profile(s)
I've heard good things about Relationship Hero (disclaimer: from the person who founded it).
I’m going to assume that date coaching is a thing now. It wasn’t when I was single.
Is your friend very shy? That would make it harder. If that is part of the problem, he could try getting regular exercise. Cardio and weight training relieve anxiety and help with self confidence.
If he is up to it, being able to dance a bit would give him a chance to meet potential partners.
My friend is not shy, and has been in the dance scene for years to no avail. I think he knows he has a problem that needs personal coaching attention.
I grew up in Russia (but left to the US as a teenager, on my father's H1B visa). This means I still know a number of people in Russia who are disproportionately techy, and statistically I'd expect some of them to be interested in no longer being in Russia right now. (Some will have left already, some will want to stay no matter what.) Is there a more effective way to look for jobs that might sponsor them than "ask your company if they'll sponsor a visa, ask your friends to ask their companies if they'll sponsor a visa, etc."? Also, consider yourself asked :)
I work for a biggish consulting company and our policy is to sponsor advanced degree holders (MBAs, PhDs, MDs etc.) coming in at the consultant level but not for analysts who usually come in at the undergrad (BS/BA) level. That's for the U.S., not sure what the policy is in our international offices. I've also asked about expanding sponsorship to further down the ladder and it's definitely being considered but I don't think that policy is likely to change at least in the short term.
If you know anyone who might be interested they can reach me for more info at gbz.uraarffrl@tznvy.pbz (rot13)
Note that I asked this again lower down in the advert post, and they (Dave92f1) said that they're willing in principle but haven't done it in practice, and also that they'd consider remote.
Also, if you want to send a resume my way, my company (http://www.cyberoptics.com/) is looking for at least one software person and at least in principle willing to sponsor people; I'm lastname at gmail.
Scott, thank you for helping set me on the path towards effective altruism. Your writing was deeply influential to me in high school and early college, and I think it was a really big part of why I got into EA (where I get a lot of self-esteem from these days). Since I think its relevant, I'm a senior software engineer at a FAANG and I donate around 30% of my pre-tax income, so include some fraction of that in your total impact!
[Context, I'm reading through 'What got you here won't get you there'. It recommends thanking the top 25 folks most influential in your professional life. Scott handily qualifies for me.]
Just want to say that donation of 30% of pre-tax income is pretty darn impressive.
Looks like you are really walking the walk.
Good for you.
Thanks! I really appreciate the encouragement; I don't have many EAs in my social circle so the positive reinforcement is rare and valued :)
You're welcome!
Seconded. I'm in the middle of a career switch from lucrative but soulless software dev to medical bioinformatics that's socially useful, very interesting, very frustrating, and paying next to nothing.
Scott's writing, especially UNSONG, has been one of the main things that pushed me to finally do it and ruin/fix my life.
Congrats! I'm almost surprised you managed to find a software related job with poor pay :)
A Harvard Business Review reports that 'study after study puts the failure rate of mergers and acquisitions somewhere between 70% and 90%' (https://hbr.org/2011/03/the-big-idea-the-new-ma-playbook).
Has anyone seen studies citing a failure rate of one country invading another country? It seems like this might help forecasters take the outside view of Russia-Ukraine. I can't figure out the right search terms.
I expect a wider range than the 70%-90% for M&A failure. One reason is the difficulty of identifying invasions due to proxy warfare. (Should the Bay of Pigs landing by anti-Castro Cuban exiles be classified as the US invading Cuba by proxy, or an abortive civil war?) Another reason is the difficulty of defining failure, since political goals are harder to evaluate than corporate profits/losses.
The rates probably vary by technological era, as new weapons make offense or defense easier.
What are your favorite pieces of fiction from Scott?
Personally -- even though it seems crazy to pick this one since I'm sure it was low-effort compared to many other stories -- I would have to say [and I show you how deep the rabbit hole goes](https://slatestarcodex.com/2015/06/02/and-i-show-you-how-deep-the-rabbit-hole-goes/). I think the ending is something like the greatest thing ever. Also very fond of [Sort by Controversial](https://slatestarcodex.com/2018/10/30/sort-by-controversial/) and of course UNSONG.
Looking at the [#fiction tag on SlateStarCodex](https://slatestarcodex.com/tag/fiction/), I also realize that there are some I haven't even read yet.
I have a theory that Scott pseudonymously wrote an alchemical allegory disguised as a bad Harry Potter fanfic, but nobody got the joke, so he had to write a whole essay explaining it. If true, that is my favourite.
"Chametz" was fun! And I very much enjoyed "Ars Longa, Vita Brevis"
https://slatestarcodex.com/2017/04/13/chametz/
https://slatestarcodex.com/2017/11/09/ars-longa-vita-brevis/
Was just thinking about World War II is Not Realistic, not sure if that counts as fiction per se but it's got to be in my top 5 or so http://web.archive.org/web/20160908033911/http://squid314.livejournal.com/275614.html
Reading "Sort by Controversial" and the comments thereof made me learn the origin of the phrase "not by one iota", which is now one of my favorite facts. Is this what people learn in Sunday School? Why did no one tell me?
> The First Council of Nicaea in 325 debated the terms homoousios and homoiousios. The word homoousios means "same substance", whereas the word homoiousios means "similar substance". The council affirmed the Father, Son, and Holy Spirit (Godhead) are of the homoousious (same substance). This is the source of the English idiom "differ not by one iota." Note that the words homoousios and homoiousios differ only by one 'i' (or the Greek letter iota). Thus, to say two things differ not one iota, is to say that they are the same substance.
https://ksuweb.kennesaw.edu/~tkeene/ogtHomoousios&Homoiousios.htm
Except, is that etymology true? The online dictionaries I check don't mention it as an origin for that meaning and instead say it's from iota being the smallest letter and therefore almost insignificant. (The do mention that 'jot' derives from this, as iota also is transcribed as jota.) And Wiktionary quotes it as being from the New Testament "until heaven and earth pass away, not an iota, not a dot, will pass from the Law". So that predates Nicaea (I presume?).
So, uh, is this a scissor statement? Discuss at your own risk.
One piece of evidence against this etymology - Hebrew has an expression "On the tip of a yod", which means "decided based on a really tiny difference between two otherwise equal things", which feels like the same expression. Yod is the Hebrew alphabet version of iota and is just a really small letter (אבגדהוזחטי - yod is the little one on the left). Iota is also a pretty small letter. So if "on the tip of a yod" and "by one iota" have the same origin it was probably from the graphics of it rather than some complex greek spelling.
Something I thought about just now - Hebrew narrowly missed a chance to have a letter called 'Yoda'
I just followed this link to reread the pills story and it struck me that of course William van Orange
I had missed this one.