Open Thread 142

This is the bi-weekly visible open thread (there are also hidden open threads twice a week you can reach through the Open Thread tab on the top of the page). Post about anything you want, but please try to avoid hot-button political and social topics. You can also talk at the SSC subreddit or the SSC Discord server – and also check out the SSC Podcast. Also:

1. Okay, now the adversarial collaboration contest is actually over. Thanks to everyone who sent me their entries. I’m going to start posting them sometime over the next few weeks. If you haven’t sent it yet, I may still post yours if you can get it in before I’m done posting everyone else’s, so hurry up!

2. New mistake on the Mistakes Page. I had previously argued technological progress wasn’t slowing down; based on the work of Tyler Cowen and Ben Southwood I now think it is; my previous position was mistaken.

3. Some good responses to my posts on therapy from various therapists and therapy patients; unless I get around to collecting them all in one place I’ll just link @QiaochuYuan on Twitter.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

406 Responses to Open Thread 142

  1. ana53294 says:

    I have seen the claim that people without insurance use the ER in the US to access ordinary medical care that doesn’t need an ER, driving costs up for hospitals. But then some of the Vox stories @MrApophenia linked are about people with insurance being denied because they visited the ER unnecessarily.

    So, a question. Let’s say it’s a Sunday night, and you have a serious allergic reaction (no difficulties breathing though) to some exotic seafood. It’s not deadly, but it is really, really uncomfortable. Or you get ill during an especiall long bank holiday weekend. Where do you go? In Spain, we have clinics that do these special shifts during holidays. You may need to drive further, and call them, but you don’t need to go to the hospital. In my case, when this happened to me, I went to an emergency clinic, got a steroid shot by a nurse, and went home. The emergency clinic also happens to be 30 minutes away from home, whereas the hospital is an hour away.

    When I was a kid, my parents also drove me to the hospital in the middle of the night when I woke up with a terribly painful ear infection. Pediatric emergency care is usually given in hospitals, because there’s a lack of pediatricians*.

    Clinics handle many things. Stitches, shots, antibiotics, wound cleaning & treatment, etc. They don’t have X-rays or other machines, but anything that doesn’t require machinery or a lab can be done in a clinic (with less waiting time).

    Do you have emergency clinic services in remote areas of the US? Do people with urgent non-lethal issues have access to doctors?

    *Although we have more than other countries. Kids until the age of 14 get a pediatrician in Spain.

    • Edward Scizorhands says:

      It’s not deadly, but it is really, really uncomfortable

      • Well... says:

        I didn’t learn about urgent care until I was in my late 20s, and only because I started working in healthcare IT. Aside from the fact that a lot of people just don’t know urgent care centers exist (confirmed by research I did in the aforementioned job), aren’t urgent care hours relatively limited?

        • Eric Rall says:

          They often are, yes. Of the two urgent care centers closest to me (I live in a small-to-medium exurb on the fringes of the Bay Area), one is open 8am-5pm M-F and closed on the weekend, and the other is open 2pm-9pm M-F and 11am-6pm on weekends.

          I prefer the later model: if you have an established primary-care doctor, you can often get a same-day appointment for an acute complaint during standard business hours (often with your regular doctor, or if not, with another doctor in the same practice), so having an urgent care clinic open evenings and weekends closes more of the gap between emergency medicine and primary care.

          • albatross11 says:

            Yeah. If I need to see a doctor, I call my regular doctor’s office, and get scheduled with the next available person. (I really don’t see my regular doctor that often, tbh, but she did my last physical, so she has a baseline.) If I really want/need to see a doctor after hours for a non-emergency, I go to an urgent care. If I think there’s something really serious or maybe life-threatening going on with me, I’ll go to the hospital. (Fortunately, this doesn’t happen very often!)

            Mess up your knee in a way that’s still sore a week later and want it checked out = call your regular doctor’s office.

            Twist your ankle Friday at 8PM and want someone to check to make sure it’s not broken = go to the urgent care.

            Fall down the stairs and break your leg = go to the ER.

          • anonymousskimmer says:

            The mid to high tens of millions of people in the US alone do not have a primary care doctor.

            These are more likely the people who use urgent care centers or ER as their first recourse when in need.

      • ana53294 says:

        70% of centers in the United States open by 8:00 a.m. or earlier and 95% close after 7:00 p.m.

        That’s why I said Sunday night.

        What if you fell and need stitches at 2 am? There are many things that could happen during the night, that would be very miserable to wait until the morning.

        • EchoChaos says:

          That would typically be an emergency room visit, yes. “Emergency” doesn’t just mean “imminent loss of life” to most Americans.

          Which is why waits can be long if you “only” have a deep cut or an uncomfortable but not life-threatening case of poisoning.

          • albatross11 says:

            Yeah, a basic problem with going to the ER is that they’re subject to having some more serious case come in, at which point they absolutely will (and should) put you on hold while they deal with the guy having the heart attack or the victims of the nasty car wreck.

    • Corey says:

      There’s a regulation arms race wherein insurers try to keep people out of ERs, and then states try to keep insurers from doing that after the inevitable backlashes when people get denied for these kinds of problems.

      If you have no hope of ever having much money or good credit then getting denied doesn’t matter – the bills are just more noise.

      It’s heavily disincentivized for people who do want to have good credit and have experience with the healthcare system. Even if you’re not denied, everyone except the hospital itself will bill separately and be out-of-network (providers you don’t choose have no incentive to join networks), so price controls don’t apply. You’re stuck with the difference between the usual insurance rate and whatever they decide to bill you, in addition to whatever your deductible/coinsurance is on out-of-network care.

      Some (all US?) places require insurers to cover emergency care as though in-network, but then you can get caught up in arguments about what’s truly emergent, and balance billing still applies so in practice that may not reduce your costs much.

      At least with urgent care you can verify whether the doctor is in network ahead of time, there may still be pitfalls (lab? equipment like braces getting billed by a third party?) but not as many.

      • albatross11 says:

        I think several states have been trying to ban surprise bills (the hospital and surgeon are in-network, but one of the OR nurses is an independent contractor who’s out-of-network, and she’s sending you a bill for $5000 that your insurance doesn’t cover). This seems optimized for fraud and inherently corrupt, but I gather there’s a lot of pushback against those laws from people who make money doing it.

    • albatross11 says:

      Where I live, there are urgent care centers that are open 24 hours. This is basically where you go when you need to see a doctor (or more likely a PA) and it’s urgent but not all that important. If you’re having a heart attack or you know you just broke your arm, you go to the hospital instead.

      • acymetric says:

        I live in a mid-sized metro area (somewhere north of 1.5 mil). A quick check suggests there is one 24 hour urgent care within an hour drive, not counting actual emergency rooms. There is one more open until 9. The rest (and there are many) close between 5 p.m. and 8 p.m.

        I’m not sure 24 hour urgent cares are terribly common in the US.

        • Nick says:

          That astounds me. I grew up in small town Ohio and there was an urgent care every town or two, I’d guess. I wonder if they’re less common in certain states or in metropolitan areas for some reason.

          • acymetric says:

            A 24 hour urgent care? A quick perusal of the parts of Ohio I’ve spent time in doesn’t turn anything up (my perusal may not be a good one though, admittedly).

            I can certainly believe that some states/cities have regulations that make it less likely for there to be 24 hour non-ER urgent care facilities.

            My suspicion of the actual reason is that there isn’t enough market/demand for 24 hour urgent cares, because of ERs, leaving people who need urgent but not emergency care during off hours/days at the mercy of their insurance company.

          • Nick says:

            Oh, I feel stupid—I read right past the repeated mentions of 24hr. I checked google maps and the ones I’m seeing from my hometown area are all only open like 10-12 hours a day, not 24.

    • DragonMilk says:

      One reason is that emergency room visits are necessarily booked as in-network, even though the doctors are out of network.

      But if you didn’t need to go to the emergency room, then that visit is completely out of pocket.

      So many insurance companies, rather than dealing with the out of network hospital/doctor, would rather pass on the cost and negotiation process to you

      • acymetric says:

        I think people understand how it works and why. The point at hand is:

        But if you didn’t need to go to the emergency room, then that visit is completely out of pocket.

        Who determines this need, and how, and for things that are deemed “not emergency” but probably need to be addressed sooner than “in a day or two” are there other places besides emergency rooms where treatment can be had (particularly evenings/nights/weekends/holidays)?

  2. Two McMillion says:

    Over the course of my life, I’ve had several experiences which, at first blush, appeared supernatural. I’ve been able to devise natural explanations for most of them, but the one below is by far the weirdest, and the one which most defies easy explanation.

    When I was young, about eight or nine, my parents drove out to inspect an undeveloped piece of property they were considering buying. Much of the property was wooded, and I was excited by the prospect of seeing it. Once we were there, and my parents were distracted by talking to the seller, I ran off into the woods, eager to do some exploring. It wasn’t look before I looked around and realized I didn’t know where I was. I became extremely frightened, and started to cry.

    Now my parents are, and always have been, very religious people, and I had been taught about the existence of guardian angels. I decided to pray and ask for an angel to come and lead me out of the woods. So I closed my eyes, prayed, and when I looked up, I saw something.

    I want to be careful how I describe it, because I don’t want to be overly dramatic. I saw a figure, vaguely human shaped, standing next to a nearby tree. But it was clearly not a human, because I could see through it. The edges were sort of fuzzy, so it was hard to tell and at the same time clear where it began and ended, if that makes sense. I can say quite honestly that I’ve never seen anything like it since. I assumed that this was the angel I had prayed for and followed it out of the woods to where my parents were looking for me.

    So my question is… what the heck happened to me?

    The easiest explanation would probably be that my memory is wrong somehow. If so, it must have become mistaken very quickly, since I reported this story to my parents and several others soon after. The general response was disbelief, including a pastor who told me very firmly that I was lying. I don’t really know how to evaluate this alternative, since it depends on my only real source of information about the event (my memory) being mistaken.

    Another option is that I had some sort of hallucination. I’m not a mental health expert, but I’ve never had any hallucinations since then, and I’ve never been diagnosed with any mental illnesses, nor do I think I have symptoms of any. There is a history of mental illness in my family, however- my mother has borderline personality disorder and depression. The thing that makes this possibility weird for me is that the thing, whatever I saw, apparently imparted to me information I didn’t have- ie where I should go to return to my parents. I mean, maybe that information was buried in subconscious or something, but it seems like a hallucination shouldn’t be better then chance at imparting that info, right? “Kids runs off in random direction in woods” doesn’t usually end well.

    So yeah. What other natural explanations are there for that experience? Am I missing anything? I guess from your perspective another alternative is, “McMillion is lying”, but that one’s not really an option for me.

    • pqjk2 says:

      My theory:
      A part of your brain (say, a subconscious sub-agent) held information that it knew your conscious mind needed but did not have. The sub-agent accessed your subconsciously held imagery of an angel in order to communicate the information up to your conscious mind.

    • Well... says:

      You said it stood next to a nearby tree. Since you were in the woods, was there anywhere else it could have appeared except next to a nearby tree? How far away from you was it?

      You said you followed it out of the woods. This means it was moving. Was it walking? Floating along? Disappearing and reappearing? Also, how fast? Did you have to run to keep up with it? Did it seem to take you on a direct path, or did it seem like a meandering path? Did it traverse any obstacles that you had to duck under or climb over?

      Details there could help suggest some explanations.

      My default explanation (note: I am not a child psychologist or physiologist; this is purely based on my experience both as having been a kid and now as a parent) would be that as a young child, even at 8 or 9, you’re highly open to the power of suggestion. You can be made to “see” things relatively easy just by being told those things are really there. You might not see them in the normal way of light reflecting off those things and into your eyeballs, then bouncing off your retinas and being converted to a signal that travels along your optic nerves to your brain, but the end result still feels the same. I suppose you could call it a hallucination, except there’s an element of will in it; as in, you will the hallucination into being. Between your religious upbringing and various images of guardian angels in popular media (which adhere reasonably closely to what you described), you were kinda primed to “see” this thing.

      As for how it successfully led you out of the woods, I doubt you were really all that lost to begin with. You don’t have to go too far into the woods to feel like you’re in the middle of nowhere, especially if you’re exploring terrain that’s new to you and you’re a small person, when really if you just use your sense of reckoning to go back in the direction you came from, it’s not long before you start to recognize landmarks or even just see your destination through the trees.

      • Two McMillion says:

        You said it stood next to a nearby tree. Since you were in the woods, was there anywhere else it could have appeared except next to a nearby tree?

        Fair point, haha.

        How far away from you was it?

        Five feet or so, I think? Maybe a little more or less.

        You said you followed it out of the woods. This means it was moving. Was it walking? Floating along? Disappearing and reappearing? Also, how fast? Did you have to run to keep up with it? Did it seem to take you on a direct path, or did it seem like a meandering path? Did it traverse any obstacles that you had to duck under or climb over?

        I didn’t notice its exact method of locomotion. I think at the time I assumed it was walking but I don’t recall seeing that exactly. I did not have to run or climb over any obstacles, and the path was… hmm, I’m not sure, actually. I want to say straight more then meandering. Sorry if that’s not helpful.

        Your idea is interesting. I don’t have the expertise to evaluate but I’d be interested in readings on the subject. I have thought a little bit about what percentage of paths would get me back to somewhere I recognized- was it 10%? 50%? Unfortunately the area has since been cleared, so I can’t go back and inspect it.

        • Well... says:

          Since you were in the woods, was there anywhere else it could have appeared except next to a nearby tree?

          Fair point, haha.

          I mean, there might have been a clearing, or you might have been near a creek or something. Again, more details helps.

          Five feet or so, I think? Maybe a little more or less.

          If a human-sized object is five feet away, it’s difficult to see the whole thing at once without moving your eyes, but details about its texture, color, shape, etc. should be pretty clear. Yet you described this thing as hazy, translucent, not sure where it began and ended, etc. One possibility is it was made of heavenly cloud-matter that doesn’t conform to the kinds of things human eyes are used to seeing. Another possibility is you were imagining it and your underdeveloped kid-brain couldn’t work hard enough to invent all the details and keep them constant. 😀

          I also find it pretty incredible that you could follow something for more than a few feet and not notice the way in which it was moving, especially if it was humanoid yet moving in an unnatural way.

          Yeah, I too don’t have expertise to lend confidence to my explanation, but it still feels right. Wish I could make suggestions about where to read more, but I can’t.

    • Phigment says:

      I’m religious, so I’m comfortable with supernatural explanations.

      I think the question you have to ask, however, is what you’re willing to accept as an explanation. And what counts as non-supernatural. Obviously, we’re not going to prove anything definitively on this discussion board.

      I can, for instance, suggest that you didn’t see an angel, you saw a friendly Bigfoot, which lead you back to your parents. You seemed to be able to see through it because it was camoflaged to evade detection, which is how Bigfoots (Bigfeet?) have managed to remain legendary all this time.

      Does that count as a “non-supernatural” explanation? It’s not, theory, reliant on anything miraculous, but it’s still pretty unlikely.

      Maybe it was a trick of perspective, where a gap in the trees produced a humanoid silhouette when viewed from a specific position in specific lighting, and that happened to be in the general direction of your parents.

      Maybe your sub-conscious brain had a better sense of direction than your conscious brain, and, stressed by fear and primed by your belief in angels, it manufactured a vision of an angel to prompt you to backtrack your earlier course to your parents.

      Maybe a group of army rangers was conducting a secret training exercise in the area, got annoyed that some kid was crying in the middle of their sneaking practice and decided to lend a hand. Sneakily.

      Maybe God was helping, but went for low-touch and sent a vision of an angel to point you the right way, not an actual angelic being. Seems like something God could do. People get visions in the Bible.

      Maybe there were psychedelic mushrooms growing in that patch of woods and Young McMillion put one in his mouth.

      Maybe it was an actual angel.

      Maybe it was latent psychic powers connecting you to your parents.

      Maybe aliens abducted you, very rapidly probed you, and then returned you with your memory of the event fuzzed out. Maybe your parents know, and are hiding the truth from you. You whole life could actually be a one-off X-Files episode.

      • BlackboardBinaryBook says:

        Maybe it was a trick of perspective, where a gap in the trees produced a humanoid silhouette when viewed from a specific position in specific lighting, and that happened to be in the general direction of your parents.

        Maybe your sub-conscious brain had a better sense of direction than your conscious brain, and, stressed by fear and primed by your belief in angels, it manufactured a vision of an angel to prompt you to backtrack your earlier course to your parents.

        That’s where my skeptical secular mind went. The description sounds like it easily could have been a beam of sunlight through tree branches, blurred by tears. The young brain immediately made the connection “holy (pun intended) crap, an angel!” And faith + memory reconsolidation/alteration handled the rest.

    • DragonMilk says:

      The whole concept of dismissing the “supernatural” is interesting to me as it feels like the modern western intellectual is fixated on a paradigm that essentially assumes:

      1. Actual reality is fully in line with human perception
      2. Science is built upon what is replicable, and must necessarily be perceived
      3. “Natural” means within the realm of said science

      Given this, why are people so skeptical of the supernatural? If there is anything at all beyond human perception or current understanding, people look for a “natural” explanation.

      What if there exists things that human perception can only partially catalog methodically into the annals of science?

      To me it seems like a false dichotomy, and in a sense everything starts supernatural, and “natural” just means people have a working framework of repeatability, upon which can often be harnessed to make advancements in tech.

      • acymetric says:

        If you use a different meaning for the word than everyone else, you’re definitely going to have trouble understanding them when they use it. You can’t just break every word into its parts and derive a definition solely from the literal meaning of those parts.

      • John Schilling says:

        People are skeptical of specific supernatural claims which, if true, should be accompanied by far more photographic proof than is currently at hand. Obligatory XKCD. Supernatural claims like “homeopathic medicine works” and “I receive supernatural guidance manifesting as voices / feelings inside my head”, are much more widely accepted but interpreted and discussed in a way that is more compatible with the speaker’s overall belief system and so avoid the “supernatural” world.

      • Protagoras says:

        In line with what acymetric’s comment, “supernatural” doesn’t really mean anything like “outside normal science.” It has become a catchall term for assorted spooky stuff, stuff which is not in the “beyond human knowledge” category, but rather the “thoroughly investigated and found to be nonsense, but still weirdly emotionally resonant” category.

  3. BBA says:

    I will be starting a blog soon, and posting here sporadically. I’m unlikely to return to regular posting, because after some reflection (and an outburst elsewhere) I think I have a fundamental disagreement with the premises of this community. If the truth is everything that doesn’t go away when you stop believing in it, then I think the truth is much smaller than we suspect, and a lot of what we think is “truth” is built upon unstated assumptions and gerrymandered definitions to make an orderless universe conform to our models of it. With apologies to Kronecker, God made the positive integers, and everything else is a social construct.

    I was a math prodigy as a child, and the concept I keep thinking back to is Euclidean and non-Euclidean geometries. (This is a bit arcane, but since the object-level issues I’m referring to are way too controversial for a whole-number thread, geometry it is.) Euclid based the Elements off four simple straightforward postulates and a complicated fifth one about parallel lines. For centuries mathematicians tried in vain to prove the parallel postulate from the other four. Finally a few tried assuming the postulate false and seeking a contradiction, and instead found that there are other, internally consistent geometries where, for instance, two lines can both be parallel with a third line but intersect each other.

    My point is that the parallel postulate is neither true nor false, at least as far as four-postulate geometry is concerned. I can put on my Euclid hat and prove that the angles of a triangle add up to precisely 180°, or I can put on my Lobachevsky hat and plagiarize, plagiarize, plagiarize prove that the sum is less than 180°, but in the abstract, without knowing which system I’m in, I can’t prove anything about the triangles. And I think a lot of what the rationalist movement is trying to do is use Bayesian inferences to prove the postulate true or false and handwave away the inconsistencies with whatever theory it is. Or, perhaps, act as if the postulate is self-evidently true, when it’s only true for a certain subset of parallel lines, while others violate it. And then set a definition of parallelism that, explicitly or implicitly, only includes the first subset, ignoring the second subset even though it could also be considered “parallel.”

    I’ve gotten pretty good at changing my perspective, looking at a situation from a different set of postulates to reach a different set of conclusions. This leads to the unfortunate state where I can’t figure out which version of the “truth” I actually believe to be true. But it’s a position I’d rather be in than one where I dogmatically believe in one “truth” and just ignore all the inconsistencies and the studies that don’t replicate. In the interesting spaces of the universe, there are many truths, or none, depending on how you look at it.

    So there. I’m an irrationalist and I disagree with everything this community holds dear. But if you’ll have me back, I’ll try not to make too big a mess, doods.

    • Etoile says:

      I own a book of Tom Lehrer notes and lyrics; that song unfortunately doesn’t have the notes added, but it does have the words – hehe. Appreciate the reference.

    • Dino says:

      Are photons waves or particles? I somehow manage to be both a scientific rationalist, and also a mystic who acknowledges having had experiences that would be described here as “woo”. As Whitman said –
      “Do I contradict myself?
      Very well then I contradict myself,
      (I am large, I contain multitudes.)”

    • zardoz says:

      Please don’t go away. I’ve always enjoyed your posts.

      There are some things I disagree with in the capital-R rationalist movement, but from the outside, most of these arguments look like us debating whether we’re in the People’s Front of Judea, or the Judean People’s Front. I don’t know what you’re referencing with the argument about geometry or different perspectives. I guess maybe it’s a CW topic?

    • Iago the Yerfdog says:

      I have no idea what object-level issues you’re talking about, so that could change my judgment, but in the abstract it sounds like you’re becoming a meta-rationalist. Have you read any of David Chapman’s work by chance?

    • PedroS says:

      Please come by again when your blog is up to tell us its URL. I am keenly interested in following it..

    • Well... says:

      This leads to the unfortunate state where I can’t figure out which version of the “truth” I actually believe to be true.

      Why is that unfortunate? For me, getting into that state is a marker of intellectual success because it means I have likely discovered the strongest arguments on both sides, or at least discovered the irreducible difference between the sides and found it to be worthy of its controversy.

    • albatross11 says:


      Isn’t the solution to this concern (we’re arguing over premises so we never get anywhere) to look for actual evidence about which world we’re in? That won’t always answer things, but it often will let us determine some stuff. We can determine whether gravity makes things fall up or down in our world by experiment, rather than trying to reason it out from first principles.

      Otherwise, your comment reminds me a bit of Scott Adams’ discussions about “watching different movies.” Basically I think his idea is that two different people with different starting models of the world could watch the same events and see completely different things due to different premises. (Imagine a theater where we all started out watching different opening scenes, and now we’re all watching the same middle scene together–it’s possible I’m laughing and you’re crying at the same scene, due to our different worldviews, knowledge, and assumptions.) To use a not-very-CW-around-here example, this is how you could get some people honestly convinced that Obama was a secret Muslim, who could find evidence that re-enforced this belief, while most of the rest of us found this a really silly idea.

    • Aapje says:


      The universe is clearly not orderless or else you wouldn’t have been able to write that, but it isn’t perfectly ordered either. It is very complex. Models are abstractions/simplifications of that complexity, resulting in something manageable, although you inevitably lose accuracy in the process, compared to actual reality. However, we cannot operate based on reality, because it is too complex.

      ‘Truth’ is not so much a thing in itself, but part of important processes. If we act, our actions result in an outcome that we must interpret. This interpretation requires a model, where we interpret what actually happened better if our model is more truthful. If we want to predict the outcome of an action, we also need a model. If we want to reason back from a desired end state to the actions that we need to do to achieve it, we also need a model.

      More truthful models result in better predictions and better interpretations.

      A complication is that many people want to predict things for which we have poor models. This desire doesn’t go away just because you tell people that the truth doesn’t exist. People also typically can’t do all the things that all the models suggest will result in an outcome, if only because models can contradict each other. So you still need to be able to distinguish better models from worse models.

      But it’s a position I’d rather be in than one where I dogmatically believe in one “truth” and just ignore all the inconsistencies and the studies that don’t replicate.

      IMO, people here are far more likely to recognize/admit that the models they are using have inconsistencies, exceptions or poor evidence than pretty much every other group.

      • albatross11 says:

        I will point out that sometimes, people talk about “truth is socially defined” or “what is truth?” because they’re making a serious point that the truth is hard to get at, but a lot of times, ISTM that they make those comments because they don’t want to acknowledge some unwanted truth.

    • meh says:

      My point is that the parallel postulate is neither true nor false, at least as far as four-postulate geometry is concerned.

      I’m not a prodigy, but how is something neither true nor false?

      • albatross11 says:

        It is consistent with the rest of your premises for it to be either true or false. You’re proving things from a starting set of premises, so any statement you want to make may be true or false based on those premises, but it may also be consistent with those premises for it to be either true or false.

        • meh says:

          is ‘either’ not different from ‘neither’?

          • Iago the Yerfdog says:

            I think the point they’re trying to get at is that four-postulate geometry is underdetermined.

            It’s completely possible to do geometrical work without either accepting or rejecting the parallel postulate. You just can’t prove as much because you’re limited to results true in both Euclidean and non-Euclidean geometries.

            For such work, the truth or falsity of the parallel postulate hasn’t come up yet, and when it does come up you can freely choose which truth-value you want it to have based on what you’re trying to do.

      • Iago the Yerfdog says:

        To add to what albatross11 said, formal systems are never really true or false, only applicable or inapplicable. For example, space-time as a whole is non-Euclidean, but it’s close enough to Euclidean in most contexts that we can use Euclidean geometry, including the parallel postulate.

      • Dacyn says:

        The parallel postulate isn’t really an assertion until you specify a notion of geometry for it to apply to. Euclid imagined his axioms as applying to the idealized intuitive geometry. When the parallel postulate is interpreted in the context of the idealized intuitive geometry, then it is true.

        Bolyai and Lobachevsky came up with another notion of geometry such that the parallel postulate is false when interpreted in the context of that geometry, but the other four postulates are still true. I would think of it like, there are different languages that all share the same words, but not the same meanings. So a sentence which happens to be grammatically correct in both languages could be true in one but false in the other.

    • Radu Floricica says:

      Myeah, was thinking I have something to say, and then I get to Aapje’s comment. 🙂

      Anyways, I do have a point. What you’re saying, don’t take it the wrong way, isn’t particularly new. For one thing, we live in a world defined by quantum physics. It’s kindof a given that pretty much evertything we see, think or do is a big illusion. (Another point scored for ancient religions, btw) For another thing, like Aapje said, we live in an extremely complex world. To function in it we have to carve up that complexity in abstractions, so yeah, everything depends on the way we carved it, including what we call true. But you should remember that things are a lot less random than they seem. Colors seem like a perfect example of a random definition (why is Red that particular bunch of wavelenghts?) – and yet turns out most human languages name the same colors. Because at some level it makes sense to carve reality this way.

      But my point is another. That’s usually a precocious teenager’s kind of inner crisis. Are you ok? Is there anything else in your life that’s pushing you a bit off balance? Maybe the problem is elsewhere.

    • aristides says:

      Just a reminder, the SSC community is very different than the rationalist community. I personally, dislike and disagree with most rationalist beliefs, but I feel at ease here. I agree there are many possible truths, and the benefit of SSC, is that it explores one possible set of truths very well. I will look forward to your blog, and hearing your thoughts unfiltered.

    • Viliam says:

      I think the sum of angles in a triangle must be a random number, because otherwise free will couldn’t exist, and then the judgment in the afterlife would be just a cruel mockery of justice…

      Nope, just kidding.

  4. Etoile says:

    Question for SSCers with kids, or with expertise in early childhood development.
    Let’s say you have a child age 2-3 who has a very good listening memory – like, hears a song once or twice and then sings it back to you; recites a poem back after several tries; mimics things even in an unfamiliar language. What sorts of developmental activities would you do with this child to capitalize on this ability?

    • Recite poetry to her when putting her to bed.

      And expect to hear a very small voice, from the back set of the car, saying:

      “Larth Porsena of Cluthium, by the nine gods he swore … “

      • Well... says:

        My wife and I got on a Steely Dan kick about a year ago. It lasted several months. Now our kids know the words to most of the songs on Aja, and sometimes sing them in public. It hasn’t gotten us into trouble yet, but I’m sure one day it will.

        • Etoile says:

          That’s funny. Unfortunately part of the songs I now get is whatever youtube children’s channels play (the same nursery rhymes over again, plus a few I didn’t even know).

    • Rebecca Friedman says:

      I don’t have a lot of expertise, but judging by what I know of linguistics that’s very comfortably within the critical period for language learning. Are there any immersion programs she might enjoy? TV in other languages, if she watches TV, or songs? Just listening to a language at that age and getting the phonemes in your head can make language-learning without an accent at a later age possible instead of impossible. And if she’s a brilliant imitator and enjoys it…

      (To my knowledge as of about a decade ago; I haven’t kept up with the research, but it seemed pretty firmly established, at least if you heard the language regularly.)

      • Etoile says:

        Thanks! She gets Russian (and random songs in other languages) fairly regularly, probably 10-20% of the time. I have no pretensions towards perfect fluency for her, but I did have a similar thought about at least getting the phonemes and the intuition for the grammar.

    • Björn says:

      I would suggest activities where you sing with your child. This can be singing along to the radio, going to church or parent-child singing programs in music schools. I googled a little bit, most children’s choirs accept children beginning with 4 or 5, so this will be an option in a few years. When your child is 6, you can think about getting her music lessons in some instrument. I highly recommend getting her a teacher who teaches her to sing along to what she plays.

      • albatross11 says:


        Note that there’s a huge range of quality of choir teachers. The one at our church is amazingly good; the one at our daughter’s Catholic school was abysmal for years and now has been replaced. (Hopefully the new one will be better.) If she’d only had music instruction at her school, it would have been worse than nothing.

      • Etoile says:

        We do lots of singing! I was wondering more about capitalizing on it to learn other content, not just music and poems. I can teach her some (I played piano and had a very good instructor, even if I wasn’t as good a student), but yes, I do hope to find her a teacher if she finds music lessons interesting.

  5. b4mgh says:

    Back in OP 141.5 @Plumber gave an answer to the question “You have a billion dollars. It’s yours free and clear. What are you going to do with it?” that included the construction of a neighborhood arranged in hexagonal blocks.

    I was hoping he could explain the logic behind that particular choice, and that others with knowledge or interest in urban planning could say more about the factors taken into consideration when determining the shape and size of blocks in an urban area.

    • Well... says:

      Would the roads be arranged in straight lines meeting at 120˚ angles, or as a series of circles circumscribed about the hexagonal blocks, with smaller triangular blocks in between them?

      • DragonMilk says:

        Funny enough, I made a city in city skylines using a hexagonal grid to avoid having to stop.

        Problem is a lot of weaving of traffic means high volumes will wreck the system. Before you do infrastructure, try it in City Skylines first!

      • Plumber says:

        @Well… says:

        “Woulld the roads be arranged in straight lines meeting at 120˚ angles…?”

        Yes, so they’d be “Y”/three-way intersections instead of “cross”/four-way ones.

        • Eric Rall says:

          A pure-hexagon tessellation like what I think you’re describing wouldn’t have any straightaways longer than the sides of a single hex. So a car driving several blocks or more would need to make a turn at each block. Is this an intentional feature of your design here, perhaps as a traffic-calming measure or a way of discouraging people taking shortcuts through the neighborhood? I went back to look at your original post, and I saw you talked quite a bit about walking distances within the neighborhood: and if you want/expect people to mainly be walking within the neighborhood, making driving any distance within it awkward makes a fair amount of sense.

          If it isn’t intentional, you can avoid it by mixing hexagons and triangles, perhaps using the triangles as mini-parks or for commercial or public service buildings.

          • Plumber says:

            @Eric Rall,
            Yes exactly!

            Slowing down motor vehicles was a goal (@#$/ the Fire Department, more kids will be saved with slower traffic than will be lost by them going slower to the rescue, they can have more stations if they press the issue, but them wanting their big ass trucks being able to go fast will not decide the streets!).

            I’ve lived on wide streets for 20+ years, and I live next to “two Ford Model T wide streets” (effectively one-and-a-half modern car width streets) that were layed out in the ’20’s, so I see the difference between them (my street now has speed bumps, but the narrower nearby ones don’t need ’em!).

            Since narrower streets are only effective past the intersections, I wanted a slowing mechanism at the crossings, hence “Wye” junctions, and hexagonal blocks.

          • Well... says:

            See, nobody here has a problem with proposals like this, but God forbid I should speak up about 55 mph speed limits…

          • Eric Rall says:

            See, nobody here has a problem with proposals like this, but God forbid I should speak up about 55 mph speed limits…

            I’m okay with Plumber’s proposal here as a thought experiment, or in the context of a hypothetical property developer spending his own money to set up a neighborhood like this which people could move into or not as they pleased.

            I would not be okay with it as a universally-legislated policy, and not just because I’m a libertarian. IMO, there are better solutions to the main problem Plumber’s trying to solve here. My personal favorite is residential cul-de-sacs with foot/bike trails connecting the ends of the cul-de-sacs. Look up “Columbia, MD” on Google Maps for an example of what this looks like in practice. The streets in front of the houses are quiet and safe, since there’s no through traffic at all. Meanwhile, the connecting trails keep the wider neighborhood walkable, and the cul-de-sacs connect directly to larger, busier arterial roads allowing quick access by car in and out of the neighborhood.

          • Nornagest says:

            nobody here has a problem with proposals like this

            I absolutely have a problem with it, but I objected the first time it was brought up and I can only repeat myself so many times before I start feeling like I’m wasting my time.

          • bullseye says:

            I’ve lived in Peachtree City, Georgia, which has the same features you describe for Columbia. It’s nice, except that for most residents nothing is within walking distance except for more houses and maybe a park. Lots of people drive golf carts on the bike trails in order to get groceries.

          • Nick says:

            I’m not a fan of the proposal, either, or of planning cities like that in general, but I felt it would be ungracious to post a long criticism of it the same place Plumber’s naming me “west coast ross douthat.”

          • Anthony says:

            If you take away a little bit of land from the corners, you can make each of the three-way intersections into traffic circles, which will allow smoother flow without needing to increase speeds.

          • A Definite Beta Guy says:

            If you’re trying to calm traffic, I think those narrow streets are going to be more effective than trying to arrange everything as a hexagon (or any other kind of street design that makes traffic flow effectively impossible). Right now I live in a suburb laid out as a streetcar suburb, built and designed back in the 30s. The streets are narrow and speeding is incredibly uncommon because of how narrow the streets are.

            Most residential neighborhoods post-war and especially recently are very wide, practically collector road width. Even with the gentle curves in them as is typical in subdivisions, people still speed more, because the streets are just too wide.

            I also believe we have a lot more mature trees and smaller/shorter driveways, so people are more apt to park on the street, which again calms the traffic.

            Also both bikes and walking are stupid and these problems will eventually be resolved/substantially reduced by plug-in cars driven by AI. Even if it is not resolved today, it will be resolved in the next few decades, and whatever urban design you decide on today will last through several generations of transit.

          • DragonMilk says:

            I am so tempted to post my City Skylines city to show why this doesn’t work.

            It’s basically a tesselation hexagon where you’re traveling ~30% greater distance to avoid stopping, but to the extent you need to turn or go somewhere local, you’re going to be weaving through a lot of traffic, so increasing the load proves disastrous

          • Nick says:

            Also both bikes and walking are stupid and these problems will eventually be resolved/substantially reduced by plug-in cars driven by AI.

            screams for ten minutes

          • John Schilling says:

            Also both bikes and walking are stupid and these problems will eventually be resolved/substantially reduced by plug-in cars driven by AI.

            Or by the people who think bikes and walking are stupid vanishing into VR pods 24/7 and so absenting themselves from the world the rest of us live in.

            AI will be used to tinker with their feeds so that they naively imagine that their votes are still being counted and that they still wield power and influence in the real world.

    • Plumber says:

      @b4mgh says:

      “….that included the construction of a neighborhood arranged in hexagonal blocks.

      I was hoping he could explain the logic behind that particular choice…”

      Oh sure, I just don’r like crossroads/four-way intersections, as I think they’re dangerous because of collisions and selling your soul.

    • Nearly Takuan says:

      The only people I’ve met who like this thing are the exact (very wealthy / “upper middle class”) people who own land on the block. It frustrates all other drivers — which I realize is a goal, but the remainder of the metro area is set up in a way where almost nobody can afford to live somewhere they wouldn’t have to drive to work (those who are poorer still take multiple bus transfers, = 1.5 hour commutes = 12-13 hours of commitment for an 8-hour workday)

      Biking through there also sucks: I used to have the good fortune to live within actual biking distance of my workplace, and since Ladd’s Addition is technically bike-friendly the nearby streets don’t have bike lanes in that area. However, the triangles and smaller hexagonal junctions are actually really hard to turn through properly if there are any other bikers or pedestrians nearby. I saw a bunch of other bikers fall or even run into each other. Possibly this is partly because despite the city’s efforts to encourage a biking culture by making other forms of commuting more unpleasant, actual bikers often seem to think they don’t have to regulate their speed or watch for stop/yield signs. But the poor visibility and awkward left-right-left-right turns didn’t help either.

    • Lambert says:

      As a European, I propose that city building should be done the proper way.
      However, folks these days are terribly impatient and unwilling to wait the centuries needed to properly optimise a city layout.

      As a compromise, I suggest simulating the city on a petri dish and seeing copying what the slime molds do.
      Or maybe just throwing a plate of spaghetti at a map.

  6. Plumber says:

    From Does Who You Are at 7 Determine Who You Are at 63? in The New York Times:

    “…To spend time with a child is to dwell under the terms of an uneasy truce between the possibility of the present and the inevitability of the future. Our deepest hope for the children we love is that they will enjoy the liberties of an open-ended destiny, that their desires will be given the free play they deserve, that the circumstances of their birth and upbringing will be felt as opportunities rather than encumbrances; our greatest fear is that they will feel thwarted by forces beyond their control. At the same time, we can’t help poring over their faces and gestures for any signals of eventuality — the trace hints and betrayals of what will emerge in time as their character, their plot, their fate. And what we project forward for the children in our midst can rarely be disentangled from what we project backward for ourselves….”

    An essay on the making and impact of thr 7 Up film series, which started as a documentary on a group of British seven year-olds in 1963 “from different backgrounds”, and it was mostly about the British class system (which the initial installment lays on thick) but, as the years have gone by and the filmmakers visit and document the lives of the group every seven years, it’s become much more than that, an amazing chronicle of the human condition, and our common humanity. 

    At 7 they’re all charming children (though the class differences are obvious, which was the filmmakers intent), at 14 and 21 the group are quite egalitarianly mostly morose adolescents, from 28 to 42 the differences in their backgrounds start to have ‘disperate impacts’, but there’s plenty of surprises! At 49 and 56 (the last one I saw, “63 Up” is coming to theaters now) the individual fates of the group really come down to their different personalities, individual tragedies and triumphs, and it’s all quite touching, and you care about these different folks.

    Anyway, I highly recommend reading the essay and viewing the films.

    • Well... says:

      it was mostly about the British class system

      All British films and TV shows are about the British class system.

      • Any that are in favor of it?

        I can think of at least one Kipling story that I think is, and elements of Dorothy Sayers’ novels. Both of which see it more as a division of labor than a status hierarchy, and both of which depend on the upper class choosing to take on responsibilities associated with their position.

        • Plumber says:

          @DavidFriedman says:

          “Any that are in favor of it?

          I can think of at least one Kipling story that I think is, and elements of Dorothy Sayers’ novels. Both of which see it more as a division of labor than a status hierarchy, and both of which depend on the upper class choosing to take on responsibilities associated with their position”

          Even though the obvious original intent of the “7 Up” film was a criticism of it, with the later installments the impressive noblesse oblige that the “toffs” exhibit in their charity and civic work looks like a point in favor of the British class system (one of the working class ladies also is impressive in exhibiting some of the same virtues as the upper class men, albeit with more of a ‘chip’ on her shoulder).

          Not so much the women, but the working class men in some ways look to have had easier lives than the upper class ones. 

          It really is a remarkable chronicle of different lives.

      • Plumber says:

        @Well… says:

        “All British films and TV shows are about the British class system”

        Are You Being Served? is about the British class system? 

        Benny Hill is about the British class system? 

        Black Adder
        is about the British class system? 

        Blake’s Seven is about the British class system?

        Dad’s Army is about the British class system? 

        Doctor Who is about the British class system? 

        Excalibur is about the British class system? 

        I’m Alright Jack
        is about the British class system? 

        Monty Python is about the British class system? 

        The Italian Job is about the British class system? 

        Okay, I can see your point, but Doctor Who was totally about Leela (Louise Jameson) in that animal skin mini-skirt!

        • Iago the Yerfdog says:

          EDIT: I just noticed the last line of your comment, so sorry for not catching the sarcasm. That said, I’ll leave this up since I’m particularly proud of my analysis of the show.

          I can’t speak too much for the others, and I think the original statement was meant as hyperbole, but…

          Are You Being Served? is about the British class system?

          Yes. Yes it is. Very much so. I mean, most of the humor comes from sexual innuendo and Mr. Humphries’ heavily-implied-but-never-confirmed sexuality, but most of the rest of the humor comes from the fact that Grace Brothers is a microcosm of the British class system.

          You have Mr. Harmon, who represents the working class and is the only one who is actually good at his job and knows what’s going on.

          The clerks are the middle class, including Ms. Brown, who recently graduated from the working class.

          The floor walker, Captain Peacock, was an officer (but not a gentleman) during the war, and so is right on the border of the middle class and upper middle class. He’s a bit pompous and detached, but still has some grounding in the realities of the store. He’s the snob, the social climber.

          Mr. Rumbold is the upper middle class manager, and he’s a clueless, if well-meaning, imbecile most of the time.

          Mr. Grace represents the upper class. While charming and classy in his own way, he is blissfully detached from what’s going on, at death’s door (as many assumed/hoped the peerage and monarchy was/are), and is a hopelessly decadent dirty old man.

        • Well... says:

          I stand by my point.

          Also see: Keeping Up Appearances, Downton Abbey.

        • Anthony says:

          While I prefer the first Romana, the second Romana:

          is the daughter of Edward Ward, 7th Viscount Bangor, and his fourth wife, Marjorie Alice Banks; as such, she is entitled to use the courtesy title “The Honourable”.

  7. kalli says:

    Speaking of investing. Any recommended books / blogs / podcasts for good insight into the options for a (middle class) individual to take care of and grow their savings. The rates I see on saving accounts (in the EU) are really low, but am unsure where one should go (bonds, funds, stocks etc), especially with a lot of talk about a looming recession.

    I feel like knowing this is part of becoming an adult. But find information on it hard to grok, a lot of get rich quick schemes and lot of jargon when you look online. Is getting a financial advisor worth it?

    • Nornagest says:

      The Personal Finance subreddit ( is a pretty good beginner-level resource.

      Start here for the European version. The general strategy — which is also what a financial planner will tell you — is to pay off necessities, then keep enough cash on hand to cover major unanticipated expenses, then pay off high-interest debt, then invest into diversified funds of some sort, ideally tax-advantaged. Exceptions apply in various situations (if you’re saving up for a house or a car, for example, investing that money introduces short-term risk for little long-term return).

    • Read up on the efficient market hypothesis and the equity premium puzzle. tl;dr is passive investing is best, and if you aren’t going to need the money in the next few years, you can afford to take risks.

    • kalli says:

      Thanks for the info and links Nornagest and Alexander Turok.

  8. TheContinentalOp says:

    Conservatives and Libertarians will point to Cosmetic Surgery, IVF, and LASIK as examples of healthcare where markets worked to lower prices and improve quality.

    Liberals and Progressives respond, “When you’re having a heart attack, you’re in no position to comparison shop.”

    So my question is: What percentage of healthcare spending is so urgent that immediate treatment is necessary with no time to consider other options?

    Fun anecdote: My friend was suffering from abdominal pains (turned out to be appendicitis) and she used Yelp to select which ER to go to.

    • Statismagician says:

      There aren’t good figures for this anywhere I’m aware of, but I strongly suspect that it’s not really a useful avenue to consider the problem from – the fixed costs of hospitals and health care systems are largely determined by firstly the need to maintain sufficient space and staffing to handle emergencies and round-the-clock care, and then secondly the obvious economies of scale from co-locating other medical services where you already have a bunch of them. Cosmetic surgery, IVF, and LASIK are all by definition luxury options, which I think I can say uncontroversially respond differently to market forces than do necessities.

      • Cosmetic surgery, IVF, and LASIK are all by definition luxury options, which I think I can say uncontroversially respond differently to market forces than do necessities.

        Food is a necessity, yet most are fine with it being provided by the free market.

        • Statismagician says:

          There are multiple Federal agencies whose main job is to regulate the food supply, including guaranteeing access to subsistence-level food, and they only look as unobtrusive as they do because we live in a time of unprecedented plenty.

          • Cliff says:

            The regulation on food is far, far, far less than on healthcare.

            luxury options, which I think I can say uncontroversially respond differently to market forces than do necessities.


            LASIK is also not really a luxury good anymore. A few hundred dollars per eye. How would you draw the line between necessity and luxury? Glasses are a necessity, contact lenses a luxury? Vegetables a necessity, packaged foods a luxury?

          • cassander says:

            There are multiple Federal agencies whose main job is to regulate the food supply, including guaranteeing access to subsistence-level food, and they only look as unobtrusive as they do because we live in a time of unprecedented plenty.

            I’m sure that most of us on the right would be absolutely ecstatic if the federal government’s regulation of healthcare was limited to them supplying everyone with all the antibiotics, vaccines, and bone settings that their doctors prescribed and checking the purity of drugs people bought for cash.

          • Anthony says:

            None of those federal agencies directly provide food to anyone (except the military), and the only price-setting they do is to enforce minimum prices. “guaranteeing access to subsistence-level food” is done by giving money to poor people so they can afford to buy on the (almost unregulated) market.

        • Corey says:

          My canned response: Food’s just a bit more fungible than health care. A hungry person can be fed from a wide variety of foods – if corn gets too expensive they can eat beans. If someone has multiple myeloma, it doesn’t matter to them how cheap metformin is.

    • Emergency medicine could be regulated as a natural monopoly, the same way electricity is regulated.

      • albatross11 says:

        In Maryland, new hospitals and emergency rooms have to get some kind of state approval intended to ensure there’s not too much competition. The two cases I’ve seen of this approval process looked pretty sketchy to me, but I don’t know enough to be sure.

        • Cliff says:

          Yes, “Certificate of Need” laws are reviled as a classic case of rent-seeking.

          • Aftagley says:

            Disagreement here – they can be rent-seeking, but they also help reign in some stupid excesses of the medical industry.

          • Disagreement here – they can be rent-seeking, but they also help reign in some stupid excesses of the medical industry.

            Like what?

          • Edward Scizorhands says:

            In some areas of medicine, doctors have been observed to be able to manufacture their own demand. I forget the economic term for it, but back doctors performing back surgery was the prime example.

            I have trouble seeing this apply to emergency rooms, though.

            (If the government is backstopping them — which is the minimum I would expect in any foreseeable system — I can see them not wanting too many of them because of multiplication of fixed costs, but this is separate from being able to manufacture demand.)

    • Edward Scizorhands says:

      As a first pass for “no time to comparison shop, must go to hospital NOW”: emergency rooms are less than 2% of total health care spending.

      There are other features of IVF/LASIK/cosmetic surgery that make them cheaper:

      1. (Already stated: time and freedom to comparison shop.)
      2. Results are easier to compare. While they aren’t perfect measures, you can definitely tell what someone’s new vision is, or if someone got pregnant.
      3. They started from infinity costs, so they had nowhere to go but down.
      4. They are not paid for by insurance. So they have to tell you the price up-front. As compared to insurance-heavy parts of the industry, where asking for a quote varies from difficult to “fuck you for asking.”

      • Results are easier to compare. While they aren’t perfect measures, you can definitely tell what someone’s new vision is, or if someone got pregnant.

        You can tell if you’re alive, or if you still have the flu, or if you are still in pain, or if you can walk without a walker, ect. If they perform an operation to reduce your chance of further strokes from X to Y, yes, that’s hard to compare. But that’s actually a reason to care more about cost, not less. If cost is the only variable you can accurately estimate, it makes sense to only optimize for it.

        3. They started from infinity costs, so they had nowhere to go but down.

        That’s true for any medical advance.

      • Gobbobobble says:

        As a first pass for “no time to comparison shop, must go to hospital NOW”: emergency rooms are less than 2% of total health care spending.

        Does this figure include the total visit cost for when the ER is the point-of-entry and you get routed to another department for ICU/recovery/observation/etc?

    • ana53294 says:

      OTOH, dentistry, which is rarely urgent, gives you prices upfront, etc., is full of quacks, unnecessary procedures, and it’s hard to compare. And in most places, it’s not covered by insurance*.

      I use nepotism for finding a dentist: my cousin’s husband may or may not be the best doctor, but I trust him not to try to make me go through unnecessary procedures. Other people who don’t have a dentist in the family try to find a friend of a friend, or get references from friends and family. Even when I was studying quite far from home, I made sure to get my dental check ups with the dentist I trusted, because I didn’t want to find another one.

      In the UK, from what I’ve heard, partly because some people get their dentist from the NHS, and people try to get the free service, private dentistry doesn’t work very well, and people have to wait in queues even when they’re willing to pay full price. I know people who came back to Spain from the UK to get their dental procedures done.

      *Or covered by some insurances, which means there is a mixed market. It can be quite confusing.

      EDIT: To clarify: the thing that is not transparent is whether a procedure is necessary at all, not its price. That is usually known in advance. Modern medicine has two transparency problems: the price of a procedure and whether the procedure is necessary at all.

    • Aftagley says:

      So my question is: What percentage of healthcare spending is so urgent that immediate treatment is necessary with no time to consider other options?

      This question is kind of framing it the wrong way – the biggest issue is what percentage of doctors will run a bunch of expensive tests for little or no anticipated benefit and what corresponding percentage of patients are savvy enough to say no? As long as the first percentage is high and the second low, there is a need for an informed third party to help protect the patient.

      • Edward Scizorhands says:

        That doesn’t seem like a different framing. It seems a question about an entirely different part of health care spending.

        Both are good questions to ask, though.

    • ADifferentAnonymous says:

      I had the same thought a while ago and asked about it here… The answer was that it’s a very low percentage. Don’t feel like digging it up, and I don’t remember the citation, but FWIW my bias was against this conclusion–I was hoping I had come up with a clever argument that would apply to a big chunk of healthcare.

    • rahien.din says:

      Markets work best when the customer is basically outsourcing – buying a good or service that they might otherwise produce or perform for themselves, if they preferred their money to their time. But medicine is often rather arcane. Even for the physician. I can easily tell if my entree is tasty. I can easily tell if my plumbing works again. It may be much harder to tell if my medical care is good or adequate, if the cure is die to skill or luck or tincture of time. Moreover, the better doctors may select for the most difficult or refractory cases.

      Of course, patients can only choose between physicians based on the variables they can access. But that simply means it’s a drunkard’s search. (Trying to pick a dentist is a humbling window into the difficult decisions patients face in selecting their doctor.)

      So it seems like a mistake to expose those variables to the exacerbating influence of the market. It seems as though this will not necessarily result in better delivery of healthcare.

      • cassander says:

        But medicine is often rather arcane.

        So are tax accounting, home renovation, and the legal system, but we manage to find accountants, contractors, and lawyers passably well when we need them. They’re not the best functioning markets by any means, but it’s not a complete show stopper. And even if it were, how does having someone else pay for the doctor solve the information problem?

        • MrApophenia says:

          Because the someone else actually understands what the going rate is and can negotiate on that basis.

          If your hospital tells you a procedure you need costs $5000, most people have no real basis for understanding if that is an insane cost or a good deal. If they quote the same price to an insurance company, they can actually respond “Ridiculous, every other hospital in this city charges $200 for that, we’re not paying that much.”

        • rahien.din says:

          Arcane does not simply mean “difficult” or “hyperspecialized.” It also means “inexplicable” or “illegible.”

          I would pay a tax accountant because I value my time more than my money. Same as I would pay a person to renovate my house. Same as I would pay a person to navigate the human-constructed legal system.

          Medicine is different – it is stochastic and illegible. Not everyone will get well or even survive, and definitionally the times when people don’t do have their desired outcome are the times when the outcome is out of our control, often without a clear reason why.

          All this adds up to : physicians don’t fix patients because patients aren’t broken. (Spoiler: this is not an easy thing, as a patient, to accept.)

          Instead, physicians forecast the effects of the patient’s actions, elicit the patient’s health goals, and provide a curated menu of actions. Thus physicians get paid, essentially, according to the whim of actuaries and risk assessors. This is often tedious and sometimes infuriating, but it’s probably correct. As a community we want medical care overall to regress to the proper means. We monetarily judge physicians not on how enticing their treatments are to the individual, but on the downstream evidence of the care they provide to the community.

          But a patient isn’t interested in helping the world regress to a mean. They want to achieve their desired state of health. A great many people seem to want to be “fixed.” Just look to the suspicion lodged in Caplan’s proposal that we “buy health, not healthcare”: if we are not healthy, it must be because the physicians are withholding health from us!

          Health isn’t really a good or service. Health is an action.

          Which is why we need to empower physicians to advise people appropriately. Your physician may sometimes have to tell you “This sucks but we will manage” or “You will have to do such and such thing that you probably won’t like.” You really don’t want their livelihood to suffer when that situation arises. Otherwise they may have a financial incentive to lie to you, or not to care for people with tough diseases.

          You may think that’s pontificating bullshit. But more often than not, when patients are dictating care decisions according to their own self-interest and demanding the types of legible services that the market would utilize as incentives, their care and outcomes suffer.

          A true example: in India there is a hospital with two main wards, one for the government-care rabble, and one for the privately-insured VIPs. The former isn’t very nice. The latter is downright fancy and caters to every patient request. Outcomes are consistently better among the poor than among the rich.

          Certain areas within medicine seem to be “fixing people,” such as LASIK. Fine – expose elective and legible procedures such as these to the market so that even more people can elect into them. Of course there is vast room for improvement in our delivery of medical care.

          And of course all professions face these challenges to some degree. That’s why you pay the lawyer for his time even if he tells you that there is no case, and why you pay the renovator a fee for their mere estimate.

          But this forecasting is the very soul of medicine – as Osler said, it is a science of uncertainty and an art of probability. You have to pay the weatherman even when it rains (and your bookie, when your dog doesn’t place).

          Don’t make the mistake of generalizing from non-central exceptions.

          • albatross11 says:

            Can you think of other occupations that have similar properties? Maybe investment advisors?

          • albatross11 says:

            W.r.t. accessing quality of doctors, this raises a question that’s answerable but I’m not sure how good the data is to answer it: How much difference in outcome can I expect if I get, say, a 10th %ile doctor vs a 50th %ile one vs a 90th %ile one. Let’s assume I can’t manage to go to the Mayo Clinic or something at get the very best doctor in the world at my problem–how much difference is it likely to make whether I get someone on the low vs high end of normal? Is there good data on this?

            The advice I’ve always heard (and followed when it applied) is that you should care a lot about how often a doctor has done some procedure. (When I had surgery a few years ago, one of my main questions for the surgeon was how often he did this kind of surgery. “A couple times a week” was a pretty reassuring answer.)

          • rahien.din says:

            A financial or investment advisor would be an apt comparison. Or a bookie, or a meteorologist.


            Your other question is much harder to answer. What are the percentiles describing? Could be the chance of a good outcome, or patient satisfaction, or efficiency of care, or professional reputation, or something else. Even if we say “quality, by any measure or some ideal measure,” it would depend on the disease. You might need the world’s best surgeon to separate twins conjoined at the head. You might not need the world’s foremost pediatrician for your routine viral URI.

            We also have to think about how we want the system to allocate cases. Medical care is a scarce resource. The most difficult cases should be cared for by the most proficient physicians – more of their time should be spent on these. So, if you have a routine illness, your case shouldn’t be referred to the Mayo Clinic, even if you want that. You need to see the person tasked with routine illnesses: an urgent care center, a PA or NP, a trainee, or even (as in some underserved areas) a physician whose medical training has been fast-tracked so they can fill care gaps.

            So, it may be that your illness does not need the attention of a modern-day Osler (common things are common). And in order for the system to function, it may be that a lesser-quality physician is the correct allocation for your case.

          • Edward Scizorhands says:

            The former isn’t very nice. The latter is downright fancy and caters to every patient request. Outcomes are consistently better among the poor than among the rich.

            This would have been great in the discussion we had about Robin Hanson’s “healthcare spending doesn’t improve health.”

            I can totally believe your anecdote is true. It might also be rational. The graphs in section 1.1 here [1] show that some places need more medical spending, and “the poor of India” are surely in that zone, so even minor spending is likely to get huge payoffs. Meanwhile, the richer people already have that low-hanging fruit picked, will not get anywhere near the same outcomes, and are paying — perhaps rationally, perhaps not — for comfort instead.

            I will agree that patients suck as consumers. Thread OP identified A) areas where markets work great, B) areas where markets would never work, and C) asked how we could identify which is which.

            Even though patients can suck as consumers, it is not an excusae to throw up our hands and say “it’s too much work!” Things can get better.

            First off, we can identify more things that should be pushed into category A. E.g., an MRI is an MRI is an MRI. They don’t care about outcomes. There are still legitimate reasons why someone might pay premium prices (such as location or off-hour/immediate access), but we should still be able to get quotes, not get blank stares when asking about the price [2].

            If people have bad ideas about what makes good health care — and I agree, they do — that is going to make problems whether the health care is negotiated in a group or individually. Helping people figure this out is going to be a really hard task, just like building a bridge is a really hard task. But not doing anything at all because it’s hard is the most expensive option.



          • albatross11 says:

            I’m thinking in terms of quality of doctors, defined in some ideal way. Like, there’s surely a bell curve of skill of surgeons, with some surgeons being amazing one-in-a-million talents and others being the guy who just barely got through his surgical residency and frankly should have gone into some other area of medicine. And similarly, for pediatricians/GPs, cardiologists, oncologists, etc.

            My question is, how important is it in most cases to have (say) a doctor from the right half of the bell curve. My sense is that for most stuff, it’s not all that big a deal. Maybe one way to look for that is to look at outcomes for patients of PAs, who have way less extensive training than doctors. If the patients do about the same then it’s probably not a big deal whether you get a great or mediocre doctor.

            What if you have cancer? How mcuh should you be worried about getting the best oncologist in town vs a randomly-selected one? Or what if you need some fairly routine surgery like having your gall blader removed? Would it be worthwhile to somehow try to get the best surgeon in town, or would any surgeon that did a lot of gall bladder removals be fine?

          • rahien.din says:

            @Edward Scizorhands,

            If memory serves, the rich patients had a greater chance of hospital-acquired infections, despite being in big private rooms and having probably better baseline health. So it’s not that the rich had less relative improvement, it’s that they endured greater absolute harm. (Would you link to that prior discussion? Probably too late to comment, but sounds interesting to read.)

            We can probably stop short of saying that patients suck as consumers… that doesn’t feel right. To the extent that patients suck as consumers, it is the physician’s responsibility to advise them. And to throw it back onto the physician even further : when money was no object and you could order whatever test you wanted, physicians did a really awful job deciding what to order, in terms of raw economics but also in terms of medical reasoning. So it all comes back to the fact that this system needs a great deal of improvement. We do have a tremendous amount of work to do.

            It does seem (maybe this is my only real contention) that the vicissitudes of the market are not where we should lodge much of medical decision-making. But you’re absolutely right, some things are clearly delivery-of-service. I have probably underestimated the amount thereof – see what I mean? On the whole, physicians aren’t great at this aspect of the healthcare system.


            You’re probably right. For most individual cases, it might not make too much of a difference. And common things are common – so the chance that a mediocre physician encounters a rare-disease-masquerading-as-a-routine-disease is pretty low.

            Also, your litmus test of “How often do you do this procedure / see this disease?” is likely to be reliable.

          • Edward Scizorhands says:

            The discussion about Robin Hanson is just one quarter-thread back:


    • MrApophenia says:

      As Edward Scizorhands mentioned below, a huge part of the problem with comparison shopping ERs is simply that a lot of the time they won’t tell you how much they will charge you in advance. Prices will not be provided until you get the bill.

      Vox did a project a couple years ago where they asked readers to send ER bills; I know a lot of folks here aren’t exactly fans of Vox, but the results were pretty interesting. The thing I found really striking was that the prices seemed almost arbitrary – how much a procedure cost at one hospital would have little relation to the same procedure at the next hospital over, with a procedure that costs a couple thousand dollars at one hospital costing tens of thousands at another in the same city, for no clear reason anyone will explain. The main unifying factor being that nobody knows what the cost will be before they get the treatment.

      Here is a collection of the articles they wrote based on it:

    • Corey says:

      There are many more difficulties in care shopping other than urgency.

      My mantra: in any healthcare situation, you have to agree to unlimited financial liability in advance; “market” does not seem like a good term for it.

      Even some services that seem eminently shoppable (e.g. MRIs) can have pitfalls (how well does their EMR integrate with the reading doctor’s EMR and your PCP’s EMR? Will there be hidden costs somewhere in the chain, e.g. a separately-billing professional who will of course be out of network? Is there actually a difference in quality?)

    • Garrett says:

      Estimates of ER spending of healthcare in the US I find credible range from about 2-4%. Pick whatever your preferred number is. This is the *upper* bound of non-comparison-shopping options.

      When it comes to comparison shopping, there are levels involved. For example, for something completely elective (eg. cosmetic surgery) you might care that whatever provider you pick can get it done on a Friday so that you can recover over the weekend and be back at work on Monday, if appropriate. In other cases it might be a recommendation from someone you know, or merely “accepts my insurance”.

      Of that 2-4%, you need to evaluate how much of that is truly emergent and time sensitive. Surprisingly, not a lot. A lot of ER visits are for things which could be handled at an urgent care if it were available 12 hours/day or if they could be seen there without paying. For the people actually paying, these could be cost-compared as well if there was a reasonable way to know in-advance what someone was going to need to have done. Hospitals definitely could make this more transparent but haven’t.

      Even with cases like abdominal pain needing surgery such as your friend it’s still possible to select a facility based on more limited criteria such as reviews (as opposed to general price shopping). For my patients with a suspected heart attack I can take them to a handful of different facilities with immediately-available cardiac cath labs if they have a preference. Unfortunately, for several reasons including law, I can’t advise a patient to go to one facility over another other than for specific medical need. One of the larger hospitals it’s really hard to bypass is a hospital I generally dislike and I would personally avoid if possible.

  9. slatehunter says:

    Hi, I’ve been hunting for a psychiatrist in San Francisco for months and I haven’t found anyone who can make an appointment in the next 3 months. Is there some secret way to find a psychiatrist? At first I was carefully reviewing people to find a good match, but now I just want to get any appointment so that I can try to get help, as I’ve been steadily deteriorating mentally. I don’t know if I can make it 3 more months. It was hard enough to get to here. I’d been trying therapy, but 12 months of therapist hunting yielded exactly one therapist who can’t even see me reliably or regularly. I’m hoping that finding some medication help will proffer enough stability that I can start to find my way out.

    • cassander says:

      does menlo park count as san francisco?

    • Corey says:

      I think it’s better to use psychiatrists for medication management, and people with other credentials for talk therapy. They can use brain-space for psychology stuff that psychiatrists have to allocate to doctor-y stuff.

      Also, depending on the urgency and nature of the issue, a primary care doctor might be able to hook you up; if it’s anxiety or depression they can score you Lexapro or Wellbutrin etc. (My wife manages Lexapro through our PCP, I have a psychiatrist and use other meds).

    • mingyuan says:

      I got a referral to a psychiatrist’s office via my primary care doctor, and someone was able to see me within a week or two. (I was in a similar situation (and also in SF) and getting meds really really helped). So if you see any sort of doctor at all, I recommend asking them for a referral to an in-network psychiatrist (though depending on your financial situation, it sounds like you’re desperate enough that out-of-network might not be a deal-breaker). I don’t know much beyond that, sorry. Best of luck!

  10. AG says:

    Can one weaponize a monkey’s paw/malicious genie by wishing good things for one’s enemies?

    Or do the curses involved here sense the underlying intent, and always grant wishes to maximally fuck the wisher over?

    • Nick says:

      You could contrive to put the monkey’s paw in your enemy’s hands. The worst that could happen is they don’t use it.

      • Randy M says:

        The worst that could happen is you were wrong about the intent of the genie, it isn’t so hard to get what you want, but you were just unusually inarticulate.

      • The Nybbler says:

        Worse: The enemy uses the paw to harm you. This harm backfires and causes much harm to your enemy, but you’re still harmed.

        • Lambert says:

          Put it in the hands of another enemy of yours, who dislikes the original enemy more than they dislike you.
          Sounds like the perfect way to deal with the Middle East.

        • Nick says:

          Depending on relative harm, this might not be a bad deal. You get caught in the rain, his house floods, for example.

          • The Nybbler says:

            You don’t know the deal ahead of time. Maybe you get pneumonia and die from being caught in the rain.

          • Nick says:

            But the paw does, doesn’t it? Seems perverse if the paw’s backlash on him isn’t as bad as the consequences for you. I guess the paw might not consider consequences sufficiently far down the road, though?

          • The Nybbler says:

            I’m assuming the paw is largely indifferent to consequences to those other than its owner; mainly it wants to make the owner much worse off than he would be had he not used the paw. Also on the first two uses the paw will certainly do things that apparently help the owner in the short term, to encourage him to keep using it. So perhaps the owner kills me (who he hates) by catching me out in the rain. This floods his house, but he only realizes later (after using up the paw) that it also destroys papers he’s going to need in defense of a lawsuit that leaves him destitute.

          • Edward Scizorhands says:

            You guys remember that Homer gave the monkey’s paw to Ned and it worked great for Ned, right?

            If the monkey paw has chosen to screw you, the only winning choice is to lock it away forever.

    • ARabbiAndAFrog says:

      Any genie worth a damn would notice what you are doing.
      You: “I wish my enemy had one million dollars”
      Genie: “Done”
      You: “What’s the catch?”
      Genie: “Your enemy now has million dollars”
      You: [On your knees screaming] “NOOOO! Oh the cruel irony!”

      • Iago the Yerfdog says:

        More like you end up in a car crash with your enemy, break every bone in your body, are found at fault, and they are awarded a million dollar in damages that you now have to pay. (Or if you can’t pay that much, they get a large award — still at your expense — that they subsequently turn into a million dollar investment.)

    • helloo says:

      By default, the story can contrive itself to be whatever it wants, but also that having a bad thing happen to you doesn’t mean that it might be good for you in other ways.

      Some examples-
      Simpsons had an episode with the monkey’s paw that had Homer giving Flanders the paw as Nick’s suggestion, but it working out more or less fine.
      Touhou has a character who is cursed to have all her words proven false (ex: says tomorrow will be sunny, might cause it to instead be cloudy or raining), who uses this as a source of power as much as a curse.

      Both of those are somewhat mixed (but can be easily changed to be one example or the other) as the first one did undo world peace and nonviolence, and the second, the curse is still a curse and doesn’t need to follow her intentions or logic (it might rain fire instead).

    • cassander says:

      I believe the traditional chinese formulation of this problem is the curse “May you live in interesting times.”

  11. Urstoff says:

    A question like “is technological progress slowing down” seems perfectly suited for an adversarial collaboration given the ambiguity of the term “technological progress”.

  12. proyas says:

    Millennials are the “Echo generation” of the Baby Boomers. Given Millennials’ low birth rates, can we safely say that there will not be an Echo-Echo generation, meaning no U.S. population youth bulge thanks to Millennials having kids?

    • EchoChaos says:

      That doesn’t hold, I don’t think. The generation after the Boomers was Gen X, which was a period of noticeably lower birth rates and a lack of a youth bulge. We won’t know if Millennials will have an Echo-Echo until Gen Z starts having kids.

      • RalMirrorAd says:

        I’d assumed ‘echo generation’ meant the generation created from the baby boomers having kids.

        • anonymousskimmer says:

          Yes, and you won’t know if Millennials child bearing is a relative echo until you see how many grandchildren Gen X has.

          Millennials being an “Echo boom” is relative to the relative paucity of Gen Xers to Baby Boomers. And Baby Boomers being a boom is relative to the numbers of both the Silent Generation and GenX.

          BB: 12345
          GX: 1234
          MI: 123456
          GZ: 12345
          AA: 1234567
          AB: 12345678 (Millennials do not have an Echo boom, as GenX’s grandchildren are a larger cohort)
          AA: 1234567
          AB: 123456 (Milliennials do have an Echo boom, as GenX’s grandchildren are a smaller cohort)

      • Plumber says:

        Part of this is that the Baby Book was so long (19 years), and “Millennials” are mostly the ‘echo’ of younger ‘Boomers’ (and some older X’ers), not of the order Boomers who were young adults in the ’60’s (my Mom’s generation) which are what we usually think of as ‘Boomers’.

        And on that note I found some interesting (to me) differences in partisan leanings between older and younger Boomers and older and young Generation X (and between the very tail end of ‘Silents’ and the rest of the ‘Silent Generation’) which in case it’s a “hot-button” topic I’ll share in the 142.25 thread, suffice it to say that if you go by the average political leanings of folks born in the same year the usual generational demarcations are a bad fit

        • EchoChaos says:

          Interesting. In my mind, all generations are ~18 years long, so that doesn’t seem exceptionally long. It fits with the typical patterns too.

          That puts the Boom from 1945-1964, Gen X from 1965-1982, Millennials from 1983-2000, Gen Z from 2001-2018.

          People might fight the exact years, but those definitions make sense to me.

          I agree that “early and late” cohorts exist for all of these generations.

          I look forward to a discussion of it in .25

      • Aftagley says:


        I mean, you would be the expert on this.

    • Urstoff says:

      Why are there consistently more men than women in each age group until age 40? Is that a biological thing (slightly more eggs will be fertilized as male than female) or a social thing or a political thing (immigration, perhaps)?

      • Mostly biological, but disproportionately male immigration does contribute.

      • fibio says:

        Men have a slightly higher birth rate and a slightly higher death rate than women. It’s about 1% either way but is very visible if you look an an entire population.

        Unless you’re talking about China. In that case the answer is Communism.

      • DarkTigger says:

        Men have higher birthrates even in countries that are not under suspicion of aborting female fetuses regularly.
        Men also die more in car – and work accidents and as victims of violent crimes.
        If they are also biologically more susceptible do medical condition, or if that is an social thing, is a question not fit for an integer OT.

  13. bean says:

    Biweekly Naval Gazing links:

    Naval Gazing celebrated the 50th anniversary of Apollo 12 by looking at the Navy’s participation in the space program.

    In my continuing look at modern weapons, I’ve now covered the Harpoon family, the west’s leading anti-ship missile for most of the last 4 decades.

    The Falklands is back with Part 19, covering the last days of May.

    Lastly, riverine warfare turns to Southeast Asia, with a look at how the French used it to conquer Vietnam.

  14. johan_larson says:

    The guys at Shut Up & Sit Down play and review board games.

    A few years ago, they listed their top 25 games. Here are the top 10.
    1. Cosmic Encounter
    2. Memoir ’44
    3. Android Netrunner
    4. Two Rooms and a Boom
    5. Sherlock Holmes Consulting Detective
    6. Skull
    7. Tales of the Arabian Nights
    8. Galaxy Trucker
    9. Twilight Imperium
    10. Race for the Galaxy

    It’s quite an idiosyncratic list. Their top 10 overlaps with Board Game Geek’s top 10 at exactly one game: Twilight Imperium.

    • EchoChaos says:

      Well, reading that definitely made me want to play those games, especially with my non-hypothetical kids.

    • Machine Interface says:

      That top 10 is quite American-style centric, but looking at their full top 25, it does also include a few eurogames so there’s that.

      I’m likely getting Twilight Imperium for Christmas.

      • Aftagley says:

        SUASD is, if not anti-eurogame, somewhat skeptical of it as a form of entertainment. They tend to enjoy more directly interactive games.

        • MorningGaul says:

          Could you explain what are the difference between euro and US boardgames to a fairly casual european boardgame player?

          • fibio says:

            Generally, it boils down to a fine distinction between a game being competitive (Euro) and being combative (US).

            For example, Splendor is a classic Euro game of building a tableau and getting victory points. You’re basically racing, and while you can make some moves to spite or slow your opponents, most of the time you’re focused on optimizing your own path to victory.

            On the other hand, something like King of Tokyo is a classic US game where, while you’re still trying to hit a victory point total, but most of the game is focused on other players and their actions. Unless you thrown in the expansion you’ll spend very little focus on your own player mat.

            Of course, its a spectrum not a scale so all games are at least a little Euro and a little US.

          • Machine Interface says:

            What Fibio said is one way to put it, but another is that American-style games emphasize theme over mechanics, whereas eurogames emphasize mechanics over theme.

            The prototypical American game is all about creating a specific atmosphere and narrative experience, with lots of small, specific rules each with a very specific thematic justification. The result might gets a bit clunky rule-wise, but will creating a lasting impression. Typical US games include any dungeon crawl or RPG-like game you can think of (think Heroquest, Space Hulk, Descent, Arcadia Quest), games with “episodes” and a built-in narrative nature (Arkham Horror, Mansions of Madness), 4X games (eXplore, eXploit, eXpand, eXterminate — think Twilight Imperium, Dominant Species, Runewars) and more generally games that simulate multiple aspect of a conflict, such as diplomacy and logistics (Dune, Civilization, Diplomacy, Cosmic Encounter).

            The prototypical Eurogames starts with a well defined and harmonized set of mechanics, and the theme is added almost as an afterthought. These are games whith an elegant and smooth system that plays very well, but might not necessarily create individually memorable narratives/sessions. Typical Eurogames include worker placement games (Caylus, Agricola, Viticulture, Russian Railroads), set collection & tableau building games (7 Wonders, Sushi Go, Hanabi, Jaipur), resource collection & engine building games (Century: Spice Road, Wingspan, Terraforming Mars, Powergrid, Space Base) and tile laying games (Carcassonne, Glen More, Azul, Kingdomino).

            A classical difference is that US games are often likely to use a lot of dice, whereas most Eurogames use none at all, but this has started to become less true.

            This is of course a spectrum and nowadays more and more game tend toward hybridization. Scythe, Rising Sun, Twilight Struggle or Pax Pamir have all been described as game that combine the strong thematic elements of American games with the “holistic” approach of mechanics of Euros.

            Finally, it should be noted that wargames are generally considered outside of this classification and a category of their own, as are increasingly often so-called “economic-games” like the 18xx series and Splotter Games (think Food Chain Magnate).

          • Paul Brinkley says:

            Yet another difference I’ve heard ascribed is just historical: Euro games are simply those that came out of a gaming renaissance in Essen, Germany, starting roughly with Settlers of Catan, and starting a new wave of tabletops.

            And another ascribed difference: Euro games were a push against games where the players knew who the winner was long before the game was officially over. E.g. Monopoly – once someone’s started building houses on the richest side, there’s still an hour or so of rent trading hands, gradually erring toward that player. This was considered obviously un-fun – most Monopoly games really ought to end in an early resignation, but the dice rolling meant that a win was still remotely possible, and this snagged a person in their seat like an old woman in front of a slot machine.

            This meshes with Machine Interface’s point about mechanics. Euro games often employ various tricks to make the winner unclear until the end – “take down the leader” mechanics, cooperative play, multiple victory point awards at the end that no one strategy could vacuum up, or game phases with increasingly powerful mechanics. This had to be balanced in longer games with mechanics that gave the player a feeling of pulling ahead. The idea was to make everyone feel like they had more of a chance, while still making skill feel rewarding.

            US designers quickly noticed the quality of game mechanics from Essen and reacted wisely with their own takes. Everyone was still making thematic games (the Euro side came out with Power Grid early, for instance); the US simply entered a cycle later. Overall, everyone just got better at designing games.

            War games are indeed their own thing, which I break out into minature-based games (e.g. Mechwarrior) and historical sims (e.g. Hispania, Churchill). Both appeal specifically to people who like what-if scenarios. The former are more tactical: “what if I arm all my dragon tanks with laser MIRVs?”. The latter are for “what if the Soviets ignored the Eastern Front and just rolled all their stuff into China?”.

            The player dynamics I see in wargaming are interesting. Assuming no one’s trying to be a munchkin, everyone enjoys a winning strategy that emerges midgame, and is acting like a post-game commentator by endgame while just going through the gameplay motions.

    • Matthew S. says:

      As someone who greatly enjoys SU&SD, but not infrequently disagrees with them, there are two things to understand about what they are looking for out of a game (especially notable with the departure of Paul Dean). (Also, note that that list is almost certainly out of date, since Quinns has not been playing Android:Netrunner actively in some time, and also was pretty unequivocal recently that he thinks Blood on the Clocktower, which just came out, is the greatest game he’s ever played.)

      1. Games for them are a vehicle for social interaction. That is not to say that they don’t play strategy games, but they want banter and laughter to arise naturally from the game.

      2. They place a lot of emphasis on thematic narrative. No matter how strong the mechanics are, if the theme is not well-integrated, they won’t like it. (E.g. Elysium, a very good game which they panned.)

      One thing I will give them credit for: They do review certain games as “likely to be amazing to a specific, narrow audience, but also a miss for most people,” even when they fall into the latter category. As several of the games they have treated this way are in my top 10 (e.g. Tramways, Pipeline), I appreciate this (they get into more detail and nuance in the podcast than the videos, fwiw).

      The BGG top 10 has a different issue. Although BGG uses Bayesian averaging to penalize games with fewer ratings, games that require an extreme investment in either learning/mastering or repeated campaign play (and/or are very expensive) only get played by the small subset of users that is likeliest to enjoy them, and thus get disproportionately high ratings (e.g. Twilight Struggle, Pandemic Legacy, Gloomhaven). Some others are just inexplicable; Terraforming Mars is just okay and does not deserve to be in the top 50, never mind the top 5; fight me.

    • johan_larson says:

      Things don’t get much better if we look at the top 25 each way. BGG and SUSD’s top 25 have exactly two games in common: Twilight Imperium and Terra Mystica.

  15. anonymousskimmer says:

    @Nancy Lebovitz

    Thanks. I checked All of an Instant out from the library based on your post. It took me close to 6 weeks to finally start it (the first couple of pages weren’t gripping) and forced myself through the first chapter, but then the end of chapter 1 “And from security, curiosity arises.” and I could barely put it down.

    This is one of the best books I’ve ever read (though this is still heady off of just finishing it). It’s like Flatland, but with a far broader and important scope, threats to the basis of existence, and the means by which humanity overcomes some of their own flaws to deal with these threats, and remake reality to protect reality.

    I am struck by how much the people who do venture into the instant of time truncate themselves to do so (limiting the selves they remember to a truncated span of time of their previous life, limiting their senses to typically one sense).

    That’s all for now. I have to let this settle for a while (and need to get to bed).

    I’ll be checking out Celestial Matters next.

  16. ADifferentAnonymous says:

    Anybody have info about whether comprehensive neuropsychological testing is useful? I mean aside from its role in gatekeeping of medications and accomodations, and particularly for adults.

    Accepting both anecdotal and scientific evidence.

    • Statismagician says:

      Could you give an example? I’m not sure what sort of thing you’re talking about.

      • ADifferentAnonymous says:

        It seems to be a specific field. I’d seen it before as an ADHD screening thing, but my partner’s psychiatrist brought up the possibility of a comprehensive test. To hear the testing center tell it, this builds a detailed profile of your brain: strengths and weaknesses, learning styles, etc.

        I’m asking here because it’s kinda pricey and only partially covered by insurance. The cost is something we can easily afford, but more than we can afford to spend on snake oil.

    • alwhite says:

      I once had a client with a disease based traumatic brain injury. I referred for testing to measure a baseline of ability so we could have a clear sense if improvement was happening. I thought the test/measurement in that sense was useful.

      I’ve seen the MMPI used for diagnostic clarification purposes. It’s sort of gate-keeping-ish but seems a little different.

      It’s hard to tease out the gate-keeping function. For example IQ tests. We might suspect a person has low-IQ and therefore needs services so order an IQ test to prove it. There is a gate-keeping aspect there but there’s also an intuition check. Maybe our intuition of a person is just wrong, they don’t have a low-IQ, have a different issue, and the test ruled that out.

      I would say every measurement has a gate-keeping function because decisions have to be made somewhere. How is measurement useful to you?

      • ADifferentAnonymous says:

        Yeah, trying to look this up I see a lot about using it for traumatic brain injury and dementia. My partner (the one considering the testing) has no risk of either of these. The psychiatrist brought it up in response to suspected ADHD, but expressed some interest in a comprehensive test rather than just an ADHD screening. There are other issues besides the ADHD, but as far as I know all things considered psychiatric rather than neurological (anxiety, depression, PTSD being the big ones).

        Don’t worry too hard about what is or isn’t gatekeeping, I just didn’t want replies like “they’re very useful… For getting extra time on tests” or w/e.

        Anyway, it sounds like you haven’t personally used these tests outside of traumatic brain injury but are somewhat optimistic about their value in other cases?

        • alwhite says:

          I’m unsure about the real value and whether they are worth the cost.

          Proper ADHD testing is kind of close to a full battery. They use pieces of all kinds of tests to detect ADHD. However, this is not done all that often and ADHD is frequently diagnosed without these tests. There’s a whole debate here about overdiagnosis, but that’s not what you are looking for. Maybe this is what the psychiatrist is thinking? You might need to get more clarity to what the psychiatrist actually wants.

          By the technical rules, everything in the DSM is diagnosable based on clinician expertise and no testing required. This doesn’t always hold up. There are additional tests for ADHD. Any disorder that leads to disability services usually needs more screening than that (autism has the ADI-R or ADOS for example). Some people use the MMPI to screen for personality disorders and malingering. You don’t have to have these, but they sometimes help to convince insurance or state agencies to pay out. These are all about proving a kind of disability though. Might not be the path you want. Maybe ask “who needs to know this?” Figure out who all the players and who needs convinced of what?

  17. benwave says:

    Can any savvy investor help me evaluate the claims made here: That active funds > monkeys > passive funds > passive funds claiming to be active funds?

    The author is someone I generally respect, but the claim is surprising and I don’t have the domain knowledge to evaluate it. I turn to you, Slate Star Commentariat! Help me in my hour of curiosity.

    • Douglas Knight says:

      There are two well-known long-term substantial deviations from the efficient markets hypothesis. One is the one mentioned in the article, that small companies do better than big companies. The other is that comprehensible “value” stocks do better than fashionable “growth” stocks.

      Investing in all stocks uniformly is the same as averaging over all monkeys. But actually using monkeys increases variance and is bad.

      Truly indexing means putting your money in the same place as the average money. Then you will necessarily get average returns. Putting your money in the S&P 500 is biasing towards big stocks, which is the wrong bias.

      • RalMirrorAd says:

        Are there any ETFs or indexed funds that account for this bias?

        • ADifferentAnonymous says:

          The usual practice is to buy a mix of index funds, rather than one fund that does it all. So you’d mix a total market fund, a small-cap fund, and a value fund, effectively ending up with holding some of everything for diversification but relatively more small company stock and value stock.

          • Edward Scizorhands says:

            A week or two ago, Matt Levine pointed out a funny hypothetical: a company is at the edge between the Russell 1000 (the top 1000 stocks, more or less) and the Russell 2000 (the second top 1000 stocks). If it falls from position 999 to 1001, the Russell 2000 buys it up, potentially pushing it into the Russell 1000, causing the Russell 2000 to sell it off. Possible endless loop.

            Natural market noise plus fundamentals probably outweighs this, as well as the Russell 1000 being a thing, but if investors “rationally” put significantly more money into the Russell 2000, we can get a weird dynamic going on.

          • anonymousskimmer says:

            @Edward Scizorhands
            Aren’t the two indices operated by the same company? So why wouldn’t it just trade ownership at parity? You’d likely have some excess, as the 1000 and 2000 are unlikely to be purchased exactly proportionally, though as of now (unless I’m misreading this) the 2000 seems to be below the 1000 proportionately, meaning that a switch from the 1000 to 2000 would actually cause a small sell-off with consequent loss in market cap?

        • Chalid says:

          Google “factor investing” or “smart beta”

    • Chalid says:

      Seems ok to me.

      Market-cap-weight is more large-cap oriented than equal weight, sure, and small caps have historically done a little better than large caps, so that gives you monkeys outperforming passive. (Though do they on a risk-adjusted basis after taking into account the higher transaction costs of investing real money in small-caps? Not so clear, and also keep in mind that maintaining an equal-weighted portfolio requires constant trading.)

      He’s not saying that active funds are better, in aggregate, than passive funds. He’s saying that the particular subset of “truly active” funds is better. Also plausible; active funds in general are definitely dragged down by closet indexers, so the “real” active funds are better than the average active fund.

      (Also, it is definitely the case that the average active fund outperforms before fees.)

      I’m not sure I believe the claim that large-caps are more heavily passive-owned than small-caps. This is checkable but I’m not going to do it.

      So anyway, I wouldn’t say I believe the article, but nothing seems obviously wrong to me at a glance.

    • Artischoke says:

      The relevant question to ask is “do active funds outperform passive funds after subtracting their higher fees”. Frankly I dont care if the average active fund manager maintains a slight edge over a cheap index fund. I assume that the average difference in performance between monkeys, active managers and index funds is miniscule compared to the difference in fees.

  18. eigenmoon says:

    This Mark Zuckerberg’s interview of Tyler Cowen and Patrick Collison on progress might be worth a link alongside the link to the paper. From the interview:

    COWEN: Here’s what worries me, and it should worry you too. So as you mentioned, U.S. life expectancy is basically going up in linear fashion. But if you look at expenditures, we used to spend a few percentage points of GDP on healthcare, and now it’s about 18%. So we’ve gone up to 18%, and we’re not even boosting the rate. I’m not saying it’s the fault of any one group of people, but something has gone wrong. There’s some kind of last-mile problem. You can turn to the newspapers and read all kinds of fantastic stories — new research, new ideas, new tools — but when the rubber hits the road, people living longer, we’re spending more and more and more for exactly the same returns. So if that trend continues — and you see a similar trend in many areas: crop yields, feeding the world, other areas — the question becomes, where does all the progress go?

    • Edward Scizorhands says:

      I don’t think it’s a rational choice — either by individual actors or a central planner — to spend so much money on health care. But health care is tax-advantaged for no sane reason, and you get more of what you subsidize.

    • RalMirrorAd says:

      RandomCriticalAnalysis did a fairly long post on his own website on this topic.

      At a certain point the chronic health problems which are partly a product of high living standards will outweigh what medical science can achieve.

      • TJ2001 says:

        The other obvious outcome is that “The Government” balks at the cost spiraling out of control and simply declares the problem “Not a problem” or simply not feasible to “cure”..

        Then they put the focus on palliative/end stage care such as making people comfortable, reducing suffering, and allowing people to die quickly with dignity – while intentionally taking action not to prolong suffering or necessarily doing anything to make people live longer at massive expense…

        So for example – develop pain killers that don’t have nasty side effects, are minimally addictive, and don’t make people excessively loopy so that people don’t suffer on and on and on as they die of end stage conditions….

        • fibio says:

          So for example – develop pain killers that don’t have nasty side effects, are minimally addictive, and don’t make people excessively loopy so that people don’t suffer on and on and on as they die of end stage conditions….

          We’ll get right on that just as soon as we’re done with fusion…

  19. Freddie deBoer says:

    This is one of the most entertaining and important academic papers I’ve read in a very long time.

    • Aftagley says:

      So it’s my third grade teacher’s fault I couldn’t break 6’2″?

      Thanks for nothing Ms. Allison.

    • Clutzy says:

      Pssh. That paper is so 141.5.

    • TJ2001 says:


      I think it points out the reality students, their parents, and every college student ever has to deal with every day…..

      Some students learn because of the teacher….
      Some students learn in spite of the teacher…..

      But either way – the student is the one harmed if they don’t learn – not the teacher… The family of the student is also likely to be harmed by the poor learning outcomes… And everybody knows this…

      So for example.. In a study like this it is very hard to control for confounding variables such as parent/family involvement or student drive…

      One interesting thing I am running into (with my own kids in elementary school) is that teachers who are good at fostering parent involvement yet who not necessarily good in the classroom often have students that outperform “Good” classroom teachers who are not as good at keeping parents engaged…

      • RalMirrorAd says:

        If parental involvement as such was so decisive you would expect it to be part of the shared enviromental effects on academic performance. My default suspicion is that any measures of visible parental involvement are proxies for behavioural interest in academics. (Much like the effects of reading to children).

  20. Freddie deBoer says:

    That’s a biggie in the Mistakes column!

  21. LesHapablap says:

    What are the pros and cons of taxing businesses based on their revenue instead of profit? Is there a consensus among economists that it is a bad idea?

    • thevoiceofthevoid says:

      Con: You heavily disincentivize any endeavor with small profit margins. If your tax rate is higher than the profit margin, you can turn a profitable low-margin business into one that’s losing money.

      In general, you create a huge incentive to cut expenses as much as possible, even (especially?) at the expense of decreased revenue. This seems to me like it would lead to vastly decreased economic activity, though I’m not 100% sure (I’ve only taken one economics course).

      • thevoiceofthevoid says:

        For example (toy numbers): Let’s say your potential business would have $99/year expenses, and $100/year revenue. With no taxes, the profit would be $1/year. Thus, this business adds value to the economy (let’s assume no nasty externalities or anything for simplicity), and it’d be good if it opened/remained open. A revenue tax of only 2% would turn that $1/year profit into a $1/year loss. In contrast, as high as a 50% tax on profit would still leave you with a $0.50/year profit; no profit-tax less than 100% would make a profitable business unprofitable.

    • The Nybbler says:

      Note that some businesses are taxed on their revenue, notably retail businesses (sales tax). (And then their profits, if any remain)

      • LesHapablap says:

        Is sales tax a value-added tax or is it straight off the revenue?

        • thevoiceofthevoid says:

          No, it’s a straight usually 5-10% of the cost of the product in the US (varies by state and sometimes by city). The ridiculous part is that it’s paid by the customer (from what I gather, not an economically important distinction) but NOT included in the listed price. So, you’ll go to buy something marked as $5, then you’re rung up and have to pay like $5.32 — it’s annoying for budgeting and makes paying in cash awful.

          • Statismagician says:

            Every country I’ve spent time in except the US does include sales tax in posted prices. I’ve been told that some people like having the tax separated so they know how much they’re paying over-and-above the cost of the item, but I think this just proves those people don’t use cash very much and think about tax rates more than I do.

          • thevoiceofthevoid says:

            some people like having the tax separated so they know how much they’re paying over-and-above the cost of the item

            Sounds pretty bonkers to me, especially since you don’t actually know until after you’ve paid. If I want to think about tax rates, I’ll look up the tax rates. And despite paying them multiple times a week for the past two years, I couldn’t tell you off the top of my head what the sales tax rates are for my current place of residence.

          • albatross11 says:

            My understanding is that in many states it is actually illegal to roll the sales tax into the full listed price.

          • CatCube says:

            …especially since you don’t actually know until after you’ve paid.

            Huh? Maybe it’s radically different somewhere else, but the sales tax is a fixed percentage, and you can easily estimate tax as a number of cents per dollar (in Michigan, 7%, so if you have something that’s $1.80, you can easily estimate $0.14 tax; guaranteed to be a bit high, but enough that you can get cash out before reaching the register). My mom taught me how to do this in like 4th grade, and we got it in school, too. It’s really not as hard as you’re making it sound.

          • Statismagician says:


            What? That’s nuts, why would anybody legislate that?

          • thevoiceofthevoid says:

            Well, yeah, you can calculate it, but that’s an extra mental step and not an automatic one (for me at least. Perhaps I just have rely-on-calculator-itis). And you could also do the calculation if tax were included to find the portion of the price that was tax. To me, it’s an annoyance in a similar category to pricing things at e.g. $4.99; I’m intelligent enough to realize that the actual cost of the product is essentially 5 dollars, not 4, but it takes System 2 conscious thought and makes it more difficult to get a good sense of price by glancing at the tag.

          • Lambert says:

            It’s not like displaying prices both with and without tax is hard.
            Everywhere in the UK that sells to both trade and normies does it.

          • ECD says:

            I’m not aware of any law preventing including it in the price, so long as you state ‘tax included’. In my state at least, you’re allowed to do that.

            Almost no one around me does, at least on standard transactions. I’m uncertain why, but my guesses follow:

            1) It lets you say your price is lower than it actually is.
            2) It means you don’t have to adjust prices based on location (unlike most VAT jurisdictions, sales tax varies town-to-town and state-to-state).
            3) It means you have to add something, so folks who are exempt can just make sure something isn’t being added.
            4) It means the tax isn’t being hidden (my guess for why some place might, allegedly, ban ‘tax included’ pricing).
            5) Sales tax actually does change relatively frequently, you don’t want your profit margins (or markings) to have to change with it.

          • pjs says:

            @Lambert “It’s not like displaying prices both with and without tax is hard.
            Everywhere in the UK that sells to both trade and normies does it.”

            I thought UK sales tax was collected as a value added tax – am I wrong? So why would they make any distinction between trade and final consumers?

          • ana53294 says:

            When we changed to the euro, for a couple of years, all prices were indicated in both the euro and the peseta. The euro price was big, the peseta price was smaller.

            They also indicate the price per kg – in smoll print also.

            I don’t see why indicating the price with the sales tax would be that much trouble, unless it’s banned.

          • Lambert says:

            VAT is collected as the product or service is sold to the end user. The costs then flow backwards along the supply chain.

            When a restaurant buys a frying pan the value added by the manufacturer, retailer etc. gets taxed when a customer pays VAT on their meal.

          • RalMirrorAd says:

            WRT listing sales tax, if sales tax incidence falls 100% on consumers then estimating the sales tax is easy enough.

            But i imagine Tax incidence won’t be the same between firms/industries that pay the same rate and also won’t be the same at different rates.

            So the only legitimate rationale i can see for prohibiting showing what the effect of the tax is, if a storeowner claims that customers are paying 100% of the tax and the sales tax gets reduced to, say, 0%, would the storeowner be obligated to cut prices as much as their claims would imply?

          • anonymousskimmer says:


            Those calculations require knowing in advance any deposit taxes, sin taxes (which do seem to be rolled into the price), and exceptions to sales tax (certain foods and other necessities).

        • Eric Rall says:

          Retail sales tax is based on straight-ish revenue, at least in the US. There are some goods that are exempt from sales tax: the exact list varies by juristiction, but to a first approximation groceries (but not prepared food), prescription medications, and goods purchased for resale are exempt from sales tax.

          And the tax is framed as being paid by the buyer, with the seller just being responsible for collecting it. Economically, this is a distinction without a difference, and the main practical consequence is that goods are typically tagged with the before-tax price rather than the full out-the-door cost.

          The net result is pretty close to a VAT on goods (but not services, since US sales taxes are almost always just on goods), since only the final sale for consumption is taxed. The main differences are:

          1) Retail sales taxes are easier to evade, since you only need the last firm in the supply chain (the retailer) to commit tax fraud to evade the entire tax, whereas a VAT in that situation would still collect taxes on the wholesale price.

          2) Retail sales taxes are more visible to consumers: you get a line item on your receipt for the sales tax the retailer collected from you, and you see the difference between the posted price and what you actually paid. I have much less firsthand experience with VATs, but from what I gather they are often buried in the list price. This allows slightly higher tax rates to be politically feasible, which can be regarded as a pro or a con depending on your perspective.

          3) VATs collect revenue sooner in the process, shifting some capital costs from the government to firms in the supply chain.

          4) Sales taxes go 100% to the jurisdiction the purchase occurs in, and can trivially be evaded when the buyer and the seller are located in different tax jurisdictions (e.g. a mail-order retailer in Ohio has no obligation to collect Ohio taxes on an order from California, California has no right to collect taxes from the Ohio firm, and laws requiring Californians in this situation to report the purchase on their personal tax returns and pay taxes to California are honored more in the breach than the observance). Even when it isn’t evaded, a manufacturer in one state yields no sales tax revenue for things they sell through retailers in other states.

          VATs often approximate the same distribution of revenue through “border adjustment” provisions (collecting VAT on imports and refunding total VAT for exports), though. And I suspect this is part of why VATs never caught on at the state level in the US: border adjustments on interstate trade are unconstitutional due to provisions forbidding internal tariffs.

          • Theodoric says:

            a mail-order retailer in Ohio has no obligation to collect Ohio taxes on an order from California, California has no right to collect taxes from the Ohio firm

            I thought that SCOTUS ruled that states could collect sales tax on items shipped to them, even if the shipper had no physical presence there.

          • Eric Rall says:

            I just checked, and you’re correct: my understanding was out-of-date. The old precedent I already knew about was from Quill v. North Dakota, which required a “Physical Nexus” (a storefront, warehouse, sales office, etc) in the state attempting to collect the tax. It seems this was overturned in the 2018 case South Dakota v. Wayfair.

            Interesting split in that case, too: it was a 5:4 decision, but with Ginsburg joining Thomas, Alito, Gorsuch, and Kennedy in the majority.

        • Douglas Knight says:

          It’s VAT in most of the US, but it compounds in Hawaii, encouraging vertical integration. I probably learned this from David Friedman. Maybe other states, too, but at much lower levels with less distortionary compounding.

      • I should note that this only works because of the sales tax only taxes finished products. If you replace sales tax with a revenue tax, the retailer would pay the tax once when he sold the product and again, indirectly, when he bought it from the wholesaler.(And then again if the wholesaler bought from a farmer, ect.)

    • Lambert says:

      Sounds like a great way to get unhealthy levels of vertical integration.
      Which will make markets far harder to break into, and therefore decrease the overall level of competition in the economy.
      As well as the fact that there’s a reason why Apple doesn’t operate lithium mines.

      • Viliam says:

        Businesses of the type “take existing components; combine them / add a minor improvement / move them to another place; and sell the result for a slightly increased price” would have to choose:
        * disappear;
        * increase their price a lot;
        * get vertically integrated.

        In general, it would give large companies an additional advantage over small companies.

    • Erusian says:

      There are some states that have revenue based taxation. The pros are that it’s harder to reduce your tax burden and generally raises more money even at lower rates. The cons are that it’s more distortionary for corporate decision making, it makes it much harder to start a business, it discourages low margin industries (which tend to disproportionately owned and staffed by poor/marginalized groups), and businesses really don’t like getting tax bills when they’ve lost money. It also encourages corporate consolidation, increases the power of large corporations and decreases the power of small companies or freelancers, and discourages complicated supply chains.

      It’s one of those policies that the Chamber of Commerce doesn’t like but neither does the AFL-CIO or Teamsters really want it. In a lot of cases, it’s effectively been a way to backdoor in a sales or corporate income tax in red states that don’t want to admit they have them. (Of course, in California it’s just that the state has never met a tax it didn’t like.)

      Also, honestly, I think a lot of people intuit that if a business spends $75,000 to get $50,000 in revenue, taxing them on that revenue is a little strange. Sure, they made $50,000 but they lost money overall. There are cases like where the ‘loss’ is because of a huge salary paid to the owner. And that’s a debate worth having. (I’ve seen proposals for an owner benefit instead of profit based taxation system.) But it’s not a great case for switching to revenue based.

      • thevoiceofthevoid says:

        But then wouldn’t that huge salary get taxed as the owner’s income? Though I guess then it’d go to the Feds rather than the state.

        • Erusian says:

          It would get taxed as personal income, yes. But corporate taxes have always been a form of double taxation. That’s why the state has restrictions on owner compensation through salary but not through dividends etc. Because they get more money that way because they take two bites: as corporate profits and then when the person receives the profits as income.

      • Artischoke says:

        >The pros are that it’s harder to reduce your tax burden

        Yeah the fact that revenue is a more robust figure than income is the main benefit I can think of. Say you are willing and able to manipulate your reported revenue by 5% through creative accounting, tax fraud etc. While this would only have a minor effect on your revenue tax bill, it might wipe out your reported profits in their entirety if you can do this without affecting your costs. In this sense a revenue tax is better if you have little trust in the reported numbers, if companies finances arent very legible to the tax authority so to speak.

        For the same reason a revenue tax would also be less responsive to economic boom-/bust cycles – though if you are in favor of a countercyclical fiscal policy that is actually another argument for an income tax over a revenue tax.

        What I dont have an answer for is why anybody would prefer a revenue tax over a VAT (unless you want a progressive revenue tax because you like small companies better than big ones)

    • RalMirrorAd says:

      Taxing revenues instead of profits would require any solvent business to have pre-tax profit margins in excess of expenses + taxes, so a walmart business model where you try and sell things cheap and focus on volume would be impossible.

    • Urstoff says:

      Why tax businesses at all?

  22. TheContinentalOp says:

    For SciFi and Mystery fans in the DC area:

    We’re having a book launch party for Wildside Press’ Crime Travel anthology: 15 tales of crime and time travel.

    Barnes & Noble
    12193 Fair Lakes Promenade Drive
    Fairfax, VA 22033

    Sunday, December 8th, 1-3pm.

  23. S_J says:

    Today I learned that the tusk of the narwhal was often sold as a “unicorn horn” in Europe, in the time frame between roughly 1100 and 1600.

    An article about this practice dubs it one of the biggest hoaxes of history. I’m not sure whether it counts as a deliberate hoax, or a misunderstanding based on incomplete information.

    The trade most likely began shortly after Erik “the Red” Thorvaldsson settled in Greenland. He and his followers used whale-blubber as a source of food, alongside fish and whatever food they could raise on the arable land in Greenland. One source of whale-blubber was the narwhal. [1]

    The settlers in Greenland didn’t have much else to trade with the outside world. But they discovered that the distinctive spiral-shape of the narwhal-tusk made it an intriguing item among the wealthy and powerful of Europe.

    Books about wild-beasts-of-the-world had long included one-horned land animals among the strange creatures of far-off lands. Often, these bestiaries claimed that the horns of the unicorns had special properties. The settlers of Greenland may, or may not, have heard of unicorns. But their trading partners in the rest of Europe would likely have called narwhal unicorn of the seas, and were happy to sell narwhal-tusk as unicorn-horn.

    Thus, people from Ivan the Terrible to Charles the Bold (Duke of Burgundy) had narwhal-tusks among their collections of jewelry. The throne of Denmark has narwhal-tusks as part of its structure, and the jewels of the Hapsburgs included a scepter made from a narwhal-tusk.

    All of those items were described as “unicorn-horn” items when they were made.

    I’m a little surprised at this sequence of events: a bunch of settlers from Iceland try to make a new home in Greenland. They discover a local food-source with a unique tusk. They (or intermediaries in trade) sell the tusk as unicorn-horn. And the distinctive shape becomes part of the image of the unicorn. In the course of time, unicorns are understood to be nonexistent. But the spiral-horn-on-a-horses-head persists as the image of the unicorn.

    [1] Likely, the Greenland settlers saved themselves from scurvy by consuming whale-blubber. A similar diet was used by the Inuit people who moved into Greenland during the same century that Erik and his followers settled there.

      • John Schilling says:

        No, that was actual unicorn horn. Guy just declared it as narwhal tusk because unicorns are a protected species.

    • Douglas Knight says:

      And the distinctive shape becomes part of the image of the unicorn.

      Is that your own conclusion? I don’t see it explicit in your sources or anything else I dug up. It seems pretty plausible, but it would be nice to have someone say “I looked at all the ancient sources and none of them describe or depict a spiral.” It would be even nicer for someone to go through the medieval depictions, but that would be a lot more work.

      • I’ve actually seen a unicorn. You take a goat, transplant the horn buds early on to the center of the forehead, and they grow together as a single horn. I’m not certain, but I think it is a spiral.

        I don’t know if that method was used prior to modern times, but I don’t see why it couldn’t have been, and it could provide an alternative source for the spiral.

      • S_J says:

        That is implicit in one of the articles I linked to. I may have turned it into an explicit statement myself.

        Not sure it’s the only available explanation.

        Before today, I hadn’t seen any attempt to explain the spiral-patterned unicorn horn. @DavidFriedman provides what might be another explanation. I’m not sure whether that was ever done, or seen [EDITED TO ADD: before the unicorn-spiral-horn imagery became dominant].

  24. ana53294 says:

    I disagree with the idea that this decade, progress has stopped. It just takes a bit of time for the science to transform into practical technologies, then more time for the technology to scale up, and then for social change based on that technology. It’s more or less what happened with electricity for lightning : first it was discovered, then the lightbulb was invented, then the whole country started to use it, and then people were able to change their lives, working in better conditions, etc.

    There are many things that scaled up during this decade:


    Sure, you could order things in Amazon ten years ago, too. But you didn’t have the scale, or convenience, or the low prices you have today. And while for city dwellers online shopping isn’t that much, for country living, online shopping means we get to access a lot of the things city people get. And that wouldn’t be possible without the scaling up of warehouses, and the management of all those things.

    In Africa, you have things like Uber for trucks, only profitable and beneficial for drivers. It helps remove the middleman in Africa, where hauling was quite hard. It also made things better for truckers, since they get better paid, their trips are financed, and they don’t have to carry cash around.

    The electric cars are just scaling up, becoming practical for middle class people.


    Even in Europe, where we are quite happy with credit/debit cards and cash, thankyouverymuch, and paying by phone is still exotic, there have been many practical improvements in finance. Sure, I could go to Sweden ten years ago and buy things with my Spanish credit card – but the exchange rate and hidden fees would be bonkers. Now, I can have a multi currency account, with low and transparent fees. Remittances are a lot more practical.

    In China, things have gone crazy. You can use your phone to pay for everything, they have micro payments with phones, etc. Sure, it has a dark side too, as the CCP can control more things now, but people wouldn’t be using these technologies if it didn’t make things easier.

    In Africa, a lot of people just pay each other by phone. They don’t even have to use cash that much. This facilitates access to banking without having to build all the infrastructure, which is very hard in Africa, which lacks in infrastructures.

    These are just the things I could think of that are practical right now. There are quite a few technologies that are promising (e.g. CRISPR, cheaper desalination), but are not there yet. Then there are the technologies that are being launched up now, about to become practical.

    There’s Starlink, which will give worldwide internet access in the next ~5 years.

    • Scott Alexander says:

      Did you read the paper? Something like that was originally my argument too, but C&S try to avoid having to have this argument by using objective measures.

      • ana53294 says:

        I’m not an economist, so I didn’t understand everything, and I’m not sure my critiques are right, but things that bugged me:

        First, they have an entire section that acknowledged social innovation, pointing at the examples of China, Japan, etc. They even call it social scientific knowledge. Then they seem to swipe it aside as if didn’t matter (?).

        TFP seems to really have slowed in the US. The majority of the inventions I mentioned did not seem to affect the US that much, so that seems logical. I guess this was the decade of other countries.

        In the Patents section. They count the number of patents per billion people. They acknowledge that if you don’t do that,

        Undenominated by population, scientific progress shows what is mostly a very positive story: a massive and rapid increase over time, accelerating with the industrial revolution.

        Since human population has already grown a lot, that means that a few countries can now contribute a lot to the population. I don’t believe that the population of scientists has grown that much in the last hundred years in Western countries. Many more people do PhDs than before, but more or less the same number of positions is available. So why does the addition of billions of people in third world countries matter?

        I agree that there are fewer low-hanging fruit to pick.

        As for crop yields… Sure, there are many more scientists working on soy and corn. But if you go to any plant breeding conference, and see the kind of work that is being done, you will see why. Nobody spends that much effort on improving yield, per se anymore. They instead focus on things like NUE (nutrient use efficiency), WUE (water UE), and a lot less useful things. So the reason crop yields aren’t improving that much is that few researchers are working on that.

        As for life expectancy, well, I don’t see why people’s choices to make themselves fat and sick, get less sunshine, do less exercise and eat worse food is an indictment on science. As others have noted, life expectancy without considering quality is quite meaningless anyway.

        • anonymousskimmer says:

          There are fundamental limitations on crop yields:

          If we fully unleashed photosynthesis this would lead to plant biomass doubling every few minutes during daylight (unfeasible in terms of nutrient and water uptake, or this would cause the plant to spontaneously combust or rip apart at some point if nutrients and water weren’t a limiting factor; also, the “pink goo” problem, but from a biological source).

          So, yeah, we are tinkering at the margins; anything more would be unusable, or some monstrous weed.

          Crops aren’t processors.

    • One problem with claims about the rate of technological progress is that it’s hard to know how to define the size of an improvement. Suppose that, sometime in the next few decades, we learn to solve the aging problem, so that people no longer get physically older, perhaps can even reverse aging.

      In one sense, that could be a very small improvement, the discovery of a single team that some combination of known drugs applied in a particular way reversed the aging mechanism—there is a recent experiment where that seemed to be happening, although it’s too early to be sure that what was reversed was aging rather than only the biomarkers being measured. But in another sense, it would be one of the largest technological breakthroughs in history, as measured by its effects on individual lives and human society.

      • Well... says:

        Right. And the effects of a breakthrough (the second sense you mentioned) might only be apparent years after the breakthrough is actually made (the first sense).

      • Paul Brinkley says:

        One of my long-time working theories is that the most important innovations are those that enable other innovations in their wake. The most straightforward such innovations are those which remove some bottleneck on the exchange or spread of information. My vision was of a sequence of such innovations changing the world: storytelling, writing, the alphabet, the school, the printing press, the telegraph, the radio, the television, the tape recorder, the computer, the database, the internet, the web, the search engine, the smartphone. They aren’t that glamorous in themselves, but rather in what we see as a result of their existence. I see them as especially crucial. (It’s why I spent a lot of my life trying to develop automatic semantic integration.)

        Since such innovations are relatively ordinary, I don’t expect them to draw much attention in research that makes claims about the rate of technological progress. Regardless, “leads to other innovations” has become one of my metrics for innovative importance.

        • albatross11 says:

          The way access to information has been opened up in the last decade is utterly stunning, and I expect it will drive innovation in ways nobody can currently forsee.

    • TJ2001 says:

      Remember that “This Decade” includes the tail end of a giant global recession…. Many many very good things died on the vine because “Recession”…. I think it’s probably obvious that there is going to be a big dip in the Progress-O-Meter…

  25. johan_larson says:

    What are the two most similar countries in the world? Canadians and Americans are enough alike that they are often mistaken for each other, but one country is a midsized regional power without a region and the other is a world-spanning superpower, so as countries the two are quite different. Can we do better?

    • EchoChaos says:

      Canada and New Zealand.

      Anglo mid-sized powers that do well in international conflicts when they actually fight but are able to act a lot more pacifist than any other country due to their geographical locations enabling them to practically depend on a much larger and better armed power (USA and Australia).

      • Well... says:

        Another similarity is the social dynamics of being white majority with an indigenous minority, sprinkled with a sizable immigrant population.

        Huge differences in typical cost of living though, right? And a moderate but large-enough difference in language. And of course very different climates and landscapes.

        • johan_larson says:

          I could see Canada and Australia being paired: they are both large countries with relatively small populations, predominantly English-speaking, with lots of immigrants and a small remnant aboriginal population, rich but not the richest, and firmly under Uncle Sam’s military umbrella. I guess the biggest difference is that Australia is tropical, while much of Canada is sub-Arctic. And of course Canada is snuggled up right next to big daddy, while Australia is shacked up in a cross-ocean love nest.

          Pairing Canada and New Zealand seems a bit stranger. While the two nations are quite similar, New Zealand is much smaller and I can’t believe NZ/Aus relations are anywhere near as asymmetric as US/Canada relations are.

          • EchoChaos says:

            New Zealand is much smaller and I can’t believe NZ/Aus relations are anywhere near as asymmetric as US/Canada relations are.

            It’s hard to get anything as asymmetrical as US/Canada, simply because “the hyperpower” has a different relationship to everyone else, but NZ’s foreign policy stance is substantially to the left of Australia’s because it is essentially totally immune from attack because of Australia.

            And the population ratio isn’t quite as lopsided, with Australia only being 5-1 over NZ compared to the US’s 10-1 over Canada, but it’s still very big. And they do have larger trade with China than with Australia, but that’s pretty much brand new. Before then, Australia was their top trade partner, just like the US and Canada.

        • John Schilling says:

          Comparing Canada, Australia, and New Zealand:

          All three are mostly-white prosperous Anglospheric states with formerly-highly-oppressed native minorities and assorted immigrant minorities.

          Canada and Australia are very similar in having ~30 million people producing ~$1.5 trillion in GDP from a highly diversified industrial economy, with New Zealand being the clear outlier in scale.

          Canada and Australia also have similar amounts of habitable land, as a narrow strip attached to a vast tract of nigh-uninhabitable land with a scattering of aborigines and miners. New Zealand again the outlier.

          Canada and New Zealand are very similar in having all of foreign policy being essentially optional because they have no neighbors capable of seriously bothering them(*). Australia has to deal with e.g. China and Indonesia in ways that can be very bothersome and which Uncle Sam can’t shield them from.

          Australia and New Zealand are very similar in having an entire land mass to themselves, with oceanic borders and no immediate neighbors. Also in being inconveniently far away from the rest of the Anglospheric/Western world, with associated cultural and economic impacts. Canada is the outlier there.

          Which two of the three are most similar, depends on which metrics matter most to whoever is making the comparison.

          * Except, in Canada’s case, the United States – which is mostly uninterested in seriously bothering them.

          • EchoChaos says:

            Excellently said. I agree with this, and noted the unusual comparison of Canada and New Zealand because I found the others less interesting.

          • cassander says:

            I prefer to phrase it “New Zealand is Australia’s Canada”.

    • John Schilling says:

      Andorra, Liechtenstein, San Marino, and Grand Fenwick all seem pretty much interchangeable. Microstates with traditional Western European culture, which manage a dubious existence because tourists like to visit and occasionally set up bank accounts in countries most people aren’t sure really exist. I’d add Monaco to the list, but the casino and the famous princess give it a more distinctive image.

      At the other end of the scale, I’d say the largest countries that could plausibly be described as interchangeable (as far as the outside world is concerned) would be Norway/Sweden, or maybe Australia/New Zealand.

    • Lambert says:

      Czechoslovakia wouldn’t have lasted so long nor split so peacefully had they not been quite similar.

      The problem with this question is that it’ll be quite easy for your answer to piss off a lot of Danes/Swedes/Kiwis/Aussies.

    • Some of the gulf petrostates?

    • The Nybbler says:

      Serbia and Croatia, except to the Serbs and Croats.

    • Eric Rall says:

      I considered nominating Austria and Switzerland, as both are medium-sized, mainly German-speaking, landlocked, mountainous, fairly rich, located in center of Europe, and conspicuously geopolitically neutral.

      Counting against them as twinsies, though: very different histories through the mid-20th century, Switzerland is much more linguistically and culturally diverse (63% native German speakers in Switzerland vs 89% in Austria), about twice as densely-populated (Switzerland and Austria have about the same population, but Austria is about twice as big geographically), significantly richer (per capita GDP is about $80k vs $50k), and has substantially more decentralized political institutions.

      • Kuiperdolin says:

        And Austria is fully integrated in the UE while Switzerland is in the UE’s influence sphere at most, and not even that deeply.

    • If we’re also counting regions, northern Italy, southern Germany and Austria (the previous Tyrol county) are strikingly like each other – to the point where southern Germany has more in common with northern Italy than southern Germany has in common with northern Germany.

      (Though I suppose that is cheating, since they were one thing previously.)

    • Anthony says:

      Uruguay and Costa Rica, maybe?

    • Chalid says:

      I just want to point out that Word2Vec is perfect for answering questions like “what are the two most similar countries in the world” and “New Zealand is to Australia as ____ is to the United States.” I don’t have time to actually go do the analysis though.

    • Kuiperdolin says:

      Spain and Portugal.

      Portugal is somewhat less powerful and has slightly better food and slightly worse climate, but otherwise same population, same language if you’re serious about it, same economy, same deep-rooted popery, same diaspora sending remittances home, two lost, world-spanning colonial empires, two right-wing but isolationnist dictatorships in the 20th century that ended mostly peacefully at around the same time, sat both world wars out, entered the EU at the same time, etc.

  26. Nancy Lebovitz says:

    Details about how “AI” has been hyped.

    I’ll note that “virtual reality” has been degraded from giving full bandwidth experience (I think that’s what it meant in science fiction” to just sight and sound.

    • Rachael says:

      A lot of that article is valid, but this sentence leapt out at me:

      “In fact the apparent coherence of the interview [with GPT-2] stemmed from (a) the enormous corpus of human writing that the system drew from and (b) the filtering for coherence that was done by the human journalist.”

      (b), fair enough. But (a) isn’t an objection – it’s the essence of what natural-language AI is achieving and what any intelligence is doing. The “apparent coherence” of Gary Marcus’s hype-busting article stems from the enormous corpus of human writing and speech he’s drawn from over the course of his life.

      • anonymousskimmer says:

        I thought the key word in (a) was “enormous”.

        If you keep feeding the AI more and more human writing, does its skill ever plateau (in the absence of human feedback), or keep increasing? And is the AI able to recognize bad (or regional) source material?

    • Nornagest says:

      I can’t recall how it works offhand in any other literary SF, but there’s a bit in Snow Crash that talks about how people in the Metaverse bow to each other, because shaking hands would remind them that they’re not actually there — the avatars can shake hands just fine, but it doesn’t give tactical feedback to the operators.

      • bullseye says:

        In Snow Crash you can handle objects in VR; there’s swordfighting. If I remember correctly, being able to move through other people is a deliberate feature to keep people from getting in each other’s way.

        • Nornagest says:

          There’s swordfighting, but it’s a minor plot point that the VR swords don’t handle anything like real ones, because of the lack of tactile feedback — they go through anything, rather than getting hung up in somebody’s neck. It’s not clear if the sword feels like anything to its wielder, but RL motion control games make it clear that the controls for a sword needn’t closely resemble one.

        • Trofim_Lysenko says:

          lack of collision detection is both to keep people from getting in each other’s way, and because it’s extra overhead to no real point. If I remember correctly, one of the features of the club Hiro’s ex-business partner runs is that it has a much more full-featured physics model including collision detection and so on.

  27. greghb says:

    I know of no better outlet for Unsong-esque linguistic coincidences, so just gotta post here I guess.

    The word “Tennessee” sounds the same in Hebrew as the verb “to try” when conjugated to the second person singular feminine future: “you will try” (תנסי). From this we derive Tennessee’s nickname The Volunteer State.

    A corollary, based on the gender of the conjugation: the Lady Vols basketball team is generally better than their male counterparts.

    (Counterpoint: the state Tennessee is not transliterated this way in modern Hebrew. The first letter in the state’s name is a tet, not a tav. But, you know, TINACBNIEAC.)

    • moshez says:

      Re: Tav vs. Tet:

      Modern transliteration uses tav for “th”, mostly because of the Yiddish influence on early transliteration effort. It is pretty a-historical, and transliteration to “tet” makes just as much sense.

  28. Well... says:

    Can anyone recommend a good winter work glove? After a couple hours on Amazon I’m mostly just experiencing decision paralysis, so I turn to you SSC commentariat. My budget is about $40, though I could go a tad higher for a really nice pair. Here are my requirements:

    1. Tough and rugged, so I can use these gloves to work outdoors (chopping wood, hauling branches and tools, etc.) without worrying they will tear or puncture. I’m also looking for a pair that will last a long time, not just one or two seasons. The primary material should be some kind of leather.

    2. Warm AF, which I assume means lined with something really dense. I understand my fingers getting cold after a couple hours in sub-zero temperatures, but until then I want my hands to stay toasty. I don’t want heated gloves, though, unless the gloves are still warm AF without the heating feature in use.

    3. Waterproof or at least water resistant. I need to be able to work in slushy snow without my hands getting wet, and without my gloves absorbing a bunch of water.

    4. Reasonable amount of dexterity. I’d like to be able to reach into my pocket and take out my keys without ungloving. This means the gloves can’t be really bulky.

    5. Available in size men’s medium.

    I do NOT need to work a touchscreen with gloves on, and in fact I’d prefer gloves that don’t include the “touchscreen” feature because I assume it means extra seams and compromises to the other aspects mentioned above. (Someone can correct me if this is wrong.)

    • Lambert says:

      I don’t think you can get all of that out of a single pair.
      Insulation is bulky, toughness is inflexible and heavy wear compromises waterproofing.

      Do gloves layer well?
      e.g. liner gloves, thin fleece gloves, ski gloves, big ol’ work gloves?

      • Sandpaper26 says:

        In my experience, no, gloves do not layer well. Layered gloves drastically reduce dexterity with little warmth gain. At best, a thin pair of gloves layers well with a warm pair of mittens. Unless Well requires the dexterity of all five fingers, some thin synthetic material gloves under warm, durable mittens with separate pointer finger is the way to go. Otherwise, I have a pair of raw leather work gloves with fleece liner that I got from a home improvement store that I consider a good compromise, though waterproof works both ways (and working in warm gloves makes your hands sweat).

      • Well... says:

        I’ve been wearing my lined leather gloves under my ski gloves. They layer well in the sense of being really warm, and I think the ski gloves are waterproof, but it’s super bulky, and I wouldn’t trust its ruggedness for doing any kind of rough work outside. The reason I combined them in the first place is because I tore a finger on the leather gloves just handling some wood in my yard.

      • Anthony says:

        The conflicts between the different requirements may be able to be overcome, but not for $40.

        An underlayer of tough nitrile gloves can work for waterproofing, but the insulation getting wet would probably eliminate any useful warming.

    • The Nybbler says:

      I usually use motorcycle gloves for winter dexterity; you can reach into a coat pocket to remove keys, though not pants pockets. However, a pair that is both warm and waterproof will be well out of that budget; an inexpensive waterlogged motorcycle glove is worthless for warmth.

      • Well... says:

        Can you recommend a starting point to search for such things?

        • The Nybbler says:

          Unfortunately I don’t have any that meet your criteria; that’s how I’ve found that waterlogged gloves don’t work. My current pair is the Redline brand, a few years old, inexpensive, tough, and warm when dry.

        • Michel Delving says:

          I would recommend not bothering. By going with a motorcycle glove, you’re going to be paying for a lot of features like impact and abrasion resistance that you don’t need. The ergonomics of the gloves are also going to be off; most motorcycle gloves are formed in a way that supports gripping handlebars and makes everything else harder.

    • Statismagician says:

      You want high-quality ski gloves. Any of them will fit 2, 4, and 5; 1 and 3 should be doable with a little searching. I can’t promise waterproof and leather right off the shelf, but this is trivially solvable with a trip to the hardware store.

    • joshuarterry says:

      You might want to look at ice climbing Belay gloves. They are made for lots of manual work in really cold conditions and are hard wearing.

    • A Definite Beta Guy says:

      Can you just get a standard glove and spray it with water resistant spray once per season?

    • Beck says:

      I worked outside in cold weather for years and was never completely happy with a set of gloves.

      Mechanix makes some decent cold weather gloves, but I’ve generally just bought a pack of fairly cheap ones with a nylon grip if I needed any kind of dexterity.
      This was for framing, so you may be looking for different things.

      • Matt says:

        I worked outside in cold weather for years and was never completely happy with a set of gloves.

        Me too.

        If you’re working outside, and working hard, even waterproof gloves won’t keep moisture out, because you’ll be sweating. (This is a balancing act that I could not conquer – how to stay warm on cold days without inducing sweat) On cold days, we always had multiple sets of (cheap) gloves and a fire barrel burning, both to warm our hands and to dry out the gloves we were switching from. Note that this did not mean that our hands weren’t cold, just helped to make the cold more bearable.

    • anonymousskimmer says:

      You should probably consider going to a place like REI and trying on what they have. Specialist stores of that kind are likely incentivized to hit key points in their in-store sales stock, and would likely have a good variety. If you do this though, they keep gloves in various parts of the store, so you’d probably need to ask for help to avoid missing any.

      Any specific recommendations anyone could give you 1) wouldn’t necessarily meet your relative criteria, or 2) may actually be your best bet, but would be unideal enough that you won’t realize they’re the best without trying on multiple other gloves.

    • martoca says:

      Elk skin gloves work well for me riding bicycles down to about 40 degrees for hours on end. They are thick enough to still keep fingers warm even when wet. They are soft enough to handle small items. USD62+ Available at
      foxcreekleather dot com elkskin-riding-gloves

  29. nupi says:

    Question about discussing experimental or more out there treatments with your doctor(s). How do you best go about it without them going to label you as a crank?

    Backstory: I was diagnosed with MS a few months ago and I am supposed to go on Ocrevus in two weeks which I am mostly ok with. I am wondering how to discuss more exotic treatments with my neurologist (thinking in particular of the Metformin ideas floating around which seems a safe and cheap enough compound to try [1]) as well some of the keto leaning diets. I am not a big fan of keto in general – in my view, regular people have no business being on any of them – but anecdotal reports (I know I know) in autoimmune disorders at least has me consider it…

    [1] For that reason I also don’t think it is a silver bullet, there must be thousands of people inadvertently doing those studies…

  30. Frederic Mari says:

    I cannot help but feel the Tyler Cowen paper is mostly mistaken. Or confusing one thing – a slowing growth/slowing productivity – with something else – slowing scientific progress – that allows Cowen to preserve his preferred political biases (conservatism leaning libertarian, AFAICT).

    But take the paragraph on life expectancies and health care. The authors acknowledge that the issue of declining life expectancies in the USA is basically an institutional problem rather than one of actual scientific progress.

    Indeed, AFAIK, life expectancy has kept going up in Western Europe. Focus on the second graph. This is entirely a “choice” of the USA to, in effect, declare that the lives of poor people ain’t worth the money. It has nothing to do with scientific achievements per se.

    Similarly, I’m a bit suspicious of productivity measurements. Everyone understands intuitively what we want to measure. But, in practice, the fact that they include labor costs in the calculation makes it feel self-referential to me. It loops with growth itself too.


    Imagine I increase my employees’ wages – without increasing their numbers! – but don’t vary my output – the calculation says my productivity went down. But did it? I don’t think that’s what most people and even economists like Cowen really imply when they use the term productivity.

    Another case – imagine there’s a recession going on, so less demand, so my output drops but, esp. in the early stage of such a recession, I refuse to let go immediately of parts of my workforce – after all, the situation might turn itself around and training and retaining good people is hard enough. My output dropped but my costs stayed the same. That’s a productivity drop apparently.

    At which point, productivity gains or slowdowns are highly correlated with GDP growth itself and you’re losing the explanatory power you were looking for in the first place when you sought to measure productivity gains.

    Or so it seems to me but I’m an amateur. Maybe Tyler Cowen will correct my misunderstandings.

    • kalimac says:

      I did not read the whole article, because it’s not nearly as entertainingly written as Scott’s, but as far as I got I had a similar impression.

    • Rachael says:


      Like Scott’s original article, I would measure technological progress in terms of the invention and/or widespread adoption of innovations that make people’s everyday lives different and better. It’s a bit of a subjective measure, but if the objective numbers don’t reflect it, maybe they’re counting the wrong things.

      • Frederic Mari says:

        @Rachel :

        I would measure technological progress in terms of the invention and/or widespread adoption of innovations that make people’s everyday lives different and better.

        Tangential b/c I’m more interested in productivity gains in this conversation but I think it’d make sense to consider “output divided by hours worked (regardless of compensation), assuming full/near full capacity utilization)”.

        The evolution of *that* number might give us an idea into the evolution of productivity and the adoption/deployment of technological progress within firms…

      • Freddie deBoer says:

        Isn’t this an example of shopping for metrics that fit your preconception?

        • anonymousskimmer says:

          Well, yeah, that’s the point though. If people assume the economists mean one thing, yet the economists actually mean something else, it’s worthwhile to have an additional metric that measures the thing people think is being measured.

          The adoption of smart phones and social media has objectively made some people’s lives worse (and not just the luddites who refrain from owning them). So it would be a good idea to include this decrease in value in the metric as well.

      • Lambert says:

        One year, when the opioid epidemic was worsening and there was a nasty flu, out of decades.

        • Statismagician says:

          Also, raw life expectancy is a deeply terrible metric for reasons which will be obvious if you’ve ever been to a nursing home, and more people are realizing this (cf the increased usage of living wills, palliative care, etc).

      • Frederic Mari says:

        I would also add that the fact the UK is dropping alongside the USA reinforces the point it’s an ideological choice.

        You want austerity and not spending money on poor people when they have an health issue or going through a risky situation (giving birth, for example. IIRC Black women in the US, esp. the South, die at rates way above norm)? That’s perfectly fine (well, not really…). But don’t act surprised when said poor people then die more often and spoil your national statistics.

        As long as you’re rich/very well insured, the US health care system kicks ass.

        • Jaskologist says:

          Getting a bit CW-baity there, aren’t you?

          If you’re going to hurl accusations, at least have something to back them up. Show me in the numbers where the US and the UK both spent less money on poor people’s health. Keep in mind that this is an actual decline in health, so there better be an actual reduction in spending to go along with it, and not just a reduction in the rate of increase.

          • Nancy Lebovitz says:

            The same money– or more– could be spent, but less effectively.

          • Frederic Mari says:

            sorry – Just to be crystal clear, I meant “you” as in the US/UK societies, not you you. I have no knowledge as to where you personally stand on these issues.

            I’ll address the data issue and baconbits objections when I’m home.

          • EchoChaos says:

            @Frederic Mari

            The problem that @Jaskologist points out is that this is a very CW topic, which Scott has asked us to refrain from in this thread.

            Post about anything you want, but please try to avoid hot-button political and social topics.

            Reposting this thread on Wednesday would be good.

          • Belisaurus Rex says:

            Even if the US spent as much on illegal/undocumented immigrants as it does on full citizens, perhaps they’ve already done things in their 20 years of life before crossing the border that would decrease their life expectancy. And we all know that illegal immigrants are counted in life expectancy but don’t get the same level of care. Look at the white-only life expectancy. Going up. This makes Britain’s decrease more of a mystery.

            Edit: I guess this would apply to any immigrant, but more strongly to undocumented/illegal immigrants. Is just mentioning that immigrants often have worse outcomes than natural born Americans too much CW? Or do I have to go into whether it’s a good/bad policy for that to happen?

          • a reader says:

            Belisaurus Rex:

            But Hispanic immigrants in the US actually live more than Non-Hispanic White Americans:


            Although Hispanic Americans are twice more likely to be living under the poverty line and three times more likely to not have health insurance than Non-Hispanic White Americans, they have a longer life span than them by 3 years.

          • Frederic Mari says:

            Jaskologist and co : Point taken on the CWness of the comments/discussion. Let’s save that for Wednesday then.

        • baconbits9 says:

          You want austerity and not spending money on poor people when they have an health issue or going through a risky situation (giving birth, for example. IIRC Black women in the US, esp. the South, die at rates way above norm)? That’s perfectly fine (well, not really…). But don’t act surprised when said poor people then die more often and spoil your national statistics.

          Your impressions of US healthcare are misinformed, differences in life expectancy in the US aren’t down to healthcare spending at all. The major group component that drives the US down is mortality for black males, and specifically black urban males who have lower life expediencies than low income rural blacks despite having significantly higher incomes (estimates often in the 40%+ range).

          The two major drivers of life expectancy differences between the US and European countries are the homicide rate and deaths from automobile accidents.

          • The two major drivers of life expectancy differences between the US and European countries are the homicide rate and deaths from automobile accidents.

            The U.S. homicide rate is about 5/100,000, down from about twice that a few decades ago. The U.S. highway death rate is about 12/100,000. Are those numbers large enough so that changes in them could explain differences between the changes in life expectancy in the U.S. and elsewhere?

          • Eric Rall says:

            The overall difference in death rates in the US and Canada, for instance, is about 100 deaths per 100,000 population (864 vs 763), so no. Which is surprising to me, since I seem to remember running some numbers a while back (I think around 2009 or 2010) and finding that car accident rates explained most of the difference in life expectancy.

            Either I made a mistake in my math or misread some numbers then, or the numbers have changed a lot in the intervening time, or my conclusion relied on some factor that gave car accidents a vastly outsized impact on life expectancy relative to overall death rate (maybe the death rate gap is partly due to population age distribution, not just a difference in life expectancy at birth; or maybe car accident deaths skew a lot younger than other common causes of death).

          • @Eric:

            Your figures are relevant to explaining differences in death rates, but not differences in the rate of change of death rates (or life expectancy). I’m not sure which we should be talking about, but Frederic’s claim was about “dropping” not about relative levels.

          • thisheavenlyconjugation says:

            The two major drivers of life expectancy differences between the US and European countries are the homicide rate and deaths from automobile accidents.

            Citation needed (or more plainly, no they aren’t). As Eric Rall says, homicides too rare to have a significant effect on life expectancy.

          • SamChevre says:

            Actuarial note from an actuary: homicides and car accidents have a disproportionate effect on life expectancy, relative to death rates, because they kill relatively young people. If the typical person dies at age 90, you have to keep 50 90-year-olds alive to 91 to have the same impact on life expectancy as keeping one 40-year-old from dying in a car wreck.

          • Nancy Lebovitz says:

            My impression is that grief shortens life expectancy. If this is true, then homicides would cause more damage than is seen form the homicide statistics, though possibly still not enough to lower general life expectancy.

    • albatross11 says:

      My very sketchy understanding is that the added mortality corresponds to higher rates of suicide and opioid overdoses (aka too much fentanyl in your heroin), mostly among middle-aged working-class whites. This doesn’t look much like either a failure of technological progress or society deciding that poor peoples’ lives aren’t worth saving.

      • EchoChaos says:

        This doesn’t look much like either a failure of technological progress or society deciding that poor peoples’ lives aren’t worth saving.

        Agreed. “Poor people’s lives aren’t worth saving” is pretty extraordinarily CW in a non-CW thread too.

      • John Schilling says:

        This doesn’t look much like either a failure of technological progress or society deciding that poor peoples’ lives aren’t worth saving.

        What would it look like if society decided that poor peoples’ lives weren’t worth saving? At the margin, where “society” isn’t actually marching the poor into ovens but are just dialing down the threshold of how much it will pay to keep poor people alive, and presumably favoring outcomes that don’t make “society” look cartoonishly evil.

        • baconbits9 says:

          There are probably many ways it could look, but it wouldn’t look like the US where people’s life expediencies are largely grouped by geography and race, not by income.

          • There are two different questions here. One is why life expectancy in the U.S. is lower than in some other countries. A quite different one is why the change in life expectancy is different. To explain the latter, what you need is not a difference between the relevant countries but a difference in how they are changing.

    • Cliff says:

      This is entirely a “choice” of the USA to, in effect, declare that the lives of poor people ain’t worth the money.

      Two things. First of all, this is pretty tendentious in a country where poor people get free health care and rich people do not.

      Second of all, we’re looking for the cause of a CHANGE, right? So what is the change where the U.S. suddenly decided poor peoples’ lives were not worth the money? Was it the Medicaid expansion that expanded free health care up to 138% of the poverty line? That was several years ago. Was it the healthcare subsidies that go to low, working class and lower middle income people?

    • Wency says:

      I’m disappointed Scott didn’t open the discussion back up and explain why this article convinced him. I’ve been in the techno-pessimist camp all along, but I’m always curious to hear arguments to the contrary. I’m mostly going on my own intuition, what I’m seeing in my life and what I hear from other people. The claims of advances happening behind the scenes, in AI and so on, don’t seem to be generating much change in the real world.

      The 2010s have been the slowest decade for technological change in my life (compared to 80s, 90s, and 00s). Cowen wrote The Great Stagnation 9 years ago, I agreed with it then, and it seems to have aged well.

      For practical consumer conveniences, all I can see in the last 10 years is that our iPhones, PlayStations, and Teslas have gotten incrementally better. As for productivity in workplaces, I haven’t seen much change in my industry or anything adjacent to it. PCs aren’t able to do much more than they could then — they just boot up quicker thanks to SSDs — and we’re mostly using the same software, just with a few more features, none of them life-changing. The best feature addition is just a minor convenience: digital signatures (perfectly doable 10 years ago but not widely adopted).

      Compared to the big changes from computerization in the preceding decades, this all seems like small potatoes.

      • For practical consumer conveniences, all I can see in the last 10 years is that our iPhones, PlayStations, and Teslas have gotten incrementally better.

        We spent Thanksgiving at the house of my son and his wife, which gave me a chance to try out their VR headset (a Quest). I was sufficiently impressed so that when we got home I ordered one. Since you are including entertainment, I think that looks like a more than incremental improvement

        • Wency says:

          Interesting thought. My sense is that VR remains basically a novelty (as it’s been for my entire lifetime) — not good enough to convert non-gamers into gamers, or to cause many gamers to cut back much on playing more traditional games. But it’s fair to say the technology has improved a lot, especially when it comes to affordability, and this was mostly a development that happened in the 2010s.

          Of course, I could also point to the rise of pixelated indie games in the 2010s as a sign of technological plateau. AAA games have looked the same for so long, compared to the breakneck pace of graphical improvement in prior decades, that gamers began to revisit the styles of the 80s and 90s in search of something different.

      • albatross11 says:

        I think one area of progress that has slowed dramatically is processor speed. Moore’s law is still in effect, but now we get more transistors on the chip over time, but not faster clock speed. I think there was about 30 years of increasing clock speeds and increasing speed for a single process to do some job. Over the last 10-15 years, that’s turned on its side, and we get more cores, more special-purpose modules (AES instructions, for example), etc., but an individual process runs at about the same speed it did 10-15 years ago.

        There is still amazing progress in computer hardware, but there’s a specific way in which older computers got better every year and new computers no longer do. If the problem you’re trying to solve can be handled efficiently by lots of parallel processes (multiple cores, multiple processors in the GPU, etc.), then you get a speedup, but if it’s a single-processor job, you don’t.

        The other thing that seems to me to have happened with computer hardware is that we’re *waaaaay* down the tradeoff curve in terms of cost to build the next fab vs performance/density of the processors produced.

        • Nornagest says:

          The fact that the new processor’s running at the same clock speed as the old one doesn’t necessarily mean that a given segment of code in a high-level language will take the same number of ticks to execute, even in a single-threaded environment. Those special-purpose instructions aren’t trivial in a lot of applications, and compilers have gotten a lot better too (largely thanks to parallelization).

          • albatross11 says:

            Sure. If you’re doing AES encryptions, modern processors are *way* faster. (Also more secure, since the obvious way to implement AES in software has a timing side-channel that leaks the key.) And there are improvements in instruction-level parallelism and such that make things run a bit faster in general. But it’s still the case that if you’re doing some inherently-sequential thing, it’s usually not much faster on your new computer than it was on your five-year old one.

        • Wency says:

          Agreed. Combine this deceleration in hardware improvement with the observation that computing is still where nearly all of our observable technological progress is happening, and it’s easy to get to the conclusion that technological progress has slowed.

      • anonymousskimmer says:

        DNA synthesis costs have probably dropped a factor of 5 in the last 10 years; turnaround times are much better as well.

        Sequencing costs have dramatically decreased (though some of this is due to less rent-seeking), and sequencing throughput has dramatically increased by a couple orders of magnitude. The machines are noticeably smaller too (Pacbio Sequel II is the size of a Sequel, though with almost 8 times the sequencing throughput, and a Sequel is about 1/3 the size of an RS II, with almost 7 times the sequencing throughput, and all this with higher quality scores).

        Automation and liquid handling hasn’t changed much in the past decade, at least at the scale I’m used to using in the lab. But it has already scaled to the point where it’s difficult to see if further miniaturization would be usable (it’s probably at the point where Theranos’ idea of their miniaturized lab is becoming feasible). Automation integration is being developed, including integrated tools for small labs (though so far these integrated tools seem to require vendor-specific consumables, except for the large-scale built-to-order integration).

        So I do see advances in my field still, and pretty significant ones. And this only includes the engineering-tech advances, not the biological-tech advances (e.g enzymatic discoveries and development).

    • This is entirely a “choice” of the USA to, in effect, declare that the lives of poor people ain’t worth the money.

      Do you have any evidence that insufficient spending on healthcare for the poor exists or is the cause of America’s low life expectancy? Keep in mind that healthcare expenditures have no corrleation with life expectancy above a certain level:

    • Paul Zrimsek says:

      You may be mixing up productivity with unit labor cost. When calculating productivity, the denominator is hours worked, not labor costs.

      More on cyclical effects on productivity here (PDF). They don’t look like that big a deal over the long run.

    • Imagine I increase my employees’ wages – without increasing their numbers! – but don’t vary my output – the calculation says my productivity went down.

      No, your profits went down and labor costs went up, productivity is unchanged.

      Another case – imagine there’s a recession going on, so less demand, so my output drops but, esp. in the early stage of such a recession, I refuse to let go immediately of parts of my workforce – after all, the situation might turn itself around and training and retaining good people is hard enough. My output dropped but my costs stayed the same. That’s a productivity drop apparently.

      And it should be, you’re producing less with the same amount. If it’s a smart business decision, you’d have higher productivity later, which would be reflected in the long-term productivity stats.

    • Douglas Knight says:

      Life expectancy has very little to do with medicine, let alone money. The best interventions in recent years are fighting smoking and drunk driving. Europe did these later than America, which is why American life expectancy leveled off earlier.

  31. zardoz says:

    [Post 1 reviewing “Business Adventures”]

    I read “Business Adventures: Twelve Classic Tales from the World of Wall Street” by John Brooks recently.

    When I first started reading this, I didn’t realize quite how old it was. The audiobook I was listening to had a copyright of 2014, but the text itself was written in the mid to late 1960s. That makes it an interesting time capsule of mid 20th century America. I also didn’t quite realize how long it was. The book is organized into twelve chapters. Each tells a separate tale.

    The first chapter talks about “the fluctuation,” the unexpected stock market plunge (and subsequent partial recovery) of 1962. Over the course of just a few days, the entire US stock market declined sharply, and then suddenly rebounded.

    It is an interesting window into the old, human-powered stock market of the 1960s. Brokers took customer requests, usually by telephone. They then communicated with traders in “the pit”– basically a bunch of people who stood around in the stock exchange, shouting orders to buy and sell. Once the trades were executed, they were written on slips of paper and sent around the building by pneumatic tube. A staff of “girls” read the paper slips and typed them into the computer, so that the current stock prices could be displayed on “the tape.” But the tape couldn’t keep up when the volume of trades is high. So over the course of the day, it sometimes fell beind by several minutes.

    During “the fluctuation”, the tape ran unusually late– several hours late, in fact. Brooks talks a lot about the possible connection between the lateness of the tape and the stock market plunge. I guess the theory is that by creating uncertainty about what the true prices were, the lateness of the tape exacerbated the fluctuation.

    Ironically, in modern times, people usually worry about the opposite problem: that by allowing prices to move too quickly, the present-day computerized stock market creates extra volatility. People talked about this theory a lot during the flash crash of 2010. Of course, both theories could be true. Perhaps the old stock market was too slow, and the new one is too fast. At first glance, though, neither theory seems as well-supported by evidence as some people seem to think.

    People today sometimes worry about trades being executed a few milliseconds too late. In the 1960s, trades might not be executed for hours if the guy in “the pit” couldn’t find the other guy with whom he was supposed to trade. And apparently during the great flucutation, some brokers just went out for a drink, since they couldn’t communicate with their partners in the pit. But don’t blame them too much– as Brooks notes, they actually saved their clients a lot of money by holding fast during the selling panic.

    Brooks likes to talk about some of the individual people who experienced the crash firsthand, like the guy who stood around with his paper notebook and recorded trades in AT&T stock. It might be harder to write this kind of human interest story about the stock exchange today. The dapper grey haired gentleman with the notebook is interesting in the way that CPU #7 executing stock trades is not.

    In the final analysis, Brooks says, the cause of the downturn is still “unfathomable… It occurred, and something like it could occur again.”

    [to be continued]

    • dpm96c says:

      Thanks for the review—I’ve never heard of this, but it sounds kind of charming for the reasons you point out.

  32. jhertzlinger says:

    Speaking of mistakes… There’s some evidence that the dark-energy theory may have been a mistake.

    That, in turn, means human settlement can catch up with the universe.

  33. jhertzlinger says:

    How long before tagging something as a mistake turns out to be a mistake?

  34. bagel says:

    You can tell a lot about the threats facing an airplane by its camouflage.

    World War I started with a lot of high-visibility uniform patterns, with an almost knightly air, and mostly ended with drab colors similar to the trenches but with highly visible roundels. This was the result of the generally low altitude fighting – which allowed the planes to be camouflaged at long range against the ground, but still identify friend from foe in the melee.

    World War II saw tons of camos, from sandy to woody to watery to black for night fighting. One crazy example is the pink camos that British Spitfires used for dawn reconnaissance. Early US stuff – and some early German stuff – was painted a high-visibility yellow because friendly fire was proving more dangerous than the enemy. Testing before Normandy (and even earlier for the British Typhoon and Tempest fighters) showed a lot of friendly misidentification of planes from the ground; hence the famous black-and-white “invasion stripes”. Tanks ended up doing the reverse and painting bright five-pointed stars on their roof so the allies wouldn’t bomb them. But by the end of the war, in Europe the Allies were flying higher and higher, and winning so handily, that they didn’t need camouflage anymore. They stripped most of the paint off their planes to save weight, resulting in lots of pictures of bare-metal B-17s, B-24s, P-51s, and Spitfires. The Soviets started even a little earlier, because they couldn’t afford the paint. In the Pacific, Japan continued to pose an air threat and the US Navy never stripped the paint from their planes. They – and all the other Allies in the Pacific theatre – did omit red from their roundels to avoid confusion with the Japanese, however.

    These practices continued straight through Korea, buoyed by the fact that widespread airborne radar meant that everyone knew where to find everyone else. The US Air Force and Soviet-aligned forces flying mostly un-camouflaged planes and the US Navy painting them … well, navy blue. Strategic Air Command? All bare metal. It’s the aesthetic of the jet age, of the 1950s in many ways. Even the U-2 flew bare metal, early on.

    In fact, this continued all the way up to the introduction of the SA-2 “Guideline” missile. Suddenly, high altitude wasn’t safe; so the burly, high-speed low-drag fighter jets returned to dodging trees and anti-aircraft guns in order to stay hidden at ground level. Cover – from hills, the curve of the Earth, and the noise of the ground clutter – was the only protection against those prying electric eyes. Those camouflage patterns, unused since World War II, were rediscovered over Vietnam, with F-100s and F-105s the last frontline jets to roll out in bare metal. The only new addition was a high altitude gray, because it turns out everything looks washed out way up there in the thinner atmosphere.

    After Vietnam, we re-evaluated how we would break open an anti-aircraft network and decided to focus more on electronic warfare; we thus split our planes between high altitude cover (gray) and low altitude sneaking (traditional camo). We also added a brand new technology; stealth. A combination of materially absorbing and geometrically deflecting radar waves made planes invisible to radar, allowing the F-117, the very first stealth plane, to stroll right up to their targets without being detected. It was painted black to help it remain invisible at night, but through most of development the Skunk Works had it painted gray to serve as a high-altitude missileer, hence the “fighter” designation.

    Since then, the West has mostly owned the skies and painted planes drab grays as a result. The F-22 and F-35 are both relatively dark gray for high altitude work, which is believed to be a result of its stealth coating.

    That’s a short history of aviation camouflage! In short, a plane wearing a traditional camouflage is operating low and expects to be spotted (or not!) by eye. A plane with bare metal is operating high and expects to be seen on radar long before eye. A plane in drab gray expects to fly high and be seen by eye before radar, either because the enemy radar is jammed or destroyed. A plane only wears black at night.

    Honorable mentions go out to razzle dazzle camouflages, which try and visually confuse other planes; false cockpits, more of the same; high-visibility schemes, such as the US Coast Guard or US target drone programs use, for when you want to be seen; radar retroreflectors, so that nobody runs into stealth planes in day-to-day operations; and Yehudi Lights, active visual camouflage in World War II that allowed the submarine hunters to sneak up on the subs during the day.

    • nkurz says:

      Wonderful post, thanks for writing it up! For others like me who hadn’t heard of them, here’s a Wikipedia article about the “Yehudi Lights” mentioned at the end:

    • AlphaGamma says:

      In fact, this continued all the way up to the introduction of the SA-2 “Guideline” missile. Suddenly, high altitude wasn’t safe; so the burly, high-speed low-drag fighter jets returned to dodging trees and anti-aircraft guns in order to stay hidden at ground level. Cover – from hills, the curve of the Earth, and the noise of the ground clutter – was the only protection against those prying electric eyes. Those camouflage patterns, unused since World War II, were rediscovered over Vietnam, with F-100s and F-105s the last frontline jets to roll out in bare metal. The only new addition was a high altitude gray, because it turns out everything looks washed out way up there in the thinner atmosphere.

      This also applies to the RAF’s V-bombers, which entered service painted anti-flash white to protect against heat from nuclear detonations. After the SA-2 forced a switch to a low altitude mission profile, they were repainted in the grey and green RAF Strike Command camouflage pattern. The last V-bombers in service, the Victors converted to airborne refuelling tankers, were painted in a low-visibilty desert scheme for service in the first Gulf War.

      • bagel says:

        This also applies to the RAF’s V-bombers, which entered service painted anti-flash white to protect against heat from nuclear detonations.

        Another excellent paint scheme, with a very different – and much more direct – sort of protection!

    • bean says:

      You see the same thing with ships. Until around 1900, ships were painted in schemes that made them pretty visible, notably the Nelson Checker and Victorian Black-and-Buff. This was primarily because their propulsion tended to be quite visible by itself (first sails, then coal smoke) and because battle ranges were close enough to make it obvious where the ships were. Around 1900, improvements in fire control drove up battle ranges and made the use of grey paint profitable to confuse the enemy even if they knew roughly where you were. As oil fuel came to prominence, camouflage became even more important. Ships no longer belched smoke like they did in the days of coal. Lots of schemes were tried, but most died off post-WWII as radar became increasingly important. Also, decks started to be camouflaged in wartime, because there was a chance of ships being spotted from the air. Drazzle camouflage was used on ships whose most important enemy was the submarine, because it made it harder to figure out a ship’s course and speed, particularly if you only had a brief glimpse through a periscope. This was also the purpose of things like false bow waves, which would make a ship look like it was going faster than it was.

      I probably should write more on this. Hmm….

      A plane in drab gray expects to fly high and be seen by eye before radar, either because the enemy radar is jammed or destroyed.

      Grey is used because it’s versatile. It may not be perfect for any given environment, but it works reasonably well everywhere. A modern fighter has to be able to handle both a high-altitude dogfight and a low-level strike mission, and grey isn’t horrible at either. Planes which only fight at low level may get other schemes, but there aren’t that many of those in these days of multi-role aircraft.

      It was painted black to help it remain invisible at night, but through most of development the Skunk Works had it painted gray to serve as a high-altitude missileer, hence the “fighter” designation.

      It has a “fighter” designation because the Air Force hates A-for-attack planes as not suitable for its manly image. The F-117 doesn’t even have space for a good air intercept radar, which you need for a missileer.

      • bagel says:

        You see the same thing with ships.

        Ship camos are baller. Bring back razzle dazzle pl0x.

        Grey is used because it’s versatile. It may not be perfect for any given environment, but it works reasonably well everywhere.

        As many glorious shots from the “Star Wars Canyon” indicate, gray is not very low-observability against greens. 😉

        It has a “fighter” designation because the Air Force hates A-for-attack planes as not suitable for its manly image. The F-117 doesn’t even have space for a good air intercept radar, which you need for a missileer.

        According to Ben Rich, in his book Skunk Works, the F-117 was originally spec’d as a high altitude fighter, and only at the last minute converted to the attacker we know it as. Given that the F-117 was his baby when he ran the Skunk Works, I’m willing to take him at his word. The same is true of the F-111; although it was supposed to be a missileer, it took a nearly clean-slate redesign to turn its core ideas into the much more successful F-14.

        What exactly they were planning to do about an air-intercept-radar is a great question. I could image that they didn’t want an internal radar which would give the game away to any plane with an radar warning receiver; maybe they planned to have a friendly radar buddy-lock? Or they could have planned to modify a heat-seaking missile with lock-on-after-launch capability. Or just loaded one of those nuclear rockets in because lol Cold War. They were downright paranoid about EMCON in the F-117, not even communicating by radio.

        • bean says:

          According to Ben Rich, in his book Skunk Works, the F-117 was originally spec’d as a high altitude fighter, and only at the last minute converted to the attacker we know it as.

          First, I’ve looked at Ben Rich’s book, and at best he’s someone doesn’t let the truth get in the way of a good tale. There were lots and lots of factual errors, and with a tone that seemed calculated to appeal to young engineers who wanted to curse the world for not seeing their genius. I particularly liked the section where he claimed to have cracked passive quieting and been ignored by the Navy. Yeah, right.

          Second, I cannot believe that the USAF would get offered stealth, and ask for a high-altitude fighter first. A strike aircraft for hitting really hard targets makes perfect sense. One stealth strike aircraft might replace a dozen or more conventional aircraft, several of whom would be lost along the way. A small handful could go a long way, while a stealth air superiority fighter isn’t going to have nearly the same leverage, and you couldn’t use them without radiating. Preplanning can replace communications and active sensors, but you can’t preplan air combat missions like that. It’s pretty obvious that the F-117 got its designation because they were using numbers above 112 for purloined Russian aircraft, and it was intended as a cover.

          The same is true of the F-111; although it was supposed to be a missileer, it took a nearly clean-slate redesign to turn its core ideas into the much more successful F-14.

          The problem with the F-111 was that it was a hybrid design, badly compromised by the need to also be a low-level strike aircraft.

          Or just loaded one of those nuclear rockets in because lol Cold War.

          Those were all long gone by the 80s.

    • Nancy Lebovitz says:

      Is there some history about friendly fire becoming less of a problem?

      • bean says:

        It’s probably due to improvements in technology for managing an air battle. In WWII, most guns were manually aimed, and only loosely connected to anyone who had knowledge of the overall air battle. Now, even a typical MANPADS has an IFF transponder, and most of the air defenses are netted and know exactly what they’re shooting at. Aegis and AWACS are really, really useful for that.

      • bagel says:

        Modern electronic systems profess to identify friend from foe – when they work. Their reliability was a big problem in Vietnam and still caused some near misses in Desert Storm. But a jumpy recruit looking at two very similar planes with their defensive guns was apparently even worse. Maybe not because their rate of mistakes is all that high, but they have a lot of opportunities to make them. They see every friendly plane twice, after all, and maybe don’t see enemy planes.

        This is also the motivation behind the infamous “PT belts” in use today; soldiers were dying more for reasons like being run over by humvee drivers who didn’t see the soldiers for their camouflage, so they wear bright, reflective yellow belts over their camo outside of combat. It’s as silly as it sounds, but it supposedly does work.

    • proyas says:

      I wonder if planes will have active camouflage in the future, allowing them to change color to match the environment.

      • bagel says:

        I bet we’d be more likely to see it on helicopters before airplanes. But there have been public tests of thermal camouflage to make tanks look like cars, so it’s within the realm of possibility!

  35. ajakaja says:

    For what it’s worth all of @QiaochuYuan’s ideas seem valid and probable to me based on my experiences with therapy and with providing therapy to friends.

  36. ketchupduck says:

    I recently did a calibration assessment and found out I am very underconfident, which I understand is pretty unusual. I’m planning on doing more practice to get closer to perfect calibration. If people in a similar situation have had luck with particular strategies to improve their calibration, I’d love to hear more.

  37. Auric Ulvin says:

    Does anyone else (occasionally) see GPT-style text flashing before their eyes when half-asleep?

    Sometimes when I come home, the bus lulls me into a slumber where I close my eyes. I then read grammatically correct but meaningless sentences, the text flits across my vision as I read it. I can sometimes alter the topic but I can’t think too hard about it or I’ll wake back up again and the text will disappear.

    If I focus and close my eyes for about 15-20 seconds at any given time, I can also see distorted letters flitting across my central field of view, though they don’t form words. I do a lot of reading in my everyday life, so it’s not surprising that it would cross over into my subconscious.

    • Solra Bizna says:

      I “read” such text without seeing it, in that the words seem to flow into my head in the same way they do when I’m speedreading, but without any visual component. The sentences are incredibly GPT-like, in that they are globally nonsense but make “sense” locally. It’s almost always as if I am reading an encyclopedia article, except that the topic frequently changes. Sometimes I can hold a sentence or two in my head long enough to analyze them; the “facts” held within are always plausible-sounding but rarely match reality. Usually, when I try [to re-read one], the sentences get rerolled, giving me similarly-structured ones with wildly different content. I’m always either half-asleep or otherwise compromised when it happens.

      This experience predates my exposure to GPT-2, and is part of the reason that GPT-2 excites me so much.

    • anonymousskimmer says:

      Generative Pretrained Transformer“, a text generating artificial intelligence system.

      Acronym defined for those unfamiliar with it, as a typical googling of “GPT” brings up “GUI Partition Table”.

    • Placid Platypus says:

      I get images, but not text.

    • Rebecca Friedman says:

      I get audio rather like what you describe – like listening to my own thoughts, except I’m not consciously thinking them and the sentences don’t make sense. It’s usually immediately before falling asleep, and a good sign that I’m actually going to get to sleep. If I focus too hard on how they don’t make sense or try to control them, the effect fades; if I just listen I drift off to sleep. I’ve suspected it’s a lead-in to dreaming, but don’t remember my dreams so I don’t actually know. I’ve never seen text or images; it’s fascinating that other people do!

      • Lambert says:

        I get audio too, but sometimes it’s from a specific person.
        Often someone I’ve spoken to that day. Maybe it’s my subconcious building something like a GAN.

    • thevoiceofthevoid says:

      No text; instead, my constant inner monologue / subvocalization gradually becomes less and less coherent, and (somehow) begins to turn into experiences/dreams rather than just words.

    • Shion Arita says:

      I don’t see letters, nor do I really ‘hear’ sound, but my mind ‘speaks’ long strings of grammatically correct nonsense. It seems based on the responses of others that this is fairly common.

      Interestingly enough, people who have had damage to the brain region called Wernicke’s area can develop something called Wernicke’s Aphasia, or ‘Fluent Aphasia’, in which their speech becomes VERY GPT-2 like. I.E. it is all grammatically correct, and locally makes sense, but isn’t about anything globally.

      Here is an example.

  38. JPNunez says:

    Programmer Codexers, Advent of Code 2019 is open

    enroll with your github/twitter/google accounts and solve 1 programming puzzle every day until the 25th to try and save Santa and Xmas (or…your favorite santa related holiday)

    first puzzle is basically adding a bunch of numbers, but I promise it gets trickier later on

  39. broblawsky says:

    I know that there are a lot of people here who are interested in investment, and a lot of people who work in machine learning-adjacent fields. Has anyone tried training an algorithm to assist in market or sector timing? I’m considering trying to implement one sometime in the next month or so, and I was wondering if anyone else here had experience with that already.

    • ana53294 says:

      While I haven’t done it, and stick to passive funds, there’s a guy who has been using math models for investing for decades, and it has gone quite well for him. Maybe you find his math model accidentally!

      I wouldn’t do it with your main savings, though. Just with your play money (I put 10% into that; your tolerance could be higher).

      • DarkTigger says:

        5%/44% fees? I Thanks but I think I pass.
        I mean, if this fund performs as advertised, 44% performance fees would be acceptebale. But the added 5% management fees?

        • ana53294 says:

          The fund is closed anyway. Yes, it makes profits, because they’ve been able to have a positive result for the last 30 years or so.

    • wonderer says:

      You should read “A Random Walk Down Wall Street” before trying something like this. To summarize the book, it isn’t possible to do market or sector timing. In fact, no method that’s easily discovered or implemented can possibly let you beat the market. If such a method existed, someone would be using it already, which would move the markets in such a way that the method becomes ineffective. This is called the efficient market hypothesis.

      Here’s another way to put it: you are trying to compete against billion-dollar financial companies, which hire thousands of the smartest people in the world to work together full-time in squeaking out the tiniest possible statistical advantage. They have the best data and tools that money can buy. They think nothing of paying millions of dollars to put their algorithmic trading FPGAs (because computers are too slow) slightly closer to the stock exchange to shave a few milliseconds off the light travel time. Do you really think you could beat them at their own game? This isn’t David vs. Goliath; it’s David vs. the entire Philistine army, if God was on the Philistine side and the Philistines were the ones with slings.

      • kalimac says:

        This is why I don’t gamble on sports games. What makes me think I can beat the spread, when the spread is calculated from the accumulated knowledge of everybody else who knows the game a lot better than I do?

      • albatross11 says:

        The weaker form of the EMH is something more like: Before you try invest on your own and beat the market, explain why you know more than all the pros investing in the market. It’s not impossible that you *do* know more than they do, but it’s probably a pretty rare event….

        • Chalid says:

          You don’t need to know more than the pros. You just need to know a significant thing that isn’t already priced in. (Cue David Friedman’s story about buying Apple stock, or Scott talking about his doctor dad’s successful investments.)

          But you can take it for granted that lots and lots of pros know about machine learning, so this is not an exploitable edge.

          • Thomas Jorgensen says:

            Or you could try to exploit something that the market damn well knows but is not acting on due to exceedingly common biases.

            Wall-street is very high testosterone.

            The market has so far systematically undervalued companies with female managment. This is historical data, and in the literature, so may be arbritaged away any day now, but simulated investment funds that operate on the principle of “We hate male ceos, and will therefore execute on a strategy of buy-and-hold female-lead companies, and divest when a firm gets male leadership” keep beating the market like a drum.

            This is the exact sort of thing econ text books says should result in a real fund executing this strategy and gobbling up the market until the bias goes away, but… “misandry, LOL” is not a massive fund that exists, so, presumably this has yet to happen.

          • CatCube says:

            @Thomas Jorgensen

            So why is this still just “simulated?” They can’t find one rich woman to bankroll this sure-fire money-printing opportunity?

          • wonderer says:

            What CatCube said. Why is this simulated? Why hasn’t anyone proven just how well this works by becoming rich? It doesn’t need to be a woman–plenty of men would jump at the opportunity to both get rich AND reduce systemic biases by helping out female CEOs.

          • Chalid says:

            Speaking from experience, attributes of management (including the amount of female management) is the sort of thing that would fit comfortably into a quant model with 50+ other factors. I’ve never tested that exact signal myself, but I’m sure lots of people have done so, and it wouldn’t shock me if it works and it’s being traded. (It also wouldn’t shock me if it doesn’t work and the papers on it are wrong or for some reason untradeable). With most of this stuff, the money-making opportunity from any one signal in isolation is not that great; you have to combine lots of signals before you can get anywhere.

            “I don’t know of anyone getting rich off this strategy” is not a good argument, since the sort of people who would get rich off this strategy are likely using it as part of a complicated portfolio of strategies and wouldn’t particularly advertise any one of those strategies.

          • wonderer says:

            With most of this stuff, the money-making opportunity from any one signal in isolation is not that great; you have to combine lots of signals before you can get anywhere.

            I’m willing to believe that the best quants in the world can squeeze a tiny statistical edge out of this. What I’m not willing to believe is:

            simulated investment funds that operate on the principle of “We hate male ceos, and will therefore execute on a strategy of buy-and-hold female-lead companies, and divest when a firm gets male leadership” keep beating the market like a drum.

            Even if this was true historically (which I highly doubt), it wouldn’t be true today. Squeezing a tiny edge out of a 50 factor model is not “beating the market like a drum” based on that one strategy.

          • Thomas Jorgensen says:

            Trading floors are incredibly male. They also think they, themselves, are the pinnacle of all creation. This is an enviorment which is going to create sexist biases from hell, and the market does not need to undervalue female-lead companies much to make the described strategy a very effective engine of picketing the markets pockets.

            Lets say that the market on average underprices a female lead company by five percent. Do you find this implausible as something that could happen due to the madness of crowds?

            That does not mean you end up realizing 5 percent better returns by only owning women-lead companies. It means you realize a five percentage point return on your entire investment in every firm you are invested in whenever said company switches to a male CEO and you go “EEW, y-chromosomes are icky” and cash out – because the market reevaluates the firm at that point, and pays you the pricetag without the malus arising from bias. Which leads to stuff like this.


            There are lots of other examples of this kind of work. Maybe there is a quiet cohort of people executing on this and keeping their mouth shut, because, after all, if enough money get in on it, it will stop working.

          • abystander says:

            Past performance is no guarantee of future results. Whenever you mine historical data you can find a strategy that generates large returns, but doesn’t work going forward like buying companies using theory XYZ style of management.

            You can try a number of social justice factors like unionization and if companies with unionized workers outperform the market then the reason is that workers are more productive with their rights protected and if the companies underperform it is not the unions fault.

            And it looks like since that 2015 article came out reversion to the mean has taken place which is how I would put it has taken place.

            There are couple investment vehicles concentrating on woman lead firms but they aren’t big enough to remove a supposed bias and they are now underperforming.
            Barclays Women in Leadership ETN (WIL) 6.75% 5yr trailing return
            Pax Ellevate Global Women’s Leadership Fund Individual Investor Class 7.76% 5yr trailing return

            compared to
            SPDR S&P 500 ETF Trust (SPY) 10.69% 5yr trailing return

          • Trading floors are incredibly male. They also think they, themselves, are the pinnacle of all creation. This is an enviorment which is going to create sexist biases from hell, and the market does not need to undervalue female-lead companies much to make the described strategy a very effective engine of picketing the markets pockets.

            The traders aren’t the ones determining market prices–the investors are. So I don’t see how the views of the floor traders are relevant.

          • Chalid says:

            There are lots of other examples of this kind of work.

            Yes, and unfortunately those examples mostly don’t work out of sample, or it don’t replicate, or actually have their performance explained by other factors, or are not tradeable for some reason. Otherwise my job would be way way easier.

            The traders aren’t the ones determining market prices–the investors are. So I don’t see how the views of the floor traders are relevant.

            This is true, but the investors are mostly male too, fwiw.

          • John Schilling says:

            and the market does not need to undervalue female-lead companies much to make the described strategy a very effective engine of picketing the markets pockets.

            And yet, it appears that no one has ever done this. More to the point, you have never done this. It’s like some strange variant on the old bit about the two economists and the $20 bill on the sidewalk, but the allegedly wise economist is saying “No, this time it really is a $20 bill that nobody has bothered to pick up because stupid reasons”, and still not picking up the $20 bill.

      • broblawsky says:

        I think my main sticking point there is that I don’t believe markets are efficient right now, thanks to the proliferation of passive investors.

        • Cliff says:

          And the much larger proliferation of quants with massive pools of capital hooked up with zero lag directly to the stock markets?

          • broblawsky says:

            Hedge funds only own ~2-5% of the market, though, and I don’t think there are many quants outside of hedge funds.

          • Cliff says:

            Aren’t there quants at every trading company? Goldman Sachs, Chase, etc? Don’t these firms trade on their own accounts for corporate profit?

            Have you surveyed the literature about market efficiency?

          • wonderer says:

            Hedge funds might own only 2-5% of the market, but actively managed mutual funds are still 50% of the mutual fund market. And the very fact that hedge funds are not a bigger percentage of the market should give you pause. If they all get very good returns, why haven’t they all grown much larger? Why isn’t every rich person throwing their money at hedge funds, and why isn’t every entrepreneur trying to start a new hedge fund?

            As Cliff mentions, there are also lots of other institutions that actively manage their assets. The endowment of Yale is actively managed, and managed quite well. The same is probably true of the other elite universities.

          • thisheavenlyconjugation says:

            Why is that percentage the relevant thing? The question to me seems to be whether hedge funds are limited by the capacity of their successful strategies or by access to capital. All evidence I can see points to the former: you see hedge funds restricting access to top-performing strategies to employees because they have limited capacity, not trying to borrow money to trade with.


            Don’t these firms trade on their own accounts for corporate profit?

            Officially and in the US, not since the introduction of the Volcker Rule in 2014. They still have quants though.

    • John Schilling says:

      As always, there’s an XKCD for that. And with machine learning rather than the merely human kind, you can get to the same end result faster and more efficiently.

      • EchoChaos says:

        As Microsoft learned fast when they exposed TayTweets to the Internet, AI often doesn’t do exactly what you want when interacting with the real world.

    • Douglas Knight says:

      Even if you had the right algorithm (and RenTech claims that it mainly uses one variable linear regression, but that it takes a PhD to conscientiously do what the textbook instructs), where would you get the data? RenTech claims to have used the cloud cover in Paris. Where would you get historical data for this? Where would you even get historical data for stock prices? How much do you trust your record historical stock prices? Even if errors are very simple, they could propagate information backwards in time, which is the key problem. Historical data often drops stocks which ceased to exist, so the landscape of stocks does not look the way it did at the time.

      If you think that passive investors have made markets less efficient, then you seem to think that historical data wouldn’t be useful. But then what would you apply machine learning to?


      Not really relevant to your post, but there’s the weird story of the hackers that stole 150,000 press releases ahead of time. They only traded on 1% of them, chosen partly on the basis of liquidity and partly on the basis of how well they could predict the direction of the movement of the stock. They only got it right 77% of the time. Of course that’s plenty to make money, but it’s surprisingly low, considering how much filtering they put on them.

    • Chalid says:

      Caveat, I’m not particularly knowledgeable about machine learning and have never tried what you’re doing.

      You’re going to have data quality or quantity issues. You can get daily data on sectors, in which case you’re going to have ~20 data points per day * a few years of relevant data, and I doubt that’s enough to get you anywhere. You could get more data by modelling individual stocks instead of sectors (a few thousand per day) but then I expect that you as an individual investor are going to have to do a lot of work to actually trade them, not to mention getting the issues of getting a clean dataset free of e.g. survivorship bias. You could try to get more data by going to higher frequency (e.g. hourly) but then trading becomes difficult again, and I think also the data itself may become expensive and/or suspect (no personal experience here).

      I’m not as much of a market fundamentalist as the rest of this thread – I know a couple people who have successfully have pulled this off and managed to turn modelling they did at home into jobs managing real money (though I think they all had relevant finance experience going in). And there are professionals who essentially trade on price and volume datasets.