Open Thread 151

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 – and also check out the SSC Podcast. Also:

1. Comment of the week: Bean questions The Precipice‘s assessment of nuclear winter, links back to a comment thread from years ago challenging the Robock paper Ord relies on.

2. New sidebar ad: this one is for SafetyWing, which bills itself as “Norwegian founders with an international team on a mission to offer the equivalent of a Norwegian social safety net globally available as a membership – currently offering travel medical insurance for nomads, and global health insurance for remote teams.” It also has a Medium article where it claims its end goal is to build a country on the Internet, which is a delightfully tech-startup-Medium-article thing to claim.

3. As always, I apologize for being bad at answering emails. In some cases, I am weeks behind. In other cases, I am grateful for what you have to say but have given up on responding. In others, I have said “that’s interesting, I’ll check it out” while secretly knowing that I will never do that. Expect this kind of thing to continue.

4. Dan Wahl has an automated Unsong audiobook.

5. Calling researchers and lawyers – does anyone know the legalities of running an informal study without IRB approval? IE, if I were to email the SSC mailing list recommending people try a certain vitamin, nobody would think twice of it. If I were to email the SSC mailing list asking people with last names A-M to try the vitamin, and people with last names N-Z not to, and to record their results and send them to me, how much trouble would I be in? If the answer is “not much”, what does the requirement that “studies” have an IRB mean? Where do you cross the line from neat decentralized science experiment to official study, and what happens to people if they end up on the wrong side of it?

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

1,112 Responses to Open Thread 151

  1. gbear605 says:

    Kabbalistically, an automated singing of an unsung book sounds dangerous.

  2. Loriot says:

    One thing I’ve wondered about – how common are the stereotypical common dreams, anyway? To the best of my knowledge, I’ve never dreamed that I was naked at school for example.

    As far as I can tell, dreams are just a random mishmash mostly of things you’ve been thinking about lately, just with the part of your brain that notices inconsistency turned off.

    • Scott Alexander says:

      I’ve had “teeth falling out” and regularly (a few times a year) have “I’m in school and there was a course I didn’t realize I was in and I’ve missed it the whole semester”

      • Conrad Honcho says:

        I’ve never had “teeth falling out” but “course I didn’t attend at all until I realized it was the end of the semester” is at least twice a month.

        • AliceToBob says:

          I’ve never had the teeth one either, although perhaps I will now.

          Recently, I had a dream where I realized I was dreaming (since I could put my hand through a wall with no resistance), “woke up” relieved, but then was still able to put my hand through a wall again, after which I was really able to wake up. I’ve never had a fake “waking up” moment before, and it was interesting.

      • Buttle says:

        I have had both of those — teeth turning to chalk and falling out, and realizing just at finals time that I had registered for a class — numerous times, but thankfully not in the last twenty years or so. I used to also have the hold your mouth right and you can fly dream, that one I sort of miss.

      • Nick says:

        “I’m in school and there was a course I didn’t realize I was in and I’ve missed it the whole semester”

        I still get this one, and I have no idea why.

        • Evan Þ says:

          I never got it till after I graduated, but I’ve gotten it a few times since then. Perhaps my subconscious was adjusting to the realization that I haven’t attended any classes for the last few years.

          • Conrad Honcho says:

            It never ends. I’ve been out of college for 20 years and I still get it.

          • Edward Scizorhands says:

            Same here. WTF.

            I was really good at attendance, which is maybe part of the problem.

          • When you get to be a professor, the dream is the one where you realize that you were supposed to be teaching a course and haven’t showed up for the last month and a half.

            Or else you have to find out what class you are supposed to be teaching and where it meets, and it’s starting in ten minutes.

        • silver_swift says:

          Same here, though in my case it’s usually that I need to finish one or two more courses before I can graduate, but I’ve been to busy with my job to go to any of the lectures (I didn’t have a job while I was still studying). The dream then consists of panicking that I won’t be able to finish those courses and therefor won’t graduate on time.

          I graduated 7 years ago, I would really like for my subconscious to catch up to that fact.

        • keaswaran says:

          My version of this one is far more often “I have a plane ticket to go somewhere but I’m in the wrong airport for departure”. A week or two ago, I even had a version of this one where it was my *luggage* that was in the wrong terminal of the airport, even though we were well into the “no more flying” situation, and my local airport doesn’t even have a second terminal for my luggage to have been at (and for some reason the second terminal was a 20 minute drive away in my dream, even though nothing in my town is a 20 minute drive away from anything).

    • Rebecca Friedman says:

      I’ve had a few of them. The “it’s finals and I just realized I’ve been signed up for this/these class(es) and I haven’t been showing up for them,” for example.

      Edit: Maybe once a year or less for me; it’s a stress dream.

      • mtl1882 says:

        I’ve had that one a few times since graduating, but never during school. It still freaks me the hell out. Although in my case, it was more like I knew I was signed up for the class and had not bothered to drop it, but couldn’t exactly remember and was avoiding the issue. This situation is pretty nonsensical, but it made perfect sense in the dream, so it was fear mixed with dread and guilt over my irresponsibility. It was so vivid that one time I actually kept thinking this had happened for several hours after I woke up.

        • albatross11 says:

          I once woke from that dream a week or so after school had let out for the year, toly my grandmother (living with us) I had an exam that day I had to get to, and ran to take a shower. By the time I got out, I’d realized it was all a bad dream, but as soon as I came out of the bathroom, my grandmother was ready to drive me back to college to take my exam.

          • mtl1882 says:

            It’s weird how we usually know the difference between a dream and real life, but occasionally the distinction just fails.

        • Rebecca Friedman says:

          Yes, yes, that dream, exactly that one. In my case, usually complete with “Hmm. Somehow I thought I was such a good student that I could afford to skip a few classes, only I kept doing it a little too late and now it’s finals and I skipped the whole course.”

          (Which is completely unlike me; I was somewhat obsessive about showing up for classes, unless I was sick in which case I was very careful about telling the professor and trying to submit work remotely. I do stress about that kind of obligation. But somehow my sleeping brain thinks it’s perfectly plausible behavior; I think I got as far as “hmm, that seems odd but OK” once.)

          I never had it during school, either. Maybe I was thinking too much about the actual classes I was taking then?

      • VoiceOfTheVoid says:

        Oh god that one’s the worst.

      • CatCube says:

        Weirdly, I’ve never had this as a dream, but I’ve had it while awake a few times. I think two or three times in college, I’d be sitting at my computer one evening just surfing the internet–I think about mid-semester–and become suddenly convinced that there was a class I was signed up for and had been forgetting to attend.

        I’d check my printed copy of by schedule that I always kept posted right by my desk, then when it showed I was clear, I’d go verify online that I did, in fact, have the most current copy of my class schedule printed. It was a freaky feeling, and I think related to a few times I’d forget to go to an individual class session (I’ve occasionally even now forgotten to go to a meeting until 10 minutes after it started).

        • HeelBearCub says:


          Fuck, that’s like a weekly occurrence for me. Must be nice.

          • CatCube says:

            Outlook pops up a message for the meeting. It only happens when I’m so focused that I don’t notice the message.

          • HeelBearCub says:

            I also use Outlook.

            I get reminders on my laptop and my phone.

            Still doesn’t stop me from missing the reminder or getting lost in “that one thing I was in the middle of”.

            Just ADD things.

          • Nick says:

            This seems like a good time to ask, how the hell do I tell Outlook I did go to the meeting?

            I get reminders on my computer/phone/etc. when I’m five minutes “overdue” for a meeting that I’m sitting in. And an hour in. And a day overdue. And it gives me the option to… ignore. But I don’t want to ignore the meeting, I actually went, Outlook. What am I missing here?

          • HeelBearCub says:

            Our Outlook setup doesn’t record attending the meeting (not sure about yours, or whether that is functionality that Outlook even offers).

            We get reminders that a meeting is going to start. This is more akin to setting an alarm that wakes you up at a certain time. I just dismiss (or snooze) that meeting notification as appropriate.

            I will frequently see a meeting reminder that I haven’t dismissed sitting in the notification center, maybe even a day later. I dismiss it then.

          • CatCube says:


            I’m confused by that: I always just hit the “dismiss” button when it pops up the warning for the final time. Do you not have that on your UI?

            My typical warning time is 15 minutes prior to the meeting, then I hit the “Snooze” button so it warns me 5 minutes prior. That’s when I hit the “Dismiss” button.

      • WashedOut says:

        I get this one and “teeth falling out while I walk down the street” about 5 times a year each.

    • Loriot says:

      Maybe I just had a lot less stress about school than most people.

      I guess a followup question – how often do you remember your dreams?

      It seems to me that people dream all the time, but only remember them if they wake up abruptly, so it seems like you only dream when you aren’t sleeping well. This morning, I had several dreams I could remember due to waking up repeatedly.

      • Rebecca Friedman says:

        Almost never; unless I wake abruptly, I rarely remember much.

        I would guess I remember about one nightmare for one or less than one good dream and 4-5 neutral dreams? I think the vast majority of my sleep is neutral dreams, from what random experiments I have (ie, if something not-stress-related wakes me, it’s practically always from a neutral dream; I think I remember all my nightmares, at least briefly, which is not true of the other categories).

      • Kaitian says:

        Directly after waking up, I can remember my dreams almost every time (I don’t use an alarm clock and don’t rush out of bed). The memory goes away after a few minutes if I don’t actively try to record it. I don’t think I remember any full dream plot that I didn’t reconstruct after waking, but I do remember some general motifs without effort.

        For the original question, I very rarely have stereotypical dreams. Mine are very often about looking for something and not finding it, but not about things like school, teeth or flying.

    • Iago the Yerfdog says:

      As far as I can tell, dreams are just a random mishmash mostly of things you’ve been thinking about lately, just with the part of your brain that notices inconsistency turned off.

      Scott suggested a while back that dreams are basically what happens when your brain stops getting feedback to check its predictions, so the predictions are allowed to run free. What I took from that was actually a vindication of the idea that dreams can be meaningful: they express how the predictor has been trained and can help you tease out things you didn’t know about yourself.

      EDIT: Also I’ve never had the naked at school dream, but that might be because I was homeschooled starting with the 3rd grade.

      • Loriot says:

        I saw a suggestion once that dreams are the biological equivalent of shuffling the input data when training neural nets.

        • Solra Bizna says:

          I didn’t get really good at Tetris Attack until I started dreaming about it. The day after my first Tetris Attack dream, I found I’d made a startling jump in skill. That’s one anecdote data point suggesting the technique works for mammal-based neural net training.

      • FLWAB says:

        Scott suggested a while back that dreams are basically what happens when your brain stops getting feedback to check its predictions, so the predictions are allowed to run free.

        I think Scott’s right on the money. A while ago I realized that in all my nightmares whatever I believed was going to happen did happen. So I would be running from a monster and think to myself “There’s no way I’ll escape” and sure enough I’d run into a dead end. Or I’d hide from a monster and think “He’s going to find me! He’s going to find me!” and sure enough he would. It occurred to me that while dreams seemed to be surprising, they never actually contained anything I didn’t expect would happen on the occasions in which I did expect something. Or, more accurately, in bad dreams something bad would happen and I would be afraid that something worse would happen, that that fear would cause the worse thing to come into being. After making these realization on occasion when a bad dream starts getting worse I will halfway remember that my expectations control things and I’ll try as hard as I can to expect good things: that I will escape, or that I can overpower the monster. And it often works!

    • DinoNerd says:

      I still have the course(s) I didn’t know I was taking dream occassionally. But they are set in a University context. (Never K12.)

      I get the naked-in-public dream too – or more likely, half-dressed – in particular, wearing what I normally wear while in bed asleep. But it’s set on the street (most common) or in the workplace, never at a school.

      I’m not sure what else is steroetypically common. But my regular stress dreams involve (1) people I’m somehow responsible for making things worse in an emergency and/or creating the emergency in the first place, while I try to deal with the problems (2) moving house, including somehow having two homes rented, but not knowing about/paying for one of them; this is often combined with household members negatively “helping”, as in (1).

      I can’t recall a case where the problem creaters were part of my current household. I get people from my past, or people from my environment, or generic walk ons.

      • mtl1882 says:

        moving house, including somehow having two homes rented, but not knowing about/paying for one of them

        This is similar to the college dream some of us have mentioned, and similar to others I’ve had. It’s also similar to the forgetting to put on clothes and going out naked dream. I wonder why it’s so common to dream that you are somehow unaware of something that is in real life quite hard to miss, while simultaneously appreciating its importance and worrying about it. I understand we dream about things that stress us out (paying bills), but you’d think the result would be dreaming about screwing up your actual mortgage payment in a more realistic way–instead, your brain invents a second mortgage/house that you are somehow not aware exists!

        • Rebecca Friedman says:

          I think some of it may be that the dreaming brain isn’t very good at keeping track of details. The same way you can be talking to one person and then it’s another, or they aren’t there, and you don’t notice the transition; the dream has you out grocery shopping or whatever, and then it forgets about clothes, and then it remembers “oh hey, being naked is bad” but doesn’t remember there were clothes before it forgot them.

          At least, that’s what seems to be going on in the only variation of that dream I ever get. Other things vanish too – I may be going to fetch something, and pick it up and head back and then it won’t be there, and I won’t notice while I’m dreaming that it vanished – or if I notice, I won’t find it odd.

          • mtl1882 says:

            Yeah, dreams are ridiculous in their non-sequitur and inconsistencies. I just find the extra house or class you don’t know about one weird because it’s not just a missing or out of order piece but a totally unnecessary addition. Most of my dreams are just pieces of things that happen to me with weird context, but every so often a dream will have the strangest, most creative twist that I would never have been able to come up with on my own. It’s like, where did that come from? I wish I could think of a good example now but I can’t.

    • bullseye says:

      I used to have a lot of naked in public dreams, but I enjoyed them. (I’d do it in real life if it didn’t bother other people.)

      I used to have flying dreams too. I’d think, “Hey, I can fly! Why did I ever think I couldn’t fly?” Then I’d wake up and be disappointed.

      Once I had a flying naked dream. That was the best.

      I’ve had the teeth falling out dream a couple of times.

    • Lambert says:

      I once dreamt that I could fly, but only if I flapped my arms really hard.

      • Rebecca Friedman says:

        For me it’s swimming through air – I can fly, but I need to do breaststroke.

        I can do it in any dream if I remember I can do it, though. Currently my top tool for getting out of nightmares.

        • Garrett says:

          > For me it’s swimming through air – I can fly, but I need to do breaststroke.

          Me too, with that quirk as well!

      • Tarpitz says:

        I don’t remember ever dreaming I could fly, but I did once have a dream where I was at the top of a cliff, needed to be at the bottom, and thinking, “I wish I was [my friend] Pam – this would be a lot easier if I could fly”.

    • A Definite Beta Guy says:

      Never had naked dreams, ever.

      On the other hand, I constantly have:
      -Teeth falling out
      -Exam at school for a class I forgot I had
      -Being told I didn’t finish a class in high school, so I have to go back, but I’m not allowed in high school, so I have to sneak in
      -Driving a car from the back seat

      Surprisingly I never have a “forgot the big report” dream, and I never have bad recurring dreams about college. They are ALWAYS high school.

      • Loriot says:

        I did once have a dream where I had to redo elementary school for some reason, although I didn’t have to sneak back in.

        P.S. I find it interesting how many people here reported the teeth falling out dream, since I can’t recall ever getting that one either.

      • DinoNerd says:

        Oh my! I’d forgotten about driving car from backseat dreams. I get those too.

    • JayT says:

      The main (normal) dreams I have are the teeth falling out and the “trip and fall and wake up just before hitting ground” one. I used to occasionally get the “forgot about a test” one, but it’s been years since I’ve had it.

    • b_jonas says:

      I’ve had the dream where my teeth are loose and starting to move, and the ones where I’m naked in a public place, both multiple times. I have never dreamt that there was a school course that I completely forgot about.

      Loriot: rarely, and when I do remember, most of the time either they don’t make sense, or they only start to make sense if I can fill in the blanks of the story when I’m half-awake.

      • Garrett says:

        I occasionally get the loose teeth one and end up waking up trying to hold my teeth in place so they don’t fall out. Basically the reverse of losing teeth as a child.

        Also, I frequently have a dream I’m doing something ordinary with my mother. Invariably, she’s telling me how I’m doing it all wrong. Real-world she’s be dead for a decade now.

    • Lord Nelson says:

      I get the “you accidentally missed all your classes and are now failing school” every couple of years, but never had it until I graduated college.

      Had the teeth falling out dream once, and it was very traumatic. Had a few more dental trauma dreams, probably because I had some dental trauma in real life when I was younger.

      From childhood through college, I had extremely frequent dreams about people chasing me and trying to kill me. They actually succeeded once. The fun alteration on this dream is when someone from my family has become a serial killer and I’m tasked with stopping them.

      I’ve only had one dream about being naked. It wasn’t the typical naked dream. Instead, I dreamed that I woke up as a naked man. It was fantastic! I always wished I was a man for a variety of reasons, but most importantly, for bring able to pee wherever I wanted. When I woke up in real life, I was pretty upset that it was all a lie.

      Most of my dreams are just nonsense cobbled together from my daily life. My most memorable and amusing dream was one about a group of squirrels piloting X-Wings. I also dream semi-regularly about buying/winning Pokemon merchandise that doesn’t exist (like a lifesize Arcanine plush) and inevitably get depressed when I wake up and realize that my dream will never be…

      ETA: I forgot about the car out of control dream. I got that one a lot as a kid, but haven’t had it since I got my license.

      • Le Maistre Chat says:

        I’ve only had one dream about being naked. It wasn’t the typical naked dream. Instead, I dreamed that I woke up as a naked man. It was fantastic! I always wished I was a man for a variety of reasons, but most importantly, for bring able to pee wherever I wanted.

        Now I’m imagining a parallel universe where random places are menaced by Lord Nelson’s urine.

      • Evan Þ says:

        I had dreams about monsters chasing me when I was a kid, but never people, and never since I turned 13 or so.

        There was one dream once that a close family member had turned evil and I needed to stop them through what was shaping up to be a spy thriller plot… but then I woke up.

        Then there was another dream, when I was around 12, where my sister was doing not evil but foolish and dangerous stuff, and I needed to stop her but she wouldn’t listen to me. That stuck with me after waking up, and I’m still convinced it was a message from God with the moral that I should be better friends with my sister. I paid attention to the message, and now we’re very close friends.

    • Well... says:

      I definitely have, or have had, these known cliche dreams:

      Teeth falling out: still have occasionally but thankfully much less frequently than I used to. It got to a point where the teeth would start falling out and I’d think “Crap, it’s just like in the dreams only this time it’s real!” which made it much more terrifying.

      School class registration: mostly set in high school but I’m back there at my present age for some reason, and I find out I’m registered for a class (almost always some kind of math) way into the semester. The big anxieties are split between being way behind on the material, not being able to find the right classroom, and having missed all the homework assignments.

      Naked in public: usually it’s more like I suddenly realize I don’t have any pants or underwear on. I’m not as aware of being shirtless; I might or might not be. But these always instantly become “quest” dreams where I have to go find my pants. Over the years these have become less anxiety-producing for some reason, as if in my dreams I stopped caring if people saw my naked lower half.

      Flying: or more like, I jump up in the air but don’t realize my own strength and wind up hundreds or thousands of feet up, then get the big rush in the pit of my stomach as I start to fall back down. Dreams where I am actually flying or floating are less common than they used to be, and back when I had them they usually had an element of lucidity, where I was just aware enough that I was dreaming that I could say “Hey, it’s one of those dreams where I can fly! I’m gonna fly over there” but by the time I got there I’d forget that I was dreaming and usually the dream would change to something else.

      Paralysis: I need to run but it’s like I’m wading through clay. Or, I need to scream but all I can manage is a loud whisper.

      Curious if these recurring themes are cliches too, or if anybody else has these:

      Breathing underwater: the dream where I’m underwater and am about to give up and drown so I inhale, only to discover I can breathe perfectly well so long as I take little shallow breaths.

      Hopping: also related to the flying dream, one where I can’t fly or even run but can cover long distances very quickly by hopping like a kangaroo.

      A clean/private bathroom: wandering around a crowded/filthy labyrinthine place trying to find a suitable place to relieve myself.

      Related to the above is the dream where I’m in a public bathroom/lockerroom that’s laid out in a really bizarre way, or where I’m in a private bathroom in someone’s home and the fixtures (toilet, shower, etc.) are designed in a really surreal way.

      Finding the Thing I’ve Wanted Lately in a windfall, e.g. in a dumpster or at a rummage sale: one example of this was shortly after my first time riding a motorcycle, I had a dream where I found one in a dumpster. I lifted it out as if it was made of plastic and got on and started zooming around. Oddly it became invisible once I straddled it, but otherwise it worked fine. I’ve also had this dream a lot with different kinds of guns.

      The buffet tease: I’m at a giant buffet; several rooms full of tables, all covered with trays of food, and I have all the plates I can carry. I pile food on each plate. There is a room full of tables all bearing indescribably rich desserts too, and I cover those as well. Then once my arms are full of plates all full of food I go to sit down, and I wake up just as I’m about to sink my teeth into the first bite.

      Huge blowup with someone I care about: usually my mom or one of my brothers. In the dream we get into a fight that’s so severe I wake up upset.

      Car destroyed in the ‘hood: I’m driving one of my old cars and I park in in a questionable neighborhood. When I return to it it’s been totally ransacked, just like that scene in the movie Downtown.

      Related to the above is the dream where I’m driving (always one of my old manual transmission cars) and the brakes/steering don’t work so I get into an embarrassing fender-bender.

      • Buttle says:

        I have had a recurring variation of the clean/private bathroom dream — I find myself in a large bathroom that looks somewhat like a Pier Paolo Pasolini set, maybe with large natural rocks in it, or really outsized fixtures. I begin to pee, and then realize that it’s not in the right receptacle, and that I’m actually pissing all over the floor, or in a trombone, or something.

        When I was really young I used to wake up in the night needing to pee, and would fall back asleep, and dream that I walked to the bathroom. When I got there I would realize that I had dozed off, and still had to get up. Sometimes this would happen two or three times in a row before I finally motivated myself out of a warm bed.

      • Well... says:

        Other recurring dreams I wonder if other people have:

        Tons of turtles: I am walking somewhere, usually somewhere nature-y and wooded with a stream or pond nearby, and there are turtles everywhere. Hundreds of them, both in and out of the water, and I have to avoid stepping on them.

        I also have a variation of this dream but with fish; I am wading through water that is dense with fish. Often the water has for some reason risen much higher than normal, sort of like how the Amazon does in the rainy season. Both in these dreams and in reality I am very uncomfortable with this arrangement, as fish brush against my legs. The fish usually start out small but are gradually replaced by bigger and bigger fish until fish that either are or resemble great whites are rushing past me and I am scrambling for higher ground.

        Crazy tides: I am at a beach, usually a narrow beach with a developed beachfront of some kind, and the tide is rising very, very fast. The waves are also huge, and within what feels like 5 or 10 minutes the water is surging well into the city.

        Half-frozen pond: I’m ice-skating on a frozen pond but there are areas that are either very thin or completely un-frozen, and I have to avoid them. In some of these dreams sharks leap out of the un-frozen parts.

        A variation on this dream is the pond that’s not frozen at all, but I can sort of run around on top of the water anyway. Usually I run on it for a while but if I stop then I plunge down into it, sometimes triggering one of the fish dreams listed above.

        I grew up on a pond, and have always been very impressed when visiting the beach, so these dreams are easy to explain.

        Sports hero: I am spectating at a professional sports event and for whatever reason I get called down to the field to fill in. I am completely dominant, but that isn’t saying much because the other players both on my team and the opposing team turn out to be no better, or bigger, than little kids.

      • FLWAB says:

        I hate the buffet tease. For me it’s almost never at a buffet, but there is always tons of the most delicious foods at a family event or something. I’m looking forward to eating sooooo much but every time I try to get some something happens and I have to do something else first (like someone telling me I need to help set the table, or do my taxes, or take the garbage out) and every time I complete a task another one comes up and the tasks continue until I wake up. So frustrating!

    • Ghatanathoah says:

      I’ve definitely had the “signed up for a class and forgot to attend,” “teeth fall out,” and “am naked in public” dreams. I’ve occasionally had the “scary thing chasing you” dream.

      One ability I’ve developed lately is the ability to stop dreams just before they get really bad. For instance, once when I dreamt a maniac was going to stab me I thought “It would be really bad if I was stabbed” and avoided the knife. When I had a falling dream I realized hitting the ground would be bad and flew away.

    • MeepMorp says:

      Aaaah, sorry, I tried to click “Reply” and I accidentally clicked “Report”! Apparently it’s impossible to un-report once you’ve clicked the button! Please don’t ban either of us, sorry Scott. 🙁

      (In answer to your question: I don’t really have the “stereotypical” dreams (naked in front of class) but I have things that are basically the same tone (everyone laughs at the talk I’m giving, scary thing chasing me) on a regular basis.)

  3. johan_larson says:

    In The Atlantic, Thomas Wright has an article about what’s likely to change after COVID-19. The author discusses possible impacts on a range of regions, institutions, and countries. In the interest of not copying too much, let me highlight two of them.

    For China:

    COVID-19 is likely to be a strategic setback for China, particularly in its efforts to make inroads into Europe and other democracies. The Chinese Communist Party suppressed early warnings about the virus and an official at the Hubei Provincial Health Commission reportedly ordered a genomics company studying it to destroy “all existing samples,” causing public-health officials to lose invaluable time that could have been used to contain COVID-19. There is widespread concern that Chinese pressure has compromised the World Health Organization’s response to COVID-19 at at a time when multilateral cooperation was desperately needed. China subsequently allowed diplomats to falsely claim that the U.S. Army developed COVID-19 as a bioweapon and used it against China. The CCP is now trying to limit the damage of its errors by providing assistance to other countries in the form of face masks, respirators, and other supplies. Many see this aid as an act of confidence, but it might be evidence that the regime feels vulnerable and fragile.

    For the United States:

    COVID-19 is a disaster for Americans. The United States now has more cases than any other country. President Trump is singularly ill-equipped to handle the pandemic. For weeks, he parroted Chinese talking points that the virus was under control, and he predicted a good ending for the United States. He did not use the time he had to increase crucial medical supplies or prepare for a surge in medical personnel. In a long crisis, many people will die needlessly and the financial cost will be in the many trillions of dollars. The world has lost whatever confidence remained in the ability of Trump’s America to take charge. Leaders have watched in horror as the administration focused the bulk of its diplomatic efforts on renaming the virus. If Trump is reelected—and his polling numbers suggest he has benefited politically from the pandemic so far—substantial international cooperation is unlikely after the crisis ends and the recovery begins. Each country will go its own way.

    • Paul Zrimsek says:

      The world has lost whatever confidence remained in the ability of Trump’s America to take charge.

      What, again? How many times does that make so far?

      • Tatterdemalion says:

        Note that the “whatever confidence remained” formulation is carefully chosen…

        • Paul Zrimsek says:

          Still not good enough for iterative use. When the time comes to peddle the same trope again in a couple of months, he’ll be sorry he didn’t say “most of whatever confidence remained”.

          • Tatterdemalion says:

            100% of zero is still zero, so unless he uses an absolute rather than a proportional term I think he’s still good.

    • salvorhardin says:

      I would love to say more in response to this about the relative merits of international cooperation and internationally-redundant systems vs deglobalization and travel controls, but I don’t know how to do so without taking this in a CW direction. Perhaps worth reposting in a fractional thread?

    • Clutzy says:

      I really just don’t see how the article is accurate in any fashion and indeed I find it self contradictory. Will C-19 negatively affect China? Maybe, hopefully IMO, but things like that that I hope for rarely happen because they require a massive mobilization of political support. I see far too many connected people spouting CCP talking points for this to be likely.

      And it contradicts point 2. Because America would have to be the leader in any such effort to roll back China using C-19 as an inflection point. If its a disaster for America (not really in evidence as it ignores that China ‘s numbers are probably not right, and it ignores that the US is the biggest Western nation in total population by a lot). The article even states, “If Trump is reelected—and his polling numbers suggest he has benefited politically from the pandemic so far—substantial international cooperation is unlikely after the crisis ends and the recovery begins. ” Guess what, in that case China will certainly come out ahead.

      Overall I would put this article in the category of, “rant by a dunce”.

      • 10240 says:

        I noticed the contradiction too: China’s reputation will be that its authoritarianism successfully managed the epidemic, as long as most of the media uncritically reports the numbers given by China—including implicitly Thomas Wright, when he writes that the US has more cases than any other country.

      • Tarpitz says:

        The reason to expect a negative impact for China relative to other countries is that we should expect a global depression and increased barriers to trade to affect net exporters more severely than net importers, not because of some fuzzy reputational business.

        • Clutzy says:

          An interesting theory, but wouldn’t China happily sac 5% of its pop in the depression to position itself properly going forward? That is what it does.

          • Tarpitz says:

            How secure does the CCP look after an extended period of low or negative growth? How confident do the high-ups feel about it? I’d be surprised if it was a risk they were happy to take.

  4. Ghenlezo says:

    Thanks to everyone who came to the Mozilla Hubs meetup today. Despite some organizational stumbles on my part, it seemed to work well. So I have decided to set up another meetup for the 19th of April.

    I think the big thing we need for next time is more people willing to do a ~5 minute talk.

    If you work in an interesting area, have expertise in an unusual programming language (like APL say) are an undergraduate or PhD in a subject you think would interest others, have a historical event or case study you want to share, or a paper or result you want to explain, please consider doing a talk. Ideally, your talk should have slides but this is certainly not mandatory.

    Not only will it help our meetups, but it is a very low-stakes way of improving your public speaking, too.
    Please click this link to register for the April 19th meetup:

    Also, there is a new rule that will take effect next meetup: your avatar cannot be larger than an SUV.

    • Scott Alexander says:

      I wish I could advertise these on the open threads, but you schedule them for Sunday mornings and the Open Threads go up Sunday afternoons. Any suggestions on how I can deal with that?

      • Ghenlezo says:

        I picked Sunday at 10:30 in case anyone from Israel wanted to join, as LessWrong Israel’s Zoom meeting is what inspired me to set this up. And weekdays are not an option for me.
        I’ve already sent out the emails for the 19th, but going forward I’ll schedule them to be on weeks that don’t correspond with your visible open threads. If going forward you could put in a plug on the open thread a week before that would be more than enough.

  5. Andrew Hunter says:

    What SSC mailing list, and how was I not cool enough to get invited?

    • Scott Alexander says:

      There’s a box that says “subscribe via email” on the top right, but it just sends out automated emails about blog posts. I’ve never used it for anything else, and I don’t even know if I can. It was just an example.

  6. Rebecca Friedman says:

    So, I wanted to say thank you, everyone. During this whole situation, SSC has probably been my most reliable source of data, which has been really useful for figuring out what actions to take to best improve safety. And, honestly, between Scott and other rationalists, we got enough early warning to stock up on supplies before everyone else started panicking, and to get a lot of the adjustment out of the way early. (And to start staying home early.) I think my family’s risk was significantly lowered just from the fact we hang out in these circles.

    “Predicting that something that seems socially silly is actually true and important” is the kind of thing you would naively expect rationalists to be good at, and I am now adjusting my priors in the direction of that being correct.

    Thank you.

    • albatross11 says:


    • Conrad Honcho says:

      You folk are the best place on the internet.

    • gleamingecho says:

      actually true and important

      This seems like it’s still pretty TBD to me.

      • Rebecca Friedman says:

        I’ve been tracking data. Based on what I’ve read of the Italian data, my parents are in the high-risk age group. Based on iirc Diamond Princess data, I think Dad’s age group came out to a 9% chance of death, conditional on catching it. Diamond Princess data counts asymptomatic cases, but even if that was actually Italian data and it’s 4.5% or something, that is a deal higher risk than we usually deal with. I’m pretty happy with taking steps to prevent it from applying.

        I’m not arguing the US Government’s response is correct, but I think as private individuals, we were definitely correct to change our behavior over this level of risk, and it was definitely a very good thing that we got the data we needed to find that out. And my life, personally, would be somewhat better right now if I’d taken things more seriously earlier – specifically, stocked up completely about two weeks before I did. It would be significantly worse if I’d taken things less seriously.

        That meets my standards for “true and important.”

    • During this whole situation, SSC has probably been my most reliable source of data, which has been really useful for figuring out what actions to take to best improve safety.

      Seconding this. I sometimes forget not everyone has access to this information – I try to share the especially succinct and good comments with my colleagues, but I err on the side of not sharing*, which is probably not great. But in any case, the information here has helped me personally (if mostly psychologically, but that’s also very important) and (hopefully) others that I’ve linked to it.

      (* born of the knowledge that if I share too many links in the relevant channels, people leave the channels and I stop being able to reach them easily at all – but that’s more of an aside)

  7. Lambert says:

    Speaking of nuclear winters, many people have proposed that we geoengineer our way out of global warming by putting soot in the stratosphere, thereby causing global dimming.

    If we get hit with a nasty nuclear/volcanic/asteroid winter, can’t we just make a load of haloalkanes (non ozone-depleting, short half-life) and vent them directly into the atmosphere to cancel it out?

    • The Nybbler says:

      If we get hit with a nasty nuclear/volcanic/asteroid winter, can’t we just make a load of haloalkanes (non ozone-depleting, short half-life) and vent them directly into the atmosphere to cancel it out?

      This doesn’t seem to be the sort of thing we’d be able to do after getting hit with WWIII/VOLCANOS!!!/Sweet Meteor of Death. Except maybe VOLCANOS!!!, if they were conveniently somewhere the direct effects didn’t kill too many. In that case, we’d find our completely hidebound regulatory system would prevent it.

    • bean says:

      My concern there would be other effects not related to temperature. Specifically, reduced sunlight, which could impact things like crop growth. It’s probably better than nothing, but it’s not going to just cancel out.

    • fibio says:

      I’m not sure there’s anything within current human power to increase the global temperature by the ~5 degrees Celsius needed to counteract the cooling effect, especially with a caveat that the increase has to be temporary. Two hundred years of indiscriminate pollution has increased the average temperature by about 2 degrees Celsius, so you’d basically need to do that again two and a half times over.

  8. thisheavenlyconjugation says:

    Can someone explain why private equity/LBOs work? I get the mechanics of how buying a company mainly using debt that you then attach to that company would be very profitable, but I don’t get why it’s possible. Why don’t lenders charge more interest? Or is it just the tax advantages? (I refuse to believe that “PE investors reliably generate large amounts of value” is the answer).

    And what’s the deal with asset stripping? Is that just PE investors having higher time preference than the market as a whole?

    • broblawsky says:

      They don’t directly transfer the debt from the PE firm to the purchased company. Instead, the PE firm charges the purchased company large fees, which they use to pay off their debts. These fees are supposed to cover “administrative services” that the PE firm provides the purchased company. The PE firm might also charge the portfolio company a “transaction fee” during the initial buyout.

      • thisheavenlyconjugation says:

        I thought (at least part of) the debt was secured by the purchased company’s assets/cashflow.

        • broblawsky says:

          Yes, but that isn’t unique to PE firms. Most mergers & acquisitions are debt-financed, and the resulting merged company secures that debt with some of the assets of the new company. PE firms typically put up some actual money as an equity stake during the leveraged buyout – around 10% of the total value. However, once the PE firm has acquired the company, it can transfer money from the portfolio company to the PE firm via fees, which restore the equity the PE firm sacrificed, giving them control of the company essentially for free. Then the PE firm sells the company off via an public offering, while making sure the portfolio company takes on the debt from the buyout.

          • Loriot says:

            The risk the PE firm takes on is that they overestimated the value of the company, and hence the creditors come calling and they have to give up their stake or double down.

    • Loriot says:

      The idea behind private equity is to take over companies that are being run inefficiently, improve them, and sell them at a profit (including any profits extracted from the company prior to sale). It’s the M&A equivalent of house flipping.

    • Anon. says:

      PE investors reliably generate large amounts of value. If they didn’t, either the previous owners or the lenders would be getting screwed, which obviously can’t be happening continuously over the long term.

      • thisheavenlyconjugation says:

        Objections, in increasing order of compellingness:

        That doesn’t fit with my perceptions of discussions of PE, which seem to focus almost entirely on the financial engineering side of things and not much on generating value.

        It doesn’t fit with the many examples of LBOs leading to bankruptcies. I believe there are studies that bought-out companies do actually go bankrupt more common than comparable public ones, it’s not just that buyouts target low-quality companies. And some cases the mechanism is clear — the new owners pile on too much debt. I don’t think “bankruptcy does generate value because company’s assets go on to be used more effectively elsewhere” is the answer; you wouldn’t say a CEO of a public company presiding over a descent into bankruptcy is generating value.

        If that were true, there would be a booming trade in ex-PE people being employed as executives by public companies to generate value for them. But AFAIK this doesn’t happen (I’m sure there are a lot of execs with PE backgrounds, but not any more than from similar backgrounds like IB).

        • Loriot says:

          I believe there are studies that bought-out companies do actually go bankrupt more common than comparable public ones, it’s not just that buyouts target low-quality companies.

          That’s probably just a result of the biggest deals being done in times when people are especially optimistic. I wouldn’t be surprised if the same thing happened for public mergers and acquisitions as well.

        • JayT says:

          My impression is that PEs target distressed companies or extremely fast growing companies, and not companies with steady cash flow and lots of assets. So, it wouldn’t make sense to compare companies bought by PE firms to all companies, since the safest companies were never a target for PE firms in the first place.

          • HeelBearCub says:

            Different kinds of PE firms do different things. Lots of PE firms do things like buy companies for extremely large retirement portfolios.

            Berkshire Hathaway is a PE firm.

          • JayT says:

            Berkshire is a holding company, which is kind of the opposite of a PE. They look for stable companies that have predictable profit, and they just hold onto it and reap the rewards.

    • Tuna-Fish says:

      PE investors do reliably generate large amounts of value. In fact, typical PE deals are planned in advance so that they are only executed if they generate a lot of value instantaneously upon execution.

      PE has a horrible reputation, because to outsiders a typical PE deal looks like horrible jerks coming in and breaking stuff. This is because a lot of PE deals are done in the special cases where companies contain parts that are not just valueless, but are actively destroying value. A typical example is some very old, large company that vertically integrated a lot of things, and where the main business is no longer competitive, and is in fact producing massive losses year after year, but where the company as a whole is held afloat by some of those vertically integrated areas being competitive, essentially subsidizing the losses of the rest of the company, sometimes for literal decades. PE coming in, purchasing the shop, shutting down all the unprofitable segments and freeing the profitable ones to be their own companies is in fact generating a lot of value.

      And just to reiterate, this situation is not rare. Given how many large, old companies are effectively many smaller ones joined together at the hip, and that the expected outcome of any given company is to eventually stagnate and fail, ending up in the situation where a lot of the business is on permanent life support is not rare at all. At that point, smart leadership shuts it down and concentrates on things that still produce value. However, especially for traditional old companies that have a lot of history, shutting down the primary business segment is sometimes a choice that the leadership cannot or will not do, for reasons other than rational economic optimization.

      Asset stripping is typically not a case of PE investors having a higher time preference, but the simply the case where the market value of a company is substantially below the cumulative value of it’s assets.

      • Loriot says:

        Another case where that kind of thing might happen is empire building by overzealous CEOs. If a company branches into a lot of nonsensical side businesses, it makes sense to split it up.

      • Etoile says:

        But it feels that way, anecdotally. To be fair, I can’t see the bottom line of a lot of companies, but it’s happened to several brands that I’ve liked and have been good, and then they got worse after someone bought them out and restructured them.
        As an analogy, it feels like PE grows beautiful, tasteless tomatoes — accelerates their growth, but they don’t have time to fill up with nutrients or flavor; just volume — similarly, PE positions a company to show rapid near-term growth, at the expense of longer-term, real value and quality.

    • Rick Jones says:

      I have a related question. Why does PE generate returns greater than the market as a whole? In other words, why can’t the PE strategy be used by large public companies, like life insurance or uni endowments? Scott asked this question in his review of Picketty’s Capital in the 21st Century but didn’t really answer it. I’ve never been satisfied with what I’ve read, like that PE is more ‘agile’ etc.

      • Luke G says:

        Endowments do invest in PE funds. I don’t know about insurance; it seems like PE would be a good fit for them, but they are highly regulated and might not be able to.

        The justification for PE generating above-market returns is that it’s an extremely illiquid investment. Investors often are locked in for years. Nobody wants to invest in something so illiquid if it doesn’t offer substantially higher returns. Moreover, many investors *can’t* invest in something so illiquid.

    • Aapje says:

      It seems to me that the market underrates the risks & costs of debt, so taking a company that is worth $X and extracting wealth by debt valued $Y, doesn’t actually make the company worth $X – $Y as you’d expect, but more than that.

  9. Bobobob says:

    So trying to look for a bright side to all this, I don’t think we could have asked for a better time for a three-month shutdown (at least in the U.S.) than from mid-March to mid-June:

    –The weather is getting nicer, the days are getting longer, so people pent up in their houses are getting more sunshine
    –We’re in the middle of the annual Holiday Drought; there’s nothing major coming up except Memorial Day and July 4, and those are both second-tier travel and get-together holidays
    –We’re far enough from the end of the fiscal year (the end of September for most companies, unless I’m mistaken) that no one needs to make any hasty decisions about next FY until the fall
    –Kids have already gone through two-thirds of the school year, so they know the basic curriculum and what they’re expected to learn for the remainder of spring

    Can you imagine if a lockdown went into effect in mid-October? Then we would be faced with:

    –The complete overlap of Covid-19 and the ordinary flu (which starts petering out in the spring). I can imagine lots of people getting the flu, freaking out thinking it’s coronavirus, and overloading emergency rooms
    –The long, cold creep into winter, exacerbating the situation for folks with seasonal affective disorder and greatly increasing the Claustrophobia Quotient
    –The complete cancellation of Thanksgiving and Christmas, which I imagine would be three or four times worse for the economy than shutting down in the spring
    –About a billion editorial cartoons showing Santa Claus putting on a face mask before he goes down the chimney

    I mean, I’m about to completely gnaw off my own right leg, but it could be worse, right?

    • Loriot says:

      It was also fortunate that the Democratic primary was effectively wrapped up before the lockdowns started. Starting it a month earlier would have been a nightmare.

    • kupe says:

      **Cries in New Zealand.**

      Although here people are saying that we’ve got the best time for it:

      – Rainy winter days make staying at home easier.
      – Flu season starting with people social distancing and washing hands should mean fewer flu deaths.
      – Schools are missing the start of the year so there’s less upset around exams.
      – Tourism peaks in summer, so the hit to that large sector is somewhat lessened.

    • bullseye says:

      I would have prefered the shutdown to start in January, so we’re stuck inside when we don’t want to go outside anyway. Also, Valentine’s Day is better than the Fourth of July for being stuck inside, since you probably live with the person you’re supposed to spend Valentine’s Day with.

      The Federal government’s fiscal year ends Sept. 30 for reasons related to Congress’s budget process. I don’t think it’s especially common in the private sector. It’s my understanding that December 31 is the usual year end unless the business does a lot of Christmas sales (they want the sales and returns in the same fiscal year) or it’s a seasonal business operating in the winter.

    • Bobobob says:

      Oh, and I forgot to mention: it’s kind of a sports drought here, too, unless you’re a basketball fan. There still may be time to get an abbreviated baseball season going, and football will probably be safe if we get our act together. (I do miss the English Premier League, though–and I feel really bad for Liverpool.)

      • Tarpitz says:

        Laughing at Liverpool is 100% the best thing about the entire business, though I suppose it’s possible that it’s also the reason for the outbreak in the first place: we may be living in Daragh Curley‘s version of It’s a Good Life.

        • Deiseach says:

          I’ll overlook the “laughing at Liverpool” part because I’m feeling that we’re feeling oddly vindicated: earlier in the season, everyone was “Why are you Liverpool supporters so paranoid? Why can’t you accept that you have the league won? It’s not like anything can go wrong now!”

          Oh really? Oh really? Who’s paranoid now, huh????

    • AlexScrivener says:

      Easter is also a pretty big holiday in some parts, and is usually celebrated in groups, often in large crowded buildings which are currently closed.

      • Le Maistre Chat says:

        And what’s to stop the government from cancelling Christmas as well as Easter? There won’t be a vaccine available any time this year even if one has a successful human trial.

        • The Nybbler says:

          And what’s to stop the government from cancelling Christmas as well as Easter?

          I can see Trump shotgunning the Easter Bunny like Dick Cheney shot that lawyer, but the Grinch is clearly green, not orange.

          • Conrad Honcho says:

            The kanji for “apologize” (謝) is a combination of the primitive/kanji for “words” and “shoot.” It always makes me think of Dick Cheney.

        • Tarpitz says:

          Will they also be doing away with alms for the poor and merciful beheadings?

    • A Definite Beta Guy says:

      Eh, could be worse, but it’s still pretty bad. In Chicago, we pretty much go indoors in October, and start coming out on the sporadic nice days: it’s one of the reasons you saw all those joggers along the 606 trail through Twitter a few weeks ago.

      This is like a winter that’s twice as long, especially for city-dwellers who do not have yards to enjoy.

    • skybrian says:

      Seems like you missed a big one: it’s not hurricane season.

    • myst_05 says:

      Thing is, the shutdown will last until we have a vaccine and that’s 12 months away… So Christmas/NYE 2020 are most likely cancelled, as well as the first semester of school in September.

      • Bobobob says:

        I don’t think there’s any conceivable way the shutdown will last that long. No vaccine = vastly changed way we interact with the outside world, not indefinite lockdown. We are not talking about the Black Death here.

    • John Schilling says:

      –The weather is getting nicer, the days are getting longer, so people pent up in their houses are getting more sunshine

      Have roofs been getting more transparent? Because from what I’ve seen on my occasional outdoor excursions, the people pent up in their houses are mostly literally pent up in their houses. And I live in a suburb where most everyone has a yard and sidewalks abound.

      • The Nybbler says:

        Have roofs been getting more transparent?

        I’ve got skylights.

      • HeelBearCub says:

        I mean, I have windows in my house. And a yard.

        I imagine even the fire escapes of NYC are nicer this time of year, especially without much in the way of traffic.

        • John Schilling says:

          Are people in NYC actually hanging out on their fire escapes any significant fraction of the day? Here in Southern California, they appear to be hanging out on their lawns less than they did before the pandemic. I believe the claim “people are getting more sunshine” is armchair theorizing about how people should behave, contradicted by observation of how they actually do behave. But I would welcome more evidence.

          • Bobobob says:

            Not armchair theorizing. A bunch of people in my neighborhood (including ones I don’t ordinarily see) have been taking advantage of the nice weather to walk, work in their yards, sit outside in lawn chairs, etc.

            I guess where the word “armchair” applies is that I assume it’s more psychologically healthy to have the windows open on a nice spring day, rather than closed because it’s 40 degrees out. But I’m just one person, I have no idea how most Americans live.

          • Byrel Mitchell says:

            I’m certainly spending more time with my kids in the yard and on walks down the street. And some of the neighbors down the way are riding their horses around a lot more often.

            But I’m pretty rural; I wouldn’t expect suburbia and urbia to react the same necessarily.

          • gph says:

            More anecdata from a mid-atlantic suburb: I’ve definitely seen a lot more people out walking/biking or working/playing in their yards than usual, especially on the weekends.

    • Randy M says:

      –Kids have already gone through two-thirds of the school year, so they know the basic curriculum and what they’re expected to learn for the remainder of spring

      Ha, count this as a negative if what I heard is true–kids around here are going to need to repeat this entire school year.

      • Bobobob says:

        I would think school districts would do whatever they possibly can to keep kids from having to repeat grades. I haven’t heard anything like that in my district (though I can imagine that summer will see the rollout of an “official” curriculum and testing program for the grade my kids are currently in).

      • Beck says:

        Looks like some states are just graduating high school seniors who were passing at the time schools were closed down.
        Must be kind of an anticlimax for the kids.

        • Randy M says:

          That’s much better than pretending Seniors learn anything in the last two months of high school anyway.
          Shame for prom, graduation, etc., of course.

          • Matt M says:

            Yeah… I mean I was a nerd with few friends and my last few months of high school were probably among the most fun, relaxed, and care-free of my life.

            One of my cousins is a high school senior right now. I feel bad for him. A lot of cool memories are being stolen from him. He can never get that back.

  10. indigo says:

    I *think* the point of an IRB is to control the liability to the organization in question, and that they do this by the imposition of a code of ethics, and require that all research plans be reviewed to be in conformance.

    If this is true, then not having an IRB might not directly make it more likely for you to get in trouble, except insofar as they might warn you away from a bad course of action. (And it might be considered recklessness, if you do get in trouble for some other reason.)

    (There, now maybe someone who actually knows will correct me, for that is how the internet works.)

    • Matt M says:

      Yeah, whether this matters seems entirely contingent upon whether Scott is actually affiliated with some sort of IRB-having institution.

      If not, go nuts. It’s not like, a legal requirement, it’s an institutional one.

      If so, it probably depends on how much your institution likes you in general, and what you’re doing in specific. Recall that one of the professors who wrote the papers in the infamous “grievance studies journal hoax” was formally reprimanded and disciplined by his institution, ostensibly for violating IRB requirements by conducting an “unauthorized experiment involving unwilling human participants.” Which in a certain sense is plausibly correct, but in another sense there’s zero chance anyone would describe it that way if his political ends weren’t opposed to their own.

      • johan_larson says:

        I’m guessing the biggest risk for Scott is that someone might come for his medical license, either because someone has a grudge against him, or the licensing board feels the need to defend its scope of jurisdiction. Or, I suppose, because he does something that might actually put people in danger.

        • Matt M says:

          Yeah, given that he’s famous-ish and political-ish, that’s a very plausible scenario.

    • BlueGold says:

      Its partly liability and partly a scientific convention. Journals won’t publish anything that doesn’t go through IRB and if a research institution wants it’s people to be published then they need to have an IRB. Beyond just that, most scientists wouldn’t want to work for a place that didn’t insist on IRB because they’re an important part of ethical research practices.

      In Scott’s case it’s not so much about personal liability as much as it is that his data would be much more suspect without it. Journals will not publish anything he generates and few in the larger community would be that interested in supporting it. IRBs are not legal bodies, but there are laws associated with ethics in research that IRBs cover and go beyond. Skipping an IRB likely will not run you afoul of those laws (although given he would be working with human subjects this is debatable), but it would be a sign you don’t care about research ethics which is a red flag.

  11. albatross11 says:

    Corona-related comment:

    Someone should just give everyone working in delivery and car pickup at stores surgical masks. The only interactions I’ve had within 2m of anyone outside my immediate family in the last two weeks has been people bringing stuff out to my car (groceries and other stuff from Target) or giving me something through a drive-through window (a pharmacy). That’s probably 90% of my potential exposure to COVID-19 at this point. Putting masks on them would probably reduce the potential of having someone doing that job who’s a superspreader.

    • The Nybbler says:

      There aren’t surgical masks available.

    • AlexanderTheGrand says:

      Fully agree with the sentiment of this. And, I’m currently making masks to soon distribute to people working high-contact jobs. I’m sure an effective altruist could find a better use of my time but it’s something.

      But really, people should just be using whatever they have at home to cover their faces. If people wanted to, they’d just use a bandana, or a shirt, or a scarf. While a multi-layered, fitted home-sewn mask is better, both are so much better than people walking around with nothing. I don’t know why we don’t see more of that.

    • Purplehermann says:

      Definitely a good idea

  12. peterispaikens says:

    Regarding the non-IRB issue, my expectation would be that any data arising from that study would be unusable in the normal process of research and scientific communication. I.e. most respectable journals would disallow publishing any research results arising from data of that study if it did not have IRB approval. This means that not only other researchers would be not likely to read these results (because the usual publication venues and indexes would not have it) and it would not get included in reasonable surveys or metastudies, but also if they knew of these results, they could not base their publishable work on this data, wouldn’t cite it, etc – so in some sense it would be intentionally excluded from the “general scientific knowledge” because it does not pass this bar; and if it provides some interesting result then it would get “properly” disseminated only if/when someone else did a proper study, and they would get all the credit for that.

    Also, your employer may have explicit policies regarding IRBs making arbitrary requirements when they do/don’t apply and they may fire or otherwise discipline you for breaking these policies. Also, if you’re working on research project funded by some government agency, that funding in many fields includes a legal requirement regarding IRB process in that research.

    • Fakjbf says:

      It would still be useful in that if it gave positive results another researcher might be able to point to it when they apply for grant funding, as a proof of concept that there is a potential lead here.

      • BlueGold says:

        Nobody would point to them. Not if they want funding. Getting funding is a difficult process already. Suggesting that you’re basing your ideas on a research study that doesn’t bother with an IRB? That’s a bad sign. It doesn’t take a lot of bad signs to reject a grant application to begin with, adding one like that is a terrible idea.

  13. albatross11 says:

    Article (PDF) about the difficulties of making a COVID-19 vaccine. I think a variolation strategy would be needlessly risky to both the volunteers and the surrounding community, but testing these experimental vaccines that might provide protection but might also make things worse seems like a better bet.

    A lot of this article talks about ADE, but as I understand it (I’m an amateur, not an expert), ADE has only been observed very rarely in practice, so I think it’s not actually a huge risk. They talked about causing a damaging immune response to the virus (something that doesn’t inactivate the virus, but instead just increases inflamation around it so that the patients get a lot sicker). I think another possibility is that the vaccine might cause the immune response to go down the wrong path–a mainly antibody vs a mainly cellular response–I think there have been failed vaccines that made people sicker with the disease this way.

    If you want to go really high risk-tolerance, instead of variolation, give sets of volunteers each a different experimental vaccine, wait 3 weeks, test their antibody titers (or however you check for cellular immunity–I’m not sure what they use for that), and then if they look good, challenge them with COVID-19. That probably gets you good data on the effects of the vaccine within a few months, at the cost of being done in by an angry mob of bioethicists when you’re done saving the world.

    • Purplehermann says:

      Hey, I still don’t really get why you think variolation (lesser exposure, not an attenuated strain) is so risky.

      Assuming people are paid/volunteer and will be quarantined properly.

      Running a test to see if the lower exposure seems to be working (few severe cases, etc) on ~10,000 people should end up with less than 10 dead even with pretty rotten luck and variolation isn’t useful.

      • Clutzy says:

        I agree. If a vaccine is more than 180 days away at this point, still, we absolutely cannot wait for it. If that is how long the timescale is, the most economical strategy would be sending out notices to 20-30 year olds that they need to stock up on Date X, then an injection team will appear on date +1 and they will have to stay inside for 14 days.

        That is if the doom and gloomers are right, and they are being treated as right by the rest of the government apparatus.

        • Purplehermann says:

          I would actually prefer using hotels etc.. where possible, but yes. Isolation of these people needs to be taken very seriously.

          Should also probably get teens too, and tell the groups to get some good sleep and eat healthy the days before and through the infection

          • Edward Scizorhands says:

            Where do you house the teens?

            If I went the route of infecting the 20- and 30-something, I’d obviously start with volunteers.

          • Clutzy says:

            A hotel is an interesting idea. But overall the point is the same. The vaccine has such a long timeline to be essentially worthless outside of eventually inoculating the very vulnerable who havent yet died.

          • albatross11 says:

            We house the intentionally-infected teens in a hotel, of course. We get both a pool of immune volunteers, and a baby boom. What’s not to like?

          • J Mann says:

            If you work it right, you can get an MTV reality show out of it too.

          • Gobbobobble says:

            If you’re going to be forcibly infecting your citizens why not just weld them into their homes?

          • Edward Scizorhands says:

            The “long timeline” for the vaccine is that we are giving it to volunteers and then watching them for bad side effects, which can take a while to surface.

            At least if we give people vaccine candidates, they can be part of a controlled trial to figure out how well it works for other people.

    • Rick Jones says:

      Very informative, thank you.
      I’ve been listening to the virology podcast TWiV (the week in virology) which I recommend.

      They mentioned the mRNA vaccine mentioned in the article.

      • albatross11 says:

        The most recent TWIV I’ve listened to (I’m a little behind) discussed the fact that other human coronaviruses basically never produce long-lasting “sterilizing” immunity–where you are exposed to the virus and just clobber to and don’t get sick at all and don’t pass it to anyone. Instead, they seem to produce a couple months of that kind of immunity, and then you can get re-infected, but you fight it off more easily the second time. If that’s also true of COVID-19, then the herd immunity strategy for young/healthy people isn’t going to work as a way to protect the old and sick.

        Also, this offers a plausible explanation for why you might see people in China who tested positive and later tested negative, and then still later tested positive. They could just have had their immunity lapse. (Though false negatives or false positives are definitely a possibility.)

        I can see a few ways this might work out:

        a. COVID-19 gives you relatively long-lasting immunity–at least a couple years. Then once we have a vaccine, it becomes like a flu vaccine, and acquiring herd immunity will work to stop its spread.

        b. COVID-19 gives you a few months of immunity, and then a previous infection makes you better at fighting off the next one. Then, vaccination will still be valuable, but herd immunity won’t be available.

        c. COVID-19 gives you maybe a month or two of immunity, and then you can get re-infected just as badly. Then, vaccination won’t help.

        Both (b) and (c) suggest that we need better treatments pretty urgently. If it turns out that we can’t get herd immunity, then everyone probably gets it sooner or later unless we totally snuff it out, which looks extremely hard. If we live in that world, then we need to have something better to do than ride out what feels like a bad case of the flu, and then wait to find out if you’re going to be one of the unlucky ones who dies of pneumonia.

        If (b) or (c) is true, it’s also going to be 2-3 years at least until we can go back to anything like normal.

        • albatross11 says:

          Just to add: a world in which herd immunity is unavailable and the disease either still makes you very ill on subsequent infections or there’s a large pool of never-infected people around is probably a world where travel between countries is a whole lot more strictly limited. If Japan or Singapore or Taiwan or wherever keeps the disease from spreading, they’re going to be *very* picky about whom they let in. I wouldn’t be surprised to see a rapid COVID-19 test (nasal or throat swab, like the rapid flu test we have now) required in order to be allowed to enter many countries.

          • John Schilling says:

            If herd immunity is unavailable (and we don’t have a ubiquitous effective vaccine), aren’t there going to be infected people already and continuously in pretty much every country? What’s the use of keeping out potentially-infected foreigners if you’ve already got a local pool of carriers starting off new outbreaks every fall or whatever? And if the idea is that there are to be literally no domestic carriers so that keeping out infected foreigners is either necessary or sufficient to prevent local outbreaks, how do you get there from here?

          • albatross11 says:


            I’m assuming some countries do limit it. If so, they will be imposing a lot of restrictions on entry.

            Similarly, if there are some countries that have done the test-and-trace thing pretty effectively, so that they don’t have much community circulation going on, they’re going to be pretty unwilling to have people from less-controlled places coming willy-nilly into their relatively safe country and bringing lots of new cases.

            Suppose Iceland has done a solid enough job that they’re down to only a few new cases a week, but the US has widespread community spread of the disease. Are they going to be interested in having a lot of Americans fly there and spend a week doing tourist things? Probably not unless they can somehow be pretty sure those Americans aren’t bringing them 50 new cases next week among the employees of the hotels, restaurants, museums, etc.

            If test-and-trace is workable, it will probably be done passably well in most parts of the US. But probably not as well in Mexico, and probably not well at all in El Salvador. There’s an obvious solution there, but it’s painful for the agriculture industry and a very hard pill to swallow for a lot of Democrats.

            I don’t know how well it would work legally or practically to restrict people from different states coming in. But if Mississippi has done a crap job controling COVID-19 and Oregon has done a good job, you can imagine Oregon wanting to test out-of-state drivers coming into the country–at least with a fever check and maybe with a swab and rapid test or something.

          • John Schilling says:

            The only way Iceland can keep this or any other local infectious disease at the few-cases-per-week level is if something is locally suppressing exponential growth of that disease at an early stage. In which case, that same effect ensures that any foreign threat will be linear, not exponential.

            And, consider the relative magnitude of that linear threat. In the United States, foreign visitors in the first two weeks of their stay make up less than half a percent of the physically-present population at any time. Unless those foreigners are coming from someplace with two hundred times the COVID-19 prevalence, they represent a minority of the total risk, and that total risk is something we have presumably come to live with. In tourism-heavy Iceland, as much as 10% of the population may be recent foreign arrivals, but A: they’re mostly from first-world nations that are unlikely to be doing even one order of magnitude worse than Iceland at infectious-disease control and B: that ratio is as high as it is in Iceland precisely because Iceland has a very tourist-dependent economy.

            The United States might crack down on travel from non-first-world countries, but that would have more to do with nativist sentiment than any practical disease-reduction benefit. Iceland chasing off its tourists over such a marginal threat, seems unlikely.

        • Garrett says:

          Even in the case of (c), why couldn’t we just adjust to a life in the medium-term where everybody has to get a vaccination booster shot every few months? Other industries involve routine medical procedures (STI testing in the porn industry comes to mind). It’s not exactly efficient, but at ~$20/pop it would be part of the “cost of doing business” for most people employed.

  14. Pandemic Shmandemic says:

    Have any COVID19 statistics been gathered so far to allow us to estimate the following funnel:

    A = P(Requires non-invasive ventilation | COVID19 confirmed)
    B = P(Requires invasive ventilation (intubation) | A)
    C = P(Death | B )

    I’m assuming that patients who leave the funnel at any stage prior to C do so via recovery, that is there’s no significant number of deaths while on non-invasive ventilation.

    The idea is to try and quantify the impact of a ventilator shortage on covid19 mortality, is there anything crucial that I’m missing here ?

    • Chris Phoenix says:

      “Ventilator shortage” may also include shortage of technicians, drugs, and even oxygen.

      Also, non-ICU hospital bed shortage may preclude availability of non-invasive oxygen (which may be necessary to prevent death or serious damage) and monitoring to apply stronger measures if necessary.

      Check out this for some context and statistics. “I had never discharged multifocal pneumonia before. Now I personally do it 12-15 times a shift.” and “81% mild symptoms, 14% severe symptoms requiring hospitalization, 5% critical.” It’s anonymous (it claims to be by a New Orleans ER MD) but detailed and seems consistent with other things I’ve read.

      • Pandemic Shmandemic says:

        Thanks, shortage of non-invasive oxygen should also be similarly quantified although as your link suggests there is a lot more leeway with prescribing it.

        But it is possible in principle that COVID19 patients who get to a stage requiring invasive ventilation are 99.99% doomed even if they get a ventilator, a bed and trained staff – in which case pushing for more ventilators and associated resources is not going to help.

        From your link: “worldwide 86% of covid 19 patients that go on a vent die, Seattle reporting 70%” are the numbers I was looking for – it’s not as bad as 99.99% but should probably still be taken into account when putting resources into ventilator acquisition vs. other avenues as well as with allocating existing ventilators for COVID19 patients vs. other conditions that require them.

        Also given how much time has passed i wonder whether Seattle’s 30% have recovered or some are still on the ventilators.

      • SamChevre says:

        non-ICU hospital bed shortage may preclude availability of non-invasive oxygen

        Can you talk more about this? It seems like supplemental oxygen is given pretty often without hospital admission (both portable oxygen tanks, and “run a bit of oxygen and put a pulse-ox meter on them in a hallway (which I’ve sat with my croupy child in an ER where that was the protocol)–so I feel like I’m missing something.

        • Chris Phoenix says:

          You’re right, sorry, I over-extended, using “non-ICU hospital beds” as a proxy for general availability of medical resources.

          SCUBA divemasters are (probably? necessarily?) trained in administering oxygen in the field, for short periods of time. I don’t have a good sense of what resources can be repurposed and stretched.

          I do know that some COVID cases that don’t need ventilation need a lot of oxygen and monitoring. My dad (in a hospital in San Diego) may escape needing a ventilator, but I don’t know if a non-medical person with an oxygen tank could have kept him alive for the past few days. And it would have taken a lot of tanks – he’s been getting 60 liters per minute via cannula and 100% O2 via BIPAP.

          • albatross11 says:

            I think there are also worries that high-pressure oxygen, CPAP machines, etc., probably create an aerosol. Which means you don’t want them done in hospitals where the tiny droplets are going to float down the hall and infect the guy in the next room.

    • CatCube says:

      I know this website from the University of Washington predicts numbers of cases, numbers of hospitalizations, numbers of ventilators required, and number of deaths per day, which would imply they have assumptions for the quantities you require. I admit I’ve just been pushing the “I Believe” button for the information on their website, but they have the paper for their model here. (Hey, I’m a decent structural engineer, but I’m not afraid to admit I don’t know epidemiology and defer to experts.)

      Right now, they say that we’ll have peak deaths in the US at 2,634 deaths/day on 17 APR, with 93,531 deaths total, with a 95% confidence interval of 39,966 to 177,866. The peak ventilators required is 32,518 on 12 APR, the peak ICU beds required at 40,267 on 14 APR, and the peak hospital beds required at 262,092 on 15 APR. You can get state-by-state breakdowns of predictions as well.

      • Pandemic Shmandemic says:

        The requirement for ventilators is calculated based on how many people are expected to need one, it doesn’t address the difference in outcomes between a scenario where everyone who needs one gets one and hospitals running out of ventilators (but not beds for that matter)

      • The Nybbler says:

        That website predicts that New York state was out of ICU beds on March 19, and out of all hospital beds on March 24. Neither of those were the case. It should be viewed as having predictive abilities as somewhere between those of the late “Miss Cleo” and the “Magic 8-Ball”. It’s not necessary to know anything about how their model works to know it’s drastically wrong.

    • bpodgursky says:

      This is the closest I’ve seen to (C). In China:

      > Of the 22 who eventually required mechanical ventilation, 19 (86%) died. Another early study reported 31 of 32 (97%) mechanically ventilated patients died.

      In the UK:

      > Of the 98 patients who received advanced respiratory support—defined as invasive ventilation, BPAP or CPAP via endotracheal tube, or tracheostomy, or extracorporeal respiratory support—66% died.

      My guess is that you could only get data on A from a very heavily-tested country — maybe Korea, or the cruise liner? Either way, I can’t speak for the numbers for A or B.

  15. Chris Phoenix says:

    Can we, as rationalists, reach some kind of science-based and history-based consensus on nanoscale technologies and molecular manufacturing? A few days ago someone posted that nanoscale machines would have surfaces like “sticky unvulcanized rubber.” This is at best 50% correct for some materials, 90% incorrect for others, and cannot be a substantial argument against the feasibility of engineered nanomachines.

    Another area where we could use more facts is the meaning of the word “nanotechnology.” It has had two, very different, mainstream usages: 1) From about 1986 to 2004, engineered nanoscale machines including manufacturing systems and the technologies based on them; 2) From about 1999 to present, a wide range of nanoscale technologies that do not include nanoscale manufacturing systems.

    The laws of chemistry, physics, and mechanics that Drexler used have not changed. His work is still as valid as Babbage’s was. Babbage may have been too far ahead of his time even to be useful for futurism, and the same might be true of Drexler (but I doubt it). But to the extent we care at all, can we please at least take care to be accurate and argue from facts and science rather than careless impressions?

    • Loriot says:

      I am not an expert on the field, but I recall reading once a post by someone who is who basically said that all of Drexler’s nanotech stuff was fanciful nonsense that ignored physics. So I wouldn’t say it was ever valid in the first place.

      • Le Maistre Chat says:

        I feel like there’s a joke to be made about the famous scientist Clyde Drexler, but science basketball is only a Globetrotters thing.

      • Did he make the same claim about Feynman? Drexler was inspired by Feynman’s talk, which explicitly claimed that it did not violate any physical laws.

      • Chris Phoenix says:

        “I read a post once by someone with credentials” is not a strong basis for a belief. Show me that post and I will criticize it.

        Years ago I conversed with the head of the National Materials Advisory Board about molecular manufacturing. I was there to support their Congressionally mandated inquiry into “molecular self-assembly.” He was extremely intelligent, perceptive, knowledgeable in the field, and appropriately skeptical. He asked about a dozen good questions about the physics and mechanics. I had no trouble answering them on the fly. He then acknowledged that he couldn’t find any reason Drexler’s theories wouldn’t work.

        Over several years I talked with dozens of physicists and chemists. Most of them were skeptical but willing to actually think about the science. None of them found a showstopper. Many of them left conversations with me less skeptical than they had been.

        I am not a physicist or a chemist. I’m actually a software engineer by training. But I learned enough about physical laws at the nanoscale that I could show actual physicists and chemists how molecular manufacturing could be consistent with what they knew.

        This is not strong evidence that Drexler’s theories would work. But it is strong evidence that there are no shallow reasons why they don’t work. Drexler’s designs depend on a set of simple explanations suitable for a bright ten-year-old – which are correct enough to be accepted by, and change the minds of, many experts in their own fields.

        So, let me put it more strongly. I have seen many shallow and wrong explanations of why Drexler’s designs can’t work. Show me that post and I will rip it to shreds in detail from first principles.

    • Matthias says:

      Biology essential runs on nanotechnology. Anything bigger than a single cell is a pretty recent innovation in nature.

      Biology can make all kind sir surfaces, including non-sticky. So why shouldn’t we be able to do the same? (At least in theory.)

    • Elephant says:

      I don’t understand what sort of “consensus” you want, or why this has anything to do with “rationalists.” Many thousands of physicists, chemists, and biologists work on nanometer-scale machines. Every protein, and nearly every molecule, is a nanostructure. We know quite a lot about how to work with them, and we’re doing a pretty steady job of expanding this knowledge. People like Drexler don’t add anything to this. Their pronouncements are trivial if you know that things like “biology” exist, or ridiculous if you care about realistic machineries.

      And why do we need more facts about the “meaning” of the word nanotechnology? Sorry if this sounds snarky, but what does it matter?

      • People like Drexler don’t add anything to this. Their pronouncements are trivial if you know that things like “biology” exist, or ridiculous if you care about realistic machineries.

        Drexler et. al. know that biology exists, routinely offer it as evidence on the possibility of nanotech machinery. But living things have explored only a small part of the relevant design space, a point that has been discussed extensively in the literature.

        Would you similarly argue that there was no point in humans talking about, and trying to design, heavier than air flying machines, since birds had already solved that problem?

        • Elephant says:

          My argument isn’t that biology has already solved the puzzles of nanotechnology therefore we shouldn’t pursue it, but rather that people who study how biology does this, how molecules do it, and how physics works in these systems are the people pushing the field forward. As one of many examples, here’s a very recent paper:

          I’ll admit that it’s been a long time since I read Drexler, so I am perhaps being uncharitable, but I honestly don’t see what he adds to this. Working with your analogy: of course people should have worked on non-biological flight, but the visionaries to listen to would have been the Bernouillis and Montgolfiers and Wrights.

          • Chris Phoenix says:

            What Drexler adds is a new corner of the design space based on new materials and techniques. It’s like the difference between what the Romans did with stone and concrete – which is very impressive – and what modern architects do with steel and ferrocrete.

            If you wanted to span the Golden Gate with a stone arch, it would be pretty easy – and correct – to say “That’s not remotely practical.” But with bridge wire for suspension cables, it’s eminently practical – and what’s more, it’s worth doing.

            The Romans pushed the field forward quite far. Their aqueducts and buildings are still marvelous to us today. But they could never have built the Golden Gate Bridge.

            The paper you link is about self-assembly. Self-assembly is comparable to piling stones on each other and letting gravity hold them in place. Piling stones is convenient, it works, you can build a lot with it – even arches! – but you can’t build a truss.

            The structure of argument from biologists usually boils down to “What we’re doing has lots of potential, so there’s no reason to look elsewhere. Also, our stuff is so cool that anything promising higher performance is impossible.”

            The difference between technologies – what can be built in the different corners of the design space – is vast. To someone who believes Drexler’s corner is reachable, today’s technological achievements look pretty quaint – and, sorry, but biology looks like a dancing bear, amazing it works at all but not what you’d take to a figure-skating competition, any more than you’d harness eagles to pull your airplane. As Harry Potter said in HPMOR, “We also have lawyers in Muggle Britain, and they’d think your lawyers are cute.” Or Heinlein’s example – a medieval time-traveler showing up at Du Pont and saying, “Sirrah, I am an alchemyste. Hast need of art such as mine?”

            Biology is valuable today, and on the medical front it will stay valuable. But don’t tell me about biological machines and self-assembly. I know about them. When it comes to computation, power conversion, manufacturing, and material strength, they’re not within orders of magnitude of what’s possible.

            If we care about what’s possible long-term – and as rationalists and altruists we should – then we need to study Drexler’s work, not get distracted by biology and self-assembly.

            EDIT TO ADD: As a concrete example, biology is full of entropic springs and rarely uses bond-strain springs. Entropic springs are temperature-dependent and require lots of relaxation time. Entropic springs can be built with unstressed linear-polymer chemistry and tuned via evolutionary selection of the monomers. Bond-strain springs simply work better and are easier to engineer.

            A second example: Sliding surfaces. Biology has sliding surfaces, but again they require time to relax and conform. Stiffer surfaces can slide faster at the same efficiency – zero static friction and low dynamic friction. Low friction between stiff surfaces (despite atomic “bumps”) has been demonstrated in the lab between two nested nanotubes.

            A third example: impermeable membranes. A single sheet of graphene should be impervious to helium atoms at room temperature. Biology has nothing that can match this.

            A fourth: Bacteria can self-replicate in as little as 15 minutes, but macro-scale organisms take months to years. There would be a strong selection pressure to doing it faster, but biology demonstrably can’t. Molecular manufacturing claims that a kilogram-scale factory, given suitable blueprints, can build a duplicate factory from small molecules in less than an hour.

            I could continue for many pages.

      • Chris Phoenix says:

        In the thread a few days ago (on the masks post) a lack of common understanding of that word led to bad argument structure. It’s an important word with two very different meanings, and we should either abandon it or learn both meanings and know which one we’re using.

        The consensus I want is on whether Drexler’s corner of the design space – reliable, relatively rigid, mechanically engineered, non-self-repairing machines that close the manufacturing loop – is physically possible and has anywhere near the performance he calculated.

        If Drexler was right, then there is an important set of physical technology which we will have to deal with at some point. What year? 2040 or 2140 or 2400? I can’t predict – but 2040 seems plausible, and as long as the answer isn’t “never” then the effective altruists and existential risk people both should pay attention to this possibility.

        The reason I’m approaching rationalists about this is that, in theory, we form beliefs on evidence, we are willing to adjust our beliefs based on true information, and we are willing to believe things that are unintuitive. I freely admit that Drexler’s conclusions are unintuitive to many scientists. I’ve talked with dozens of those scientists, and in many cases I pointed out things that changed their minds about whether his designs could work.

        I believe that if we looked at molecular manufacturing rationally, we would conclude that no known facts demonstrate its impossibility, and best available calculations indicate that it might actually work more or less as described. That is the consensus I want – either that, or some actual fact-based criticism that points out a showstopper. But I spent years looking for that criticism, speaking at scientific conferences on four continents and engaging with anyone who wanted – and I never found it.

  16. C_B says:

    SafetyWing straight up a Terra Ignota hive.

    • Aotho says:

      Hah, my thought too!: “So it begins…”

      That means I can get a near-super-sonic flying car soon too, please?

  17. VoiceOfTheVoid says:


    Testing the notification system, does this show up in my email inbox?

    • VoiceOfTheVoid says:

      Seems it does, wonderful!

    • Aotho says:

      I also want to try the notification system that you recommended me, even though I didn’t get any confirmation emails for about a day now; maybe it will work all the same?


    • Aotho says:


      The first attempt didn’t succeed, trying again.

      • VoiceOfTheVoid says:

        Did you replace “YourUsernameHere” in the URL with your username and refresh the page before you entered your email and submitted? I have no idea why it’s set up like that, they really should just have a text box for the username, but that’s how it is.

        For your convenience:

        • Aotho says:

          Yup, I’ve done that even the first time, it’s just that the subscription confirmation email never arrived, neither for @Aotho nor @YourUsernameHere.

          But I tried just recently again and it works now! I’m looking forward to what promises to be a much more delightful user experience when it comes to keeping track of replies.

          Thanks for suggesting this service!

  18. A Definite Beta Guy says:

    Lawn Care waits for no Coronavirus!

    Last year ended on a sour note for my lawn. None of my seed germinated (Kentucky Bluegrass can be pretty hard to get going), Creeping Charlie reared its ugly head again, and a bunch of dead spots popped up near the end of summer (drought stress).

    This year I am burning through my remaining supply of fertilizers and weed killers, because after that we’re switching to a more careful approach. Actually, we’re starting now:
    -Soil tests. Not mailed to University of Illinois (I kind of sat on my butt too long for a mailed out soil analysis), but an at-home one. We’re nitrogen-starved but other than that, good to go.
    -Measured out fertilizer app. Related to above point, my spreader has NOT been applying the correct amounts, so I’ve been feeding about half the recommended dose of nitrogen. Well, no wonder it’s not growing great.
    -Soil temp monitoring. We were expecting a warmer weekend in Chicago, so I threw down crabgrass pre-emergent on Friday. Turns out our soil temp is still averaging under 50…but it’ll probably go up this week.

    We’ll see how this year goes. I have a general plan to seed a few bare spots in the lawn in 30 days (this time with some extra soil and better starter fert), and will spray for that damned Creeping Charlie when it starts blooming in a few weeks. Come fall, I am going to another an overseeding with KBG to thicken up the lawn. The front is a bit acidic, so I’ll be working in soil amendments throughout the year to get it to base.

    Only other change this year is that we dug up an unused portion of the lawn to make a beet and asparagus garden. It’ll be a few years before we see returns on the asparagus, but I’m really hoping this pans out. We eat a ton of it and the harvest season is pretty long on them.

    • Lambert says:

      If N remains a problem, try growing a bunch of clover?
      It’s leguminous so it fixes nitrogen.
      You might want to check NH4+ vs NO3- levels in the fertiliser too.

      • FLWAB says:

        You ever try to get clover out of a lawn? It’s a pain. They tend to out-compete the grass if they make it at all.

        • AG says:

          Is this an issue? Is a clover-lawn not as usable as a grass lawn?

          • EchoChaos says:

            I grew up with clover lawns and love them, but they are not as “perfect looking” for the American suburban experience as a grass, so if that’s what you’re going for, they’re less desirable.

          • Lambert says:

            Opinion: lawns are overrated and should not be the default garden usage.

            Of course if they have a particular purpose such as if you have kids or a dog (especially if there’s no playing fields or commons nearby), and a bit of grass cover is nice.

  19. myst_05 says:

    Any information on why exactly it would take us 12-18 months to build a vaccine? Are there any attempts to circumvent standard regulatory protocols to rush a working vaccine in a shorter period of time? E.g. immediately inject vaccine candidates to 10,000 people at once instead of slowly expanding your trial?

    • Nornagest says:

      As far as I can tell, the 12-18 months figure is for vaccine development that technically ticks all the procedural requirements boxes, but ticks them at what passes for breakneck speed in the world of bureaucratic box-ticking. We could probably get numbers on efficacy much faster if we were willing to accept some level of risk.

      Not accepting some level of risk is, of course, insane under the present circumstances, but I don’t know if it’s insane enough to overcome institutional inertia. On the other hand, I’d be astonished if “complete lockdown will continue until we have a vaccine” and “the vaccine will take 12-18 months” both end up being true — one way or another.

    • johan_larson says:

      I had a chance to discuss this with my brother, who works in biotech venture capital. It’s really important to make sure that a vaccine doesn’t hurt people, since you’ll be administering it extremely widely, and if it is dangerous in even a small portion of cases, you’ll have done more harm than good. And of course you’ll have to verify that the vaccine is actually effective in the first place; not all vaccines offer equal protection.

      Pharmaceutical companies apparently hate working on vaccines, since they typically can’t charge very much for them, for PR reasons. For them, the ideal condition to treat is something people really want fixed, but don’t really need fixed. Boner pills are pretty much perfect.

      • Evan Þ says:

        My dad is a biostatistician who analyzes drug studies, and based on talking with him, I endorse this statement – it’s true about all drugs, but especially about vaccines since you’re giving them to otherwise-healthy people.

        • Well damn, that makes an unwelcome amount of sense.

          Not hard to imagine scenarios in which rushing a COVID-19 vaccine to broad usage ends up becoming the anti-vaxxers’ fever dreams come true. Which would in the big picture be very, very bad.

      • Rick Jones says:

        One thing to worry about with vaccines is ADE, where the antibody actually helps the virus become more virulent upon exposure. Happens in SARS classic and MERS vaccine candidates that target the spike protein.

    • John Schilling says:

      E.g. immediately inject vaccine candidates to 10,000 people at once instead of slowly expanding your trial?

      This would be very illegal, and very hard to hide, so I’m pretty sure nobody is doing it. The FDA could make it legal, but they won’t. Congress and the President could make it legal, but they’d have to work together to do so and if they didn’t have the FDA on their side it would either be legal but so bound in red tape as to be useless, or legal in a hamfisted way that constitutes blanket approval for anything advertised as “vaccine testing” even if done to the standards of the old Tuskegee institute and with vaccines that are seriously dangerous.

      • myst_05 says:

        But what’s wrong with getting 10,000 brave volunteers and asking them to risk their lives for the sake of the country? Wouldn’t it be equivalent to US soldiers being drafted to go to war overseas?

        • John Schilling says:

          I didn’t say there was anything wrong with it. I said it was highly illegal. Whether or not you see a connection between those two things is up to you, but “it’s not morally wrong!” is pretty much ineffective as a legal defense.

      • Evan Þ says:

        A third option: The President could preemptively promise pardons to anyone and everyone for all crimes associated with COVID vaccine testing that meets standards XYZ.

        I do not recommend this idea, but it would be better than the current situation where we ruin the economy since no one’s able to take risks with vaccine tests.

        • Loriot says:

          Now that would be a shitshow.

          Congratulations. You managed to come up with an idea even worse than encouraging people to deliberately infect themselves.

        • John Schilling says:

          The pardons don’t help with the part where you wind up professionally blacklisted because the FDA and everyone in their orbit knows what you did, and will remember it long after most of the world has forgotten who did what during the great coronavirus scare.

          The promised pardons also don’t help if the FDA waits until after Trump has left office to charge people with specific offenses.

          And if the plan is for Trump to actually issue preemptive blanket pardons to everyone who writes him saying “I did me some COVID-19 vaccine testing and I’m afraid I might be charged with a crime someday”, then an awful lot of truly atrocious things are going to be done under that umbrella.

          If the plan is that Trump will issue the preemptive blanket pardons only to people who have responsibly conducted accelerated vaccine testing, then Donald Trump doesn’t know how to do that, Jared Kushner doesn’t know how to do that, nobody Donald Trump trusts knows how to do that, and about the only people who could realistically do that are working for or closely affiliated with the FDA and CDC.

        • The Nybbler says:

          In addition to all the other problems, pardons cannot be issued for crimes not yet committed. The President can issue a pardon before conviction, trial, indictment, or even accusation, but not before the crime is committed.

          Nope. We’ve gotten ourselves into this place, where everything’s screwed down tight because the up-sides of regulating everything that way were legible (no thalidomide babies, yay) and the downsides were not, and anyway objections could be dismissed by regarding the objectors as lazy, reckless monsters. Now that the down-sides are finally legible, it’s far, far, too late.

    • Cheese says:

      Several things really (not an immunologist).

      You have to be sure it works, which basically means building a robust animal model with similar clinical characteristics to the human induced disease.

      You have to be sure of your side effect profile given this is going into otherwise healthy people.

      You have to be sure you aren’t making it worse, as has happened with similar vaccine candidates in the past:

      Therapy responses in vivo are fundamentally unpredictable. You can have a great candidate in vitro or in an animal model that totally fails to do anything in vivo or in human disease, and it takes you 5 years to find out why that was to inform refinement.

    • ana53294 says:

      Marginalrevolution linked an interesting essay that discusses, among other things, why it seems to be hard to develop a vaccine.

    • Garrett says:

      An interview with someone who’s gone through the vaccine development process. It includes explained technical details.

      TL;DR: The big concern isn’t that the vaccine might not work, or that vaccine administration might be dangerous, but that a proposed vaccine might make future infections worse for the patients.

      • myst_05 says:

        Couldn’t it be tested within two weeks by exposing vaccinated trial patients to the virus?

  20. Rick Jones says:

    I have a bleg. Can anyone direct me to information about specifically what went wrong with the initial CDC/FDA SARS covid 2 tests. I mean something more than “there were errors in manufacturing the reagent”. Which reagents? Were the primers poorly designed? Why were the rna extraction kits in short supply.
    I’m reasonably literate in lab procedures and it’s driving me crazy because everyone views it through their political filter ( the right; it was evil bureaucrats protecting their turf/the left; noble public servants who didn’t want to anger Trump). Knowing the specifics would be nice.

    • Silverlock says:

      I cannot answer your question, but I have a request of my own: if/when you find the answer, please write it up on SSC so we can all benefit. I am not a lab kinda guy, but I am looking for any pieces of knowledge that are at least somewhat filter-free.

    • HeelBearCub says:

      I saw this article linked elsewhere the other day.

      Before a state or local lab could use the C.D.C.-developed tests on actual patients, however, it had to insure that they worked the same way they had in Atlanta, a process known as verification. The first batch of kits, sent to more than fifty state and local public-health labs⁠, arrived on February 7th. Of the labs that received tests, around six to eight were able to verify that they worked as intended. But a larger number, about thirty-six of them, received inconclusive⁠ results from one of the reagents. Another five, including the New York City and New York State labs, had problems with two reagents. On February 8th, several labs reported their problems to the C.D.C. In a briefing a few days later, Nancy Messonnier, the director of the National Center for Immunization and Respiratory Diseases, said that although “we hoped that everything would go smoothly as we rushed through this,” the verification problems were “part of the normal procedures⁠.” In the meantime, she said, until new reagents could be manufactured, all covid-19 testing in the United States would continue to take place exclusively at the C.D.C⁠.

      The covid-19 tests use polymerase chain reaction, or PCR, a technology for whose invention the biochemist Kary Mullis won the Nobel Prize, in 1993. PCR is highly sensitive to contamination and other faults, which is why the verification step is necessary to insure accurate results. And yet while the reagent problems were, in their way, a fairly ordinary technical hiccup—Messonnier, at the C.D.C., was not spinning the situation—the cascading effects that they’ve had on the country’s covid-19 preparations suggest a much larger problem with the way the United States has structured its pandemic response. That problem was exacerbated by a President who has simultaneously underplayed the severity of the outbreak and overpromised the means available to fight it.

      So, that makes it look like a normal thing that happens, it just happened at the worst possible time. Not sure if that resonates with you.

      The next paragraph shows that part of the issue is just how fast we are actually going. From trying to find the genome to test in under a month:

      he problems with covid-19 testing in the United States have obscured what was, near the start of the pandemic, a triumph of modern medical science. On January 10th, three days after Chinese government officials announced that a novel coronavirus, now known as sars-CoV-2, was responsible for an upsurge of pneumonia cases in Wuhan, China, a team of Chinese scientists uploaded a copy of the virus’s genome to an online repository⁠, and virologists around the world set to work to develop diagnostic tests for the new disease. On January 21st, a team in Berlin, led by Christian Drosten, one of the scientists who discovered the original sars virus, in 2003, submitted the first paper to describe a protocol for testing for sars-CoV-2. (That protocol would form the basis for a test disseminated, early on, by the World Health Organization.)⁠ That same day, Messonnier announced that the C.D.C. had finalized its own test⁠, which it used to confirm the first known case of covid-19 in the U.S.

      • Rick Jones says:

        Thanks. I did see that New Yorker article but was looking for something more specific. I do like the acknowledgment of what amazing tools we have now. The virus genome was known, and published very quickly.

    • zardoz says:

      The first question is why the CDC had to develop its own test at all, given that the WHO had already started distributing a German test to many countries. As far as I can tell, nobody from the CDC has given an answer to that question (does anyone from SSC have more information?)

      The second question is why we forbade anyone but the CDC from working on a test. There are lots of private companies and universities that could have produced a working test (and at least one lab that actually did produce a working one in mid-February).

      Then there was the CDC manufacturing problem in February. According to the WaPo, “the CDC has yet to fully explain the nature of the manufacturing problem.” Rumor has it that one of the reagents in the test kits was contaminated. Someone on SSC was also saying that that at least one of the specified reagents in the CDC test was not widely available, but I don’t have a good source for that.

      • albatross11 says:

        Several independent labs put together their own test, including the lady in Washington who discovered community spread by violating about eight layers of dumb rules.

      • Christophe Biocca says:

        The second question is why we forbade anyone but the CDC from working on a test. There are lots of private companies and universities that could have produced a working test (and at least one lab that actually did produce a working one in mid-February).

        That one’s easyish: The declaration of emergency that was a part of allowing the CDC test to roll out automatically blocked the deployment of laboratory-developed tests (which under normal circumstances would have been allowed). Now why they took so long to also start granting EUAs for the commercially developed tests, inertia?

      • salvorhardin says:

        I’ve read (can’t remember where unfortunately) that the CDC developing its own tests separate from the WHO has been standard operating procedure for new diseases for decades. The general assumption has been that the CDC are best-in-class experts who will reliably and quickly come up with good tests for the US on their own, so the WHO doesn’t need to spend any time/resources making sure their stuff is suitable for US use. That assumption was true for prior diseases, not for this one, and people didn’t pivot fast enough when they realized (or should have realized) that it wasn’t true.

        • EchoChaos says:

          And also note that the WHO doesn’t actually provide test kits, because most test reagents are easier to produce in country.

          People who ordered test kits directly from China had very poor outcomes of the test kits.

      • Rick Jones says:

        The only concrete, albeit anecdotal, information I’ve found is from a provincial lab manager in Canada (Saskatchewan?) who wrote into twiv and who got some of the original CDC kits. She thinks the primers were contaminated, either with viral rna/dna or with other primers. This suggests a physical/blending error and not a design problem with the assay itself. It was quickly clear they didn’t work. She also expressed shock that non-CDC labs (let’s stop calling them private labs, mostly are affiliated with states or uni) weren’t authorized to develop their own tests, as they did quickly.
        These things can happen. Tyler Cohen has been on a outrage party about this, largely because it fits his priors narrative. But of course he can’t be asked to be held accountable for his lack of foresight about the GFC because “who knew”. One of luxuries of being in a soft science where you’re not held to any reproducibility standard.

    • J Mann says:

      The other question is why the bureaucracy took so long to rip off the bandaid and deregulate testing. This may be harder to answer, but did CDC and FDA scientists actually believe they could get on top of this testing problem (and COVID) right up until Mulvaney’s office called them out, or were they requesting liberalization and someone upstream didn’t reach quickly enough?

      • The Nybbler says:

        The other question is why the bureaucracy took so long to rip off the bandaid and deregulate testing.

        They’ve still only partially done so. And the answer is the stinger of the Scorpion and the Frog: it is their nature.

        • DeWitt says:

          Ooooooh sinister.

          Or, you know, you could try asking the organisations involved. Has anyone tried that yet?

    • actinide meta says:

      I have not seen a credible explanation of what actually went wrong, beyond the statement that the third primer/probe set failed to validate (and after a few more weeks of hand wringing the FDA allowed testing with two).

      This post claims that the CDC protocol fails basic in silico tests. I don’t know how credible these claims ate, how serious the problems found by the software are, or whether they have anything to do with the actual validation failures of the CDC reagent packages.

      • Rick Jones says:

        Thanks for that link. From this it sounds like the primers looped but note that he didn’t actually make his primers physically, just validated them on a computer. But if the actual problems seen in the lab show the same behavior then he’s on to something. At any rate, the idea of designing primers on a computer to look for these problems seems like a good idea.

  21. Chalid says:

    What changes in a few months that lets us safely get out of lockdown? Some possibilities, in no particular order:

    Scenario #1: testing is massively ramped up, and we can test-and-trace-and-quarantine well enough to keep the epidemic under control. This is what “the smart people” are arguing for. The testing part will get done eventually, one expects, though recent numbers have been disappointing. But the “trace and quarantine” part assumes a high level of competence on the part of the governments which failed to suppress the epidemic in the first place. In the US at least, I’ve seen no evidence that anyone in the federal government is actually working on building out this sort of program, and I don’t really see how this can work on the state or local level.

    Scenario #2: one or more of the various treatments that is being trialed turns out to be effective in reducing death rates. The lockdown buys us time to get some proper trials done and to scale up production of whatever is successful. It seems very likely that some drug or technique (or combinations thereof) will meet with *some* success, but I think we’d need a very big reduction in deaths (factor of 10?) before it was generally seen as acceptable to let the virus run wild.

    Scenario #3: we figure out that the virus has all along been less deadly that we thought, and there have been many many unobserved or asymptomatic cases all along – e.g. if there are 10x as many people who have been infected as we thought, then the virus is 10x less deadly. This doesn’t seem that likely to me, but we will probably know for sure in a couple weeks as several places are going to do some random-ish testing of their population.

    Scenario #4: we look at variation in country lockdown intensity, or variation within countries, and learn that a total shutdown doesn’t lead to much additional benefit over relatively mild measures like mandating mask-wearing and limiting large gatherings. Ideally we learn enough to point to specific measures that were helpful/unhelpful. And so we gradually relax the most onerous of the restrictions and things mostly return to normal for anyone under 70 who doesn’t go to church or commute on a subway.

    Scenario #5: variolation, or just letting less-vulnerable populations out of lockdown to get infected, or other herd immunity strategies. Doesn’t seem likely now but if the other scenarios aren’t working out after two months of lockdown we might be willing to accept these kinds of measures.

    Anything else?

    None of these seem very likely to me. But none seems fantastically unlikely either, and collectively the chance of at least one of these working out seems pretty good. And this is a case where partial credit counts – e.g. a combination of e.g. moderately successful treatments and the virus being somewhat less deadly than currently thought could lead to it being acceptable to leave lockdown.

    So I think I’m rather optimistic about most of us being able to safely move in the direction of normality this summer. Meanwhile, maintaining lockdowns while we gather information, produce supplies, etc. seems entirely reasonable.

    • johan_larson says:

      If we could isolate ourselves perfectly, the virus would die out within our borders within a few weeks. The hope is that we can almost make that happen by almost isolating ourselves for a slightly longer period. We won’t be able to kill the virus completely with imperfect quarantine, but the hope is that we’ll drive the number of carriers down so far that we can relax the quarantine measures, and contain any flare-ups that happen after that using boring old testing and contact tracing.

      It seems plausible, and we certainly won’t drop the quarantine measures all at once. And of course we’ll be carefully on the alert for any signs of flare-ups as we relax the quarantine gradually.

    • Matt M says:

      IMO, lockdowns will end when there is political pressure for them to end. It will have little to nothing to do with medical outcomes and everything to do with when exactly large segments of the public get sick and tired of this and start thinking the economic costs aren’t worth it.

      • John Schilling says:

        It won’t be completely decoupled from medical outcomes, because the politics for lifting the lockdown will be much more favorable if there’s some good news to point to, but basically this. A declining trend in new infections under lockdown will, if people are sufficiently sick of the lockdown, be taken as proof that the lockdown has “succeeded” and can be safely ended. Understanding that this won’t work, requires math and therefore will not be a winning proposition for anyone who needs to be reelected in November.

        • Matt M says:

          Every day for the past month there has been really good news and positive developments on COVID that are available to report.

          And every day for the past month there has been really bad news and negative developments on COVID that are available to report.

          Which get reported, and which people end up seeing, believing, and promoting, depends entirely on the motivations of the political-journalist complex.

      • cuke says:

        It seems like the feedback loop on this would be pretty fast though, right? If too many restrictions are opened back up and there are big spikes in cases and deaths, localities are likely to respond again pretty quickly.

        • John Schilling says:

          Or they’ll just decide (under pressure from a frustrated community) that this “proves” the lockdown didn’t work in the first place or that it worked only by delaying the inevitable.

          Politicians don’t like admitting they made a mistake, and they’re not going to compound that by admitting that admitting they made a mistake, was a mistake. The more mistakes they make, the more firmly they will insist that the latest decision was the right one.

          • Clutzy says:

            I mean, it would have proved the lockdown didn’t work. Everytime I go outside I see evidence that the lockdown probably isn’t going to work. The 3 times I’ve had to go to the store had massive evidence the lockdown wouldn’t work.

            My frustration level with the local populace is only equaled with my frustration with the “lockdown experts” that haven’t really explained what exactly the lockdown will do if people are still allowed to do grocery-related things. Won’t people just get infected at the grocery store (no employees ever have masks)? Won’t that keep the infection alive for whatever duration the lockdown is in asymptomatic populations only for it to explode again whenever lockdown is lifted?

          • Chalid says:

            Sure, there are still some contacts from people at the grocery store, or people walking too close on the street or whatever. But we don’t need literally zero new infections for the lockdowns to be effective.

            Under normal circumstances each carrier affects 2-3 other people. If people are on average having 1/5 the contacts that they usually do then each infected person will infect 0.4-0.6 other people on average and the infection rate will exponentially decay.

            Empirically, you can see that infection and death rates stop rising and eventually start to fall in locked down areas.

          • Byrel Mitchell says:

            @Clutzy At my local (small town) Walmart they’re only allowing one person and one cart per family. And on top of that, they’ve hired a ton of people.

            They have 2 people at the single open door, only allowing people in with a sufficient gap behind them and when there’s a cart ready for them.

            They have someone inside the door sanitizing carts.

            They have 1-2 people in each aisle, watching what things you touch and immediately sanitizing them (and wiping down shelves and handles in dead time.)

            They have multiple people monitoring the self checkouts, and they swoop in and wipe down the entire station between customers.

            They’re in the process of putting up spit shields at the pharmacy as well, but that’s not finished.

            I’m seriously impressed; I had expected that the lockdown would be mostly negated by grocery shopping, but if grocery stores take that level of precaution… I’m optimistic.

          • albatross11 says:


            You were formerly getting close enough to 100 people a week that you could catch the virus, now you’re getting close enough to catch it from 10 people a week. This is a big improvement, even though people still can catch it in the grocery store, or walking in the park and passing close to someone else, or whatever.


            Imagine a really simple model. Everyone in the population is either susceptible, infected, or recovered and immune. To predict how much spread you have, you want to know how many susceptible people each infected person will have contact with, and the probability that each contact will lead to a transmission.

            Lockdowns are about decreasing the number of people each infected person has contact with. If you don’t go to work or school, don’t ride a crowded bus or subway, don’t go to church or a bar, then the number of people you *could* give the virus to is a lot lower. Even though the current lockdowns are leaky and imperfect, it seems like they have to be massively lowering this number.

            Similarly, stuff like keeping big public events closed, widespread home delivery or curbside pickup for groceries, or switching all restaurants to take-out-only decreases the average number of susceptible people who each infected person *could* infect. And stuff like fever checks or sending home sick people from work/school decreases that number, by ensuring that some fraction of infected people get way fewer chances to infect others.

            Other social distancing measures (masks, hand sanitizer, keeping 2m apart) are about decreasing the probability of transmission given that we have contact. When everyone’s wearing a mask and hand-sanitizing after going anywhere in public, probably the probability that transmission will happen given that you are exposed to an infected person will go down quite a bit.

            You care about both these numbers, and there are better and worse ways to get them. My amateur guess right now is that community spread is a really small part of current spread of the virus–most of it is probably in hospitals and doctors’ offices and ambulances. Which makes the shortage of PPE really worrying.

          • Matt M says:

            Or they’ll just decide (under pressure from a frustrated community) that this “proves” the lockdown didn’t work in the first place or that it worked only by delaying the inevitable.

            This is my impression of what happened in China, BTW.

            They made an honest effort at authoritarian lockdown of Wuhan specifically. Maybe it worked, maybe it didn’t. But it gave them enough info to decide that sort of thing wasn’t feasible to do to the entire country for an indefinite period of time.

            Whether you call that the lockdown “not working” or the lockdown being “not worth it” probably doesn’t matter – the decision was made to go in a herd immunity direction.

          • Clutzy says:

            You were formerly getting close enough to 100 people a week that you could catch the virus, now you’re getting close enough to catch it from 10 people a week. This is a big improvement, even though people still can catch it in the grocery store, or walking in the park and passing close to someone else, or whatever.

            Again. I understand this. It still is useless unless there are a huge number of undiagnosed asymptomatics. Otherwise we just get the spike in June or July or August when we un-shelter because there will still be plenty of virus living.

            There have been investigations into the Imperial College Model which is used to justify shelter in place orders. Those show a massive spike in October, resulting in still something like 180 million dead worldwide.

          • LesHapablap says:

            Matt M,

            What makes you think China has gone with a herd immunity strategy? They are claiming they’ve had ~20 deaths in the last week, a herd immunity strategy would give them something like 5,000-100,000 a day. That would be one hell of a cover up.

          • Matt M says:

            I believe both that the disease isn’t nearly as bad as we are being led to believe, and that China is covering up a lot of deaths, yes.

          • LesHapablap says:

            @Matt M,

            Ah ok, the way you wrote that it sounded like it was common knowledge that China had gone to a herd immunity strategy.

            I don’t think it is plausible that China has uncontained spread, but if you have some evidence to the contrary I’d be glad to hear it.

        • Alyosha says:

          Hospitalization lags the point of infection by a week or more. You wouldn’t see a spike for a couple of weeks, and then the spike would continue for a couple of weeks after you shut everything back down as it is now.

      • Theodoric says:

        Possible signs we might be getting there soon:
        Gannett is cutting pay and furloughing people.
        Biglaw is is doing layoffs and salary cuts.
        Disney executives are having their pay cut, and I’m sure this will spread to other companies whose employees thought they were safe working from home.
        The economic effects of this are beginning to affect Important People, people it will be difficult to portray as a bunch of losers who should just #learntocode.

        • Matt M says:

          Agree with this too. Once the blue-check crowd starts losing their jobs, the media narrative will start to change, and public opinion along with it.

          I predict the longest this can possibly last is until the next school year. Public school teachers are entirely too sympathetic to face layoffs without massive public outrage.

    • LesHapablap says:

      By mid-May it will be clear that herd immunity has taken hold in Sweden and Italy (or it won’t). By that time, widespread antibody testing will give us much better stats on fatality rates. If it is politically feasible then I believe at that point countries will switch to a herd immunity strategy that doesn’t destroy the economy. Sweden will look like the cooler heads that didn’t panic.

      Already there are calls from economists and epidemiologists in mainstream online news sites to make sure the cure is not worse than the disease. By mid-May, when the layoffs just keep getting worse, there will be enormous political pressure to let people get back to work. At least, this is what I’m hoping for since otherwise my company will go under and all that I’ve worked to create over the last 9 years will be wiped out.

      For reference, here is the newer imperial college study that claims that 5% of Europe is already infected at March 28th, as much as 33% in Italy:

    • albatross11 says:

      The modeling papers I’ve seen (one example is here (pdf)) suggest that a hard lockdown followed by a complete opening up is very likely to just lead to a resurgence of the disease. Basically, if we locked down super hard, we could in principle kill the thing off by giving it no hosts to live in. But nobody can manage that (and if we did, we’d be reinfected by someone who was less careful), so we lock down hard enough to slow down the spread of the virus without eliminating it. Very little spread = very few people gaining immunity = very fast spread when we open things up again.

      The danger here is that we oscillate–we have a painful month-long lockdown, numbers of infections drop, and then we open things back up, and a month or so later we’ve got another exponential curve in infections and we’re once again scrambling to find enough ventilators and N95 masks. Probably we go back into lockdown in panic mode again.

      The ideal answer here is to get the number of infections under control, get the exponential curve to flatten out, and then eventually relax the lockdowns when we have extensive testing and tracing of sick people. But let’s be honest among ourselves–that might work in some places, but there are a lot of parts of the US where the local government is a dysfunctional mess. I’m not super optimistic about how well this will be done in those places. It *could* and *should* be done everywhere, and it will be done in *some* places, but the places that can barely keep the schools halfway functioning or the streets safe in broad daylight are probably going to do a shitty job here.

      So I thnk w really need to be working out intermediate measures that *are* sustainable and that can lower R_0 below 2–stuff we can do for a long time, maybe indefinitely. If we get R_0 down around 1.5 or so, then flare-ups after we relax the lockdowns will tend to die out on their own. If we have measures we can keep using for a long time, we can probably avoid the disaster of an alternating cycle of lockdowns and hospital meltdowns and new lockdowns. And that can work alongside tracing infections and quarantining people.

      Think of this like treading water–it has to be stuff we could do for a long time without totally tanking our economy, and without being so unpopular that they get swept aside at the next election.

      To specify measures that are affordable and effective at lowering R_0, I think we need to get some sense of what fraction of spread is happening from surface contamination vs close contact droplets vs longer-lasting airborne droplets. We also need to get some sense of how contagious (via each route) people are at different stages of infection. Does anyone have a reference for people trying to dig into that?

      One interesting and disturbing possibility is that there’s a small subset of people who show few/no symptoms and are contagious as hell–what were called superspreaders in the original SARS.

      Here’s my amateur first-cut analysis. Please tell me why I’m wrong or suggest things to read:

      a. From everything I’ve read, the spread of this disease is due to droplets. When you cough, sneeze, yell, sing, talk, eat, or even breathe, you’re producing tiny droplets of water/mucus that scatter in a cloud around you. This discusses droplets that spread disease.

      b. I think the critical question is what size of droplets are most important for the infection. If it’s mainly big droplets, then keeping 2m apart and wiping surfaces down with bleach solution a few times a day will be very effective. If it’s smaller droplets, they’ll stay airborne longer. In the worst case, they’re really tiny droplets that dry out and can stay airborne for hours. There’s experimental data that shows COVID-19 surviving several days on surfaces (3 days for stainless steel and plastic, 1 day for cardboard, an hour or two on copper), and also surviving at least three hours in airborne particles. It’s not clear that those particles are important in spreading the disease most of the time, but the virus can survive and still be infectious after three hours airborne.

      c. I think it’s pretty clear that some things (coughing, sneeezing, singing) make and spread a lot of droplets really far; quiet speech and eating and regular breathing probably don’t spread them nearly so far. Intuitively, the more energy you’re putting into something that involves sending air out of your mouth, the more droplets you’re going to make and the further they’re going to travel. The choir group that got such a fast spread is probably a demonstration of this. (Right away, this suggests to me we’re lucky sports and concerts got shut down–yelling and cheering is surely also going to send lots of droplets far away, and you could probably get really rapid spread that way. And also that it’s a good thing that most churches aren’t meeting in person right now.)

      d. If asymptomatic/presymptomatic people were routinely leaving clouds of infection everywhere they went that hung in the air for several hours, I don’t think the measures that have worked passably well in stopping the spread in Singapore, Taiwan, China, and South Korea would have been at all effective. Since other countries have apparently stopped it (however imperfectly), I infer that at least *most* transmission isn’t someone with no apparent symptoms spreading a cloud of sickness everywhere he goes. OTOH, if someone is coughing all the time, I think there’s a good chance they’re leaving a cloud behind.

      What I think that gets us:

      a. Anyone with a fever, or who is coughing/sneezing needs to be away from uninfected people. In particular, they must not be in a closed-in space (office, store, car, school room) with uninfected people, or the number of uninfected people is going to go down.

      b. Closed doors and separate spaces are probably a big win. For bigger droplet transmission, if I close my office door, your droplets aren’t getting in at all. I’m not sure what to do about open offices, warehouses, et.

      c. Getting everyone to wear a mask seems likely to decrease the droplets they produce, especially from normal talking. And that’s cheap–in six months, the whole country is going to be paved in masks. (Probably gray-market knock-off KN95s from shady Chinese sellers.)

      d. Spraying high-touch surfaces with disinfectant every couple hours seems like another win. It’s cheap, and will help if surface contact is a problem. That plus hand sanitizer stations everywhere probably deals with most transmission via contaminated surfaces.

      e. Decreasing the number of people in the office by allowing as much remote work as possible or working people on split shifts so there are fewer people in the hallways at the same time might work. (I share an office with someone–imagine if everyone who shared an office had to split days with their office-mate and keep the door closed.)

      Those are all low-cost things an office could do to decrease R_0. Add in social distancing rules (keep 2m apart as much as possible), this seems at least plausibly like it might prevent rapid spread of COVID-19 in an office. What else might work?

      • salvorhardin says:

        On this note, why on earth aren’t we doing, or even making any public plans to do, temperature screenings at the entrances to major crowded enclosed spaces (stores, transit stations etc)? It seems to be part of the East Asian success package; it could do a fair bit of incremental good even if it only catches a smallish percentage of spreaders (or perhaps even if it just has the second order effect of making symptomatic people more likely to stay home); and it’s not clear that there’s any technical barrier to doing it.

        • albatross11 says:

          Forehead fever thermometers are cheap and readily available. Wipe the surface off with disinfectant between people, and save one of your hard-to-get N95 masks for the guy at the door doing fever checks.

          TSA should have been doing this for the last month, IMO. Putting a symptomatic person on an airplane has a lot of potential for having a lot of new cases show up, and worse, they’re traveling, so they’ll end up all over the country.

          When offices reopen, I’d sure like to see this happening nearly universally.

          • salvorhardin says:

            Just so. Which all makes me wonder why there hasn’t been nearly as much noise around this as around mask-wearing, more widespread testing, or other such reasonable measures.

          • HeelBearCub says:

            Putting a asymptomatic person on an airplane has a lot of potential for having a lot of new cases show up, and worse, they’re traveling, so they’ll end up all over the country.

            I’m not sure if you actually meant asymptomatic here, which is how it scanned at first, or you actually meant someone who is, in fact, symptomatic.

            I guess, how much it actually makes a difference depends a lot on what percentage of infected are actually asymptomatic spreaders, and how they affect R0.

            I certainly agree that we should be detecting fevers. It seems like it should be relatively low cost and efficacious, although one wonders how large the current stock of sensitive enough laser thermometers is. Caveat being that they aren’t actually all that cheap (and I wonder how many readings they are designed to be able to give before breaking).

          • Lambert says:

            Wow the price of IR thermometers has doubled since I bought mine.
            Note that the white and blue (Healthcare-coloured) ones cost a lot more than the yellow and grey (trades coloured) ones.

            And the actual pyrometer seems to just be a lens and a thermopile, so that part is never actually turned off. I don’t think it’s unreasonable to expect one to read the temperature 24/7 from now till the end of the pandemic (assuming you hooked it up to a 9v bench supply).
            Hmmm, I wonder whether you can open one up and plug the temperature output into a microcontroller somehow. Then stick it on an rc plane and do some kind of low-budget geophys survey? Or make a glider that can actively look for thermals.

          • Conrad Honcho says:

            My employer’s doing the temperature check thing, but with the run on forehead meters they could only get rectal thermometers.

          • HeelBearCub says:

            I’m thinking that the industrial, cheap ones aren’t designed to be accurate to the tenth of a degree that matters for measuring body temp.

            The parts that break would be things like the trigger mechanism, the display, etc. Difference lens material would seem like a potential source of degradation as well.

          • JayT says:

            Is a tenth of a degree really all that important for screening? I can’t imagine that they would stop someone that had a 98.8 temperature instead of 98.6. I’d think even the cheap one could tell you if someone was at 100 though.

          • HeelBearCub says:

            I believe that’s the wrong way to think about it.

            If the thermometer is accurate to +-1 degree, and displays in 1 degree increments, and the persons actual temperature is 100.4, the reading might be 99. And it might be 99 or 100 for someone with a 98.6 temperature as well.

            The commercial IR readers care a lot more about the difference between 350F and 400F than the difference between 98.6 and 100.4.

          • albatross11 says:

            There are forehead thermometers you can buy at the drug store which seem accurate, are sold for medical use, and do not cost all that much money. I’m sure there’d be a problem if we suddenly needed a zillion of them, but it’s not like taking someone’s temperature is all that hard a problem. Along with forehead thermometers, there are ear, and temple thermometers, either of which would work passably well for screening people quickly. Or just old fashioned oral thermometers, if you don’t mind slowing things down a bit.

          • HeelBearCub says:

            There is a reason doctor’s offices have throw away covers for their thermometers. I think you really don’t want to purposely be touching the same object to everyone’s face or sticking it in their mouths in the middle of a pandemic if you can possibly avoid it.

          • Conrad Honcho says:

            Nobody liked my rectal thermometer joke? Tough crowd, SSC.

          • Garrett says:

            In EMS we have nursing homes measuring our temperatures when we bring patients in. We’ve had the staff measure EMS provider temperatures as low as 80F (low enough to be dead-dead) while standing there.

            If healthcare facility staff can’t get it right, I have little confidence in forehead thermometers applied by the lay public to be anything other than a random number generator.

        • Forehead fever thermometers are cheap and readily available.

          Can you do an adequate job by measuring infrared at a distance, which would be even easier?

          • Lambert says:

            If you’re talking about the vaguely gun-shaped things, that’s exactly what it does.

            The problem is that far IR optics is hard (glass is opaque to those wavelengths). It’s much easier to hold a cheap lens that focuses on a single thermopile an inch away from someone’s forehead than an expensive germanium lens focusing on thousands of thermopiles a number of feet away. And an IR camera would be rate-limited by the pair of eyeballs looking out for people with hot foreheads.

          • Loriot says:

            I recall reading about a project using infrared cameras in airports for this purpose. They can’t detect absolute temperature accurately enough to see who has a fever, but they can detect relative temperature very accurately and hence can detect people who are slightly warmer than other people and single them out for scrutiny.

    • Matthias says:

      It’s unlikely the US will get as competent at tracing as Singapore already is. And Singapore has just announced a lockdown..

      • Chalid says:

        We won’t get as good at tracing as Singapore, but there’s room for the US (or any country) to be lots better at testing in two months than Singapore is now. Ubiquitous testing can make up for crappy or nonexistent tracing.

        • salvorhardin says:

          This. Or even just ubiquitous sampling. Which one could do like jury duty: you get a letter saying that you are in randomly chosen test group 605, please report to the testing center on Tuesday at 8:30, you’ll get a stipend for your inconvenience and a get-out-of-work letter, etc.

          A little science goes a long way. The Bay Area locked down earlier than the rest of the country because the Santa Clara County public health department, alarmed by their first community spread case in early March, scrounged together some tests and did a random sample of people presenting with flu symptoms. The CDC had been meaning to do that widely in February but abandoned their plans after screwing up their tests:

    • mtl1882 says:

      #1 has never made much sense to me as a major exit route (I expect that we will test, trace, and quarantine in some situations, of course, and that this will be useful, particularly the ability to tell if you are immune–but as a strategy it is essentially just a version of #4 that takes way more years and money). I think some combination of #2-5 will happen, with #4 being the most likely and primary approach, made easier as medical knowledge of how to handle this increases. I think there are a lot of other practical decisions we can make that can address the situation, which are not scenarios in themselves, but which build up to a pretty solid response that make ending the lockdown entirely reasonable.

      It looks like a few conditions are much more problematic than age itself, but correlate with age. I do think there are a lot of people under 70 at risk, and they will have to be very careful, and we should have things in place to support them. If they are right now locked down with a family member who is an essential worker, this is bad news. The big issue here does seem to be public transportation, which can’t just be done away with in cities, although it seems like masks really would make quite a difference. Especially if viral load plays a big role, as it seems a huge vector for spread is just talking. But depending on the data that comes out over the next month or so, I would imagine a lot of low-risk people would be somewhat relieved to get it over with this summer and be immune, if the hospitals aren’t overloaded, since it’s not going to go away and avoiding catching it is so difficult in the long-term. Then they’d be able to help vulnerable relatives and do public-facing jobs without worrying (assuming no reinfection) about either getting or spreading it. But we need better data on the risks. In any event, we’re going to have a chunk of people with immunity pretty soon–healthcare workers, police, first responders, etc., and officials/politicians who have to meet with each other to figure things out. There’s little way they won’t catch it quickly, as many have, especially with the atrocious PPE shortage, and they can’t just all go into quarantine. That will give us more information and hopefully lessen the spread.

      • Purplehermann says:

        +1 except hopefully #1 will be useful

        Also #2 (and #3 if it’s true) would make #5 a lot more attractive even if they aren’t as large factors as we hope. Especially because they would probably both be more useful for the low risk people (i assume asymptomatic cases are relatively concentrated in healthy demographics, and a small boost in bearing the virus per case seems like it’d be more useful to serious cases for otherwise healthy people)

    • Ninety-Three says:

      The testing part will get done eventually, one expects, though recent numbers have been disappointing. But the “trace and quarantine” part assumes a high level of competence on the part of the governments which failed to suppress the epidemic in the first place.

      The testing getting done eventually assumes at least a low level of competence on the part of the FDA et al. which have thus far been obstructionist bureaucracies. I’m not saying it definitely won’t happen, but “the FDA suddenly stops sucking” does not seem like a safe assumption.

    • WoollyAI says:

      The path to a loosening of restrictions by mid-late May, based primarily on data from California.

      The lockdown worked and Covid cases and deaths peak in late April, early May. Nationwide growth rates are dropping and in CA the daily Covid case/death growth rate has fallen to 10% from 20% a week ago. This makes sense because we went into lockdown early-mid March. If the disease has a two week incubation period and testing takes time then we should start to see that lockdown change the stats in late March, early April, which is what we’re currently seeing. If these trends hold, we should expect Covid to peak in late April or early May and, at least in California, it will do this without overwhelming the healthcare system.

      It would be foolish to stop the lockdown the moment Covid reaches it’s peak but after 2-3 weeks, when everyone is sure Covid is under control, there will be a lot of social and economic pressure to loosen restrictions. Restrictions will slowly be loosened while we watch to make sure Covid doesn’t spiral out of control again. Concerts and sit-down restaurants don’t come back and telecommuting workers keep telecommuting but workers who can’t telecommute return to the office/factory and various retail stores reopen with significantly enhanced sanitation.

      Without a cure it becomes highly unlikely that Covid can be eliminated so the rest of 2020 is spent managing Covid through variously strict levels of quarantine.

      • Edward Scizorhands says:

        Concerts and sit-down restaurants don’t come back

        I think sit-down restaurants will be near the first to come back, with limited seating. It won’t be a big spreader if someone sick shows up.

        Concerts? Maybe later. Depends if they can limit seating.

        Big sporting events? Those are gone indefinitely.

      • salvorhardin says:

        The big things I wish there would be more focus on are:

        — When and how can we reopen day cares, preschools, and primary schools to ease the burden on parents of young children, and the burden on the kids themselves of not getting to play in person with their friends?

        — When and how can we let people pursue their close friendships and romantic relationships, and seek/initiate new ones, more normally?

        I am biased because those are the things that would do the most to restore my own quality of life; but I am confident that they would do a lot for a large chunk of the population too.

        • albatross11 says:

          I think preschools/ schools for small children are going to be a hard thing to reopen, because it’s hard to avoid having stuff spread among the kids. Probably try to isolate grades/classes from each other, wash hands several times a day, and aggressively isolate/send home anyone with any symptoms, but I am not at all sure that’s going to be enough. That depends on what fraction of kids are asymptomtic as opposed to having what seems like just a cold.

          • Yea I agree on both points (that this would be a big quality of life improvement for many households and that it will be hard to do).

            For me and my neighbors it would help a bit if things improved just enough for us to be allowed to, and feel responsible about, doing outdoor playdates. Like a rotating thing where my 2nd-grader and a couple of neighbor kids were taken to a local park for a couple hours each day by a parent from one of the three households, rotate and repeat. Just that much would help a lot both in terms of the hours sans kidlet, and the young ‘uns getting outside to run around and work off energy.

          • salvorhardin says:

            I would also love to see small-group playdates/informal in-home childcare group rotations again. In the SF shelter in place order they made an exception for childcares for essential workers and said that these should be divided into groups of at most 12 and kept consistent (same staffers, same kids from week to week) and physically isolated from other groups. I’d imagine that just making that a legal way to run a class/childcare for non-essential folks would be a big win, especially over the summer when plenty of kids go to small summer camp groups.

            Of course, better data on whether kids are actually important vectors of spread would be even more useful. In SF there were a few scares just before the schools all closed where clusters of kids got flu-like symptoms and were all tested. To my knowledge, all those tests came back negative, though of course that is too small a sample.

        • sidereal says:

          > When and how can we let people pursue their close friendships and romantic relationships, and seek/initiate new ones, more normally?

          I assume people are already doing this, at their own judgement.

          I think this will become a (often uncomfortable) litmus for how much other people value your presence.

          • albatross11 says:

            I keep making this calculation: How many more people am I exposed to, at what level, by doing X.

            At one extreme, imagine hanging around all day in a small room with 20 people. I’m getting a high exposure to 20 people.

            At the other, imagine I’m picking up groceries and I have to step out of the car to sign the receipt, spending < 1 minute within 3 feet of the clerk. I'm getting a very low exposure to 1 person.

            Sleeping with your girlfriend is giving you very high exposure to one person–if she's got it, you're getting it. And the same goes the other way. Whether this is worth the risk is IMO subjective and based on your situation.

      • John Schilling says:

        Restrictions will slowly be loosened while we watch to make sure Covid doesn’t spiral out of control again.

        Except in the case of boiling pots, watching to make sure a thing doesn’t happen doesn’t actually make it not happen. Between the social and economic pressures you note against doing the think that makes a resurgence not happen, and the signal lag involved, the watching may turn into just an electric monk style feel-good measure.

    • keaswaran says:

      “In the US at least, I’ve seen no evidence that anyone in the federal government is actually working on building out this sort of [test, tract, and quarantine] program, and I don’t really see how this can work on the state or local level.”

      On the contrary – this seems like the sort of thing that really does need to be run on the local level, and state and federal involvement would be a distraction. For the vast majority of people, tracing back their contacts over the past few days wouldn’t require you to leave the metro area, particularly if we still have moderate lockdown of domestic travel.

      I suppose doing it at the local level means that we need to put up with a lot more roadblocks on interstates and other soft (or hard) internal borders that haven’t been a feature of most nations, but these internal borders seem just as important as international borders in countries the size of the United States.

      • Edward Scizorhands says:

        The Federal government should be writing the D&D rule books, and states/localities should be applying them, so every state/locality isn’t re-writing everything from scratch.

        The Feds probably can’t quarantine much — I’d rather not find out — besides their own facilities. They might be able to shut down the airports.

        • Le Maistre Chat says:

          The Federal government should be writing the D&D rule books, and states/localities should be applying them,

          If the Governor of Kentucky tells me to become a Wizard, I’m ready.

        • albatross11 says:

          They can certainly keep you off an airplane. I’d like to know why we’re not currently keeping people with fevers or other COVID-19 symptoms off airplanes.

  22. John Schilling says:

    Followup on the USS Theodore Roosevelt: Captain Brett Crozier has now been diagnosed with COVID-19 and is under quarantine (and hopefully appropriate treatment) in Guam. He’s reasonably young and healthy, so will probably be fine in the end, but no guarantees.

    • mtl1882 says:

      This Twitter thread is interesting:

      I had written up a response suggesting a similar take, but then figured the story had been done to death and deleted it. I might as well include it here, because the Twitter thread gets at the disconnect that I think is happening in a lot of places. I agree he probably expected to lose his job, and felt this was worth the risk.

      I’m not sure if I understand this incident correctly, but it seems like possibly Crozier and his higher-ups were basically in two different realities. The public conversation is one in which no risk is acceptable, and we’re going to eradicate this very soon. The military higher ups know that lot of people are just going to have to ride it out soon, particularly the military which may be needed in helping, that some of their people may die, but that it probably won’t be a total disaster among that demographic. More than that, it knows this isn’t going away. They’re looking at it with a much higher risk tolerance and no illusions of a quick fix. But the people in government thinking this way aren’t willing to say this very much because it alarms the public and goes against narrative. So they don’t call up Crozier and tell him that this is the deal: they know he can’t keep his men from getting sick, but they’re going to tough it out and see what happens, and that he should ignore the news version of what constitutes responsible action.

      So all Crozier hears is that distancing is necessary to avoid big problems, and is getting vague instructions to keep his men distanced and keep things under control. He can’t keep them distanced on the ship, and he doesn’t have the necessary equipment if things go bad. He assumes the Navy wants to avert all this but is somehow negligently failing to give him the tools, and feels he has no choice but to go public, being bombarded with messages about the responsible thing being staying isolated and making sure there is access to ventilators for those who get real sick. It seems the only remotely reasonable choice if he’s following what he’s hearing around him and not getting a clear picture from the Navy. Not one soldier’s life can be risked–this isn’t war, as he puts it.

      In reality, the Navy just doesn’t see much that it can do, or isn’t ready to respond yet. It knows that on a ship, the sailors will all get sick. It doesn’t view this as an unacceptable outcome, as it would be considered in the media. It expects them all to get sick eventually, as a lockdown until a vaccine arrives is just not possible, and their jobs can’t be done from home. It knows most of them will probably have mild symptoms, and it doesn’t really have the resources to address the issue, unless maybe case-by-case if things get bad. Letting them off the boat just increases the risk of spreading it to others, and there are security concerns. They figured Crozier could figure this out without being told. Maybe Crozier would have still objected to the Navy’s approach if communication had been clearer, and felt a need to go public. But I feel like there are two movies playing right now, though most are only aware of one, and that someone in his position would be likely to end up caught between them.

      • myst_05 says:

        I agree with your assessment. The military will likely have concluded by now that they need to achieve herd immunity amongst their personnel lest they risk an outbreak in the middle of an actual conflict. They cannot achieve this goal publicly, nor could they use variolation at this point, but simply letting the virus spread around military bases seems like a reasonable option.

      • The Nybbler says:

        If this is true, the whole “the captain did the right thing in the wrong way so, sadly, we must punish him” narrative is a lie. As usual; that narrative is nearly always a lie, used when the authority doesn’t want something done but doesn’t want to say so, and has (or thinks it has) cut off all effective methods for getting it done.

        • Conrad Honcho says:

          Doesn’t this mean he did the wrong thing in the wrong way, so we have to punish him?

          • The Nybbler says:

            If you want to punish him honestly, you have to admit that “the wrong thing” was attempting to protect his crew from COVID-19.

        • mtl1882 says:

          Agreed. It’s being unwilling to admit the reality of the situation.

      • John Schilling says:

        So all Crozier hears is that distancing is necessary to avoid big problems, and is getting vague instructions to keep his men distanced and keep things under control.

        I wouldn’t be at all surprised if Crozier had been getting very specific instructions to keep his men (impossibly) distanced, with a subtext of “so that when a bunch of your people die, it’s your fault for not following instructions and not our fault for demanding 110% mission focus”.

        Understanding what really happened probably requires knowing what communications Crozier had with his commanders before the famous open letter, but I’m not holding my breath for those to be released except in a form carefully and selectively redacted by the people who relieved him from command.

        • Matt M says:

          but I’m not holding my breath for those to be released except in a form carefully and selectively redacted by the people who relieved him from command.

          Hint for the journalists out there – the person who can get you this information is an enlisted Yeoman probably making less than $40K/yr. They may be cheaper to bribe than you think! Especially given how currently annoyed they probably are at the brass for firing their heroic “friend of the enlisted man” CO!

      • Matt M says:

        I find this theory possible, but unlikely.

        A commanding officer of an aircraft carrier is a very high and prestigious position, who would be let into the “inner circle,” when it comes to matters like this. The idea that his superiors had some “secret strategy” he was just supposed to intuit, and then relieved him of command for not properly doing so, seems a bit farfetched to me. That’s not how the military works. They’re generally pretty good at telling you exactly WTF you’re supposed to do.

        It’s one thing for the top brass to have secret plans that the average Joe Enlisted never hears about for political reasons. It’s a completely different one for them to try and keep secrets like that away from a guy who isn’t quite a flag officer yet but is clearly being groomed to become one.

        • mtl1882 says:

          Yeah, this is definitely a fair point, but I’m not really thinking they have a big secret strategy. More like they’re just looking at the reality of the situation as it unfolds and realizing there isn’t much of a plan. But having trouble saying this directly. As in, the plan is keep the soldiers as distanced as possible, which was communicated. In that situation, as distanced as possible isn’t going to go very far. Crozier assumed they were going for minimizing illness as much as possible, but they were just going for minimize it a little if you can, and thought he’d get that. I would assume they’d generally talk straight to someone of his rank, though, which is why I get your doubts. But I feel like right now people might be being unusually vague or just kind of uncertain, even among high ranking military, or not realize how different assumptions are out of their bubble.

        • Aapje says:

          @Matt M

          It’s pretty typical behavior for executives to expect the impossible from their inferiors, especially when the alternative is a hard decision that will make them harshly criticized no matter what they choose.

          Deciding between potential sailor deaths and taking ships out of the running seems like such a decision.

          The reason for this behavior can be:
          – delusion
          – wink wink, nudge nudge (you are supposed to do X, but I’ll never order it explicitly, so when shit hits the fan, I’ll just blame my inferiors/a scapegoat)
          – ducking (expecting/hoping for someone else to decide)
          – stalling (perhaps the problem goes away or I’ll retire/be promoted before it gets back to me)

          That’s not how the military works.

          From what I’ve heard, at the higher levels the military is a political organization where to get promoted, you need to be able to follow the ‘rules,’ rather than just or instead of merely following orders.

          • Matt M says:

            “Deciding between potential sailor deaths and taking ships out of the running seems like such a decision.”

            On the one hand, yeah, sure, that’s a tough dilemma with no easy answers.

            On the other hand, that’s like, the military’s job. No organization on Earth has more practice, experience, and training in “making tough decisions where you sometimes sacrifice human lives in pursuit of other goals” than the military does.

            And I think that’s part of what fascinates me so much about this story. COVID in general is a story largely about just how far society has swung in the “if it saves just one life” direction and how completely unwilling we are to sacrifice any amount of safety whatsoever.

            If this same kind of thinking is now directly in-play even in the military… that says a lot. If the Commanding Officer of an aircraft carrier has a policy of “I won’t do anything that risks the lives of my men,” then pretty much the only thing he can do is sit in port.

            The brass seems to understand that… but they’re accountable to politicians who are accountable to voters who are screaming for quarantines and lockdowns.

          • CatCube says:

            @Matt M

            When I was last in theater, the death of any single Soldier, whether accidental, suicide, or combat, required an AR 15-6 investigation conducted by at least a major (O-4). The results of this investigation were shared with the family, and I believe the family’s Congressman also typically took an interest. Can you imagine the bureaucratic burden if you had to do that for something like the assault on Omaha Beach?

            This is why you have troops rolling around in huge armored vehicles while wearing armored plates making them look like alien invaders and never talk to the man on the street, only meetings with village elders and the like–even if the “correct” way might be to accept higher casualties but build relationships with individuals, imagine having to defend that to the mother of somebody killed because you let them get a snack from a street vendor while not wearing body armor. Plus, even for internal military consumption, it’s common to tell your Soldiers, “I can’t get all of you back safely, but I’ll make sure the absolute minimum number of you don’t come back.”

            This, of course, has been successful if you measure by the rate of casualties–Afghanistan has lasted longer than Vietnam, with about a tenth of the deaths. But this mentality is going to be rough in a near-peer fight, where something like an in-stride breach is a GO in training if you only take 50% casualties in the process of creating a breach for the assaulting force.

            (I had an instructor at my Captain’s Career Course who had just come from being an Observer/Controller at the National Training Center when they had their first non-War on Terror training event in years.

            He told us about evaluating an engineer unit that was doing their first in-stride breach against a near-peer force. The *right* way to do this is to rush the breaching force to the wire and quickly prepare the reduction assets–I think MICLIC in this case–because the opposing force is going to start hammering you with artillery, and not the kind of love tap that two guys with a mortar do in Afghanistan, like a time-on-target barrage with an entire battery of 150mm guns.

            What this unit did was slooooooowly creep up to the concertina wire, and send one guy out to set a robot on the ground, which slooooowly wheeled up to the C-wire to tug on it to check for booby traps while the OCs on a hill watching this cradled their heads in their hands.

            Gonna be a major adjustment in thinking when facing an enemy that can take an entire brigade and make it not exist anymore, instead of terrorists that kill people a couple at a time.)

    • Matt M says:

      According to Twitter, the SECNAV has now resigned as well.

      • John Schilling says:

        Anyone who imagines that telling four thousand technically literate twenty-somethings to delete all their social-media recordings of a speech will decrease the public visibility of that speech, more generally anyone who thinks that flying halfway around the world to publicly insult a beloved officer in front of his men is going to do anything but blow up in his face, is too naive or too stupid to head an armed service in a first-world state.

        • Matt M says:

          BREAKING: John Schilling resigns from internet commentary following incendiary comments 🙂

        • bean says:

          Since there’s been about 6 months in the last decade where the holder of the office of SecNav isn’t someone who managed to win an “a cabbage could have done better” award, I think it’s incumbent upon all of us to call on President Trump to nominate a cabbage for the position.

          In a lot of ways, it’s impressive how thoroughly Modly managed to shoot himself in the foot. Or possibly in the head.

        • Various sources claim the speaker accused the officer of being naive or stupid. Unless I misunderstood what he said, that isn’t true. What he said, as I understood it, was that unless he was naive or stupid, he had to know that a letter sent to a lot of people would leak to the media. The implication is not that he was naive or stupid but that he wanted the letter to leak.

      • bullseye says:

        I read that the SECNAV was the one who “fired” Captain Crozier. But did he actually make the decision, or did the decision makers decide, “Hey, let’s make this guy do it”?

        • Edward Scizorhands says:

          Listening to his words, he seemed pretty eager to slag off on the Captain.

          • John Schilling says:

            He claims he proactively relieved Crozier from command (the man is still a Navy captain), because he wanted to get ahead of the curve where Donald Trump told him how to deal with Crozier.

          • Matt M says:

            He claims he proactively relieved Crozier from command (the man is still a Navy captain), because he wanted to get ahead of the curve where Donald Trump told him how to deal with Crozier.

            For maximum entertainment value, I am now desperately hoping Trump appoints Crozier as SECNAV.

  23. caldera says:

    We have not stepped foot outside our apartment for the past three weeks. However, we have a child who is due for his 15-month vaccinations (specifically, “Prevnar #4 or MMR, DTaP #4 and Hib #4 vaccines”). We asked if we could administer the vaccines ourselves at home – the doctor said no. I am debating whether the risk (to everyone in the family, including me with high blood pressure) of catching covid-19 by going to the doctors office outweighs the benefit to my son of receiving these vaccines at this time.

    The pediatrician’s office is pretty high volume, but has apparently segregated the practice into sick and non-sick areas, with two separate entrances. The office also seems pretty good about precautions: all doctors/nurses wear PPE and they do not let anyone come in if anyone in the family has been coughing/had a fever, etc. I am concerned, however, about the apparently high rate of asymptomatic transmission of covid-19. Even with all the precautions the pediatrician is taking, I have to think several covid-contagious individuals enter the office daily.

    Does anyone know how important it is to administer the 15 month vaccines on time, or have any ideas for how I should weigh the comparative risks here? Thanks

    • myst_05 says:

      If your child doesn’t interact with other children, how would he catch the diseases against which vaccinations are supposed to protect?

      • Kaitian says:

        Tetanus and Hib aren’t usually caught from other sick people, the bacteria are already in the environment and cause disease if the right circumstances arise.
        Measles are extremely contagious and it’s apparently common to contract them from being in a room with a sick person for a short time. So it may also be possible to catch it from walking past someone who has it.

        It’s unlikely that the child will get any of these diseases in quarantine. But it’s certainly not impossible.

    • cuke says:

      Not a doctor here but I am with myst_05 on this. Does your child’s doctor not make a recommendation either way?

      If I were in your shoes right now (our kids are older), I would wait. I don’t know where you are, but many areas are at max exponential growth in cases right now and health systems are still scrambling to adjust their systems.

      Kids seem to be most likely asymptomatic carriers, right?

      My understanding is that most vaccines are easily caught up on later. Pertussis is one where there have been local outbreaks due to lack of vaccination, but with your child otherwise isolated, risk of that kind of exposure would be low indeed.

      • Purplehermann says:

        Infants vs kids
        Kids mostly asymptomatic, infants I’ve heard are Moe susceptible

    • caldera says:

      To follow up on this post, we ended up rescheduling the doctor for a day when we were the first appointment. Beforehand, the doctor called to make sure no one in the family had any symptoms (coughing, fever, etc.). When we got to the office, the doctors and nurses were all wearing masks, and we only saw one other patient there at the same time. It has been about three weeks and no one in our household has shown any symptoms of COVID-19. Overall, we think we made the right decision in bringing our child in for vaccinations instead of postponing.

  24. Loriot says:

    I asked someone I know who works at the CDC (albeit not on infectious diseases) what he thought about all the criticism the CDC has been getting lately. Much to my surprise, he said he agreed with it, and specifically mentioned a Washington Post article saying that the CDC had gotten complacent after successful responses to SARS, MERS, H1N1, Ebola, etc.

    He also complained about the director of the CDC being a political appointee with no experience, but that’s a bit CWish.

  25. JRM says:

    I am a lawyer, but I am not your laywer. This is not legal advice. Legal advice, you pay for. And you get a competent lawyer, not someone in a totally different field.

    I am not an expert on IRB stuff. I am REALLY not an expert on administrative law as it may affect your profession; I am an expert on criminal law, and (relative to y’all and the average law professor) constitutional law. So if you get thrown out of the psychiatrist’s guild that will suck for you and you certainly can sue me and file a complaint with the bar, but you’ll be laughed out of court and I’ll spend the next year answering the bar’s questions.

    I think it depends on the thing you are advising people to take. If you advise A-M to take LSD and N-Z to take tic-tacs, you’re creating risks for people that all of this is designed to avoid. If you’re doing a study to determine whether a small dose or a larger dose of Vitamin C is OK and you run a controlled study with people, it’s not criminal.

    I am quite sure you want administrative answers; I suspect the answers to that are, “You are almost certainly legally permitted to do this [long analysis omitted because it sounds legal advice-y] but that doesn’t mean someone will disagree and you’ll be run through the system.”

    I would 100% ignore this comment if you get advice from someone in the actual field. I believe if there were a criminal code section barring this, it would be unconstitutional, but administrative codes are different and scary.

    • mefrem says:

      Why would a criminal code section barring a non-IRB study be unconstitutional? Which provision would it contravene? And are you referring to a state or federal criminal code?

      Thank you

      • JRM says:

        Very short version: Speech cannot be regulated by criminal sanctions unless it meets certain criteria, which this clearly would not reach. Speech can be regulated by administrative agencies to a substantially greater degree.

    • Matthias says:

      Would Scott ‘selling’ his mailing list data to someone in a different jurisdiction to send oout the actual emails make a difference?

      • b_jonas says:

        Ignore the “email list” part, that was a goof by Scott. If he did such an experiment, he would almost certainly ask for it in a blog post, because more people read that, and Scott won’t send unsolicited emails to people who didn’t subscribe to such a thing, but subscribed only to reply notifications.

        (Not that he could do such a study anyway. How would he get everyone to sign the consent form with a pen, not pencil, through email?)

    • theodidactus says:

      I’m a law student, even if I was a lawyer, I wouldn’t be your lawyer, and like JRM I’m not really in the right area (Criminal law, also).

      having said that:
      I echo JRM’s comment that a blanket rule that said something like “you cannot do research without IRB approval. Criminal penalties may apply” would probably be unconstitutional, however, what you need is an agency/admin lawyer. Admin law is a wild and frightening morass, not unlike Mirkwood or Mordor…certainly an area of law that frightens me…and I’m doing criminal law.

  26. HeelBearCub says:

    Not sure if it really warrants a mention, but Boris Johnson has now been hospitalized.

    aides [said] he ‘looked dreadful’ during his isolation period

    – Provides a clear reminder of how long the disease can take to fully manifest, as he started experiencing symptoms 10 days ago.
    – We don’t lose heads of “1st world” states while they are in office all that often anymore. Who was actually the most recent one? Not sure what kind of effect that would have in the UK, or worldwide, especially with a pandemic currently raging

    • AlphaGamma says:

      – We don’t lose heads of “1st world” states while they are in office all that often anymore. Who was actually the most recent one? Not sure what kind of effect that would have in the UK, or worldwide, especially with a pandemic currently raging

      Johnson is Head of Government not Head of State.

      If Poland is first world, there was President Lech Kaczynski who died in a plane crash in 2005- I’m not sure how much power the Polish President exercises on a day-to-day basis. That crash did have a huge effect on Poland, but many other important figures in the Polish government also died.

      Austrian President (AFAIK a mostly ceremonial post) Thomas Klestil died of a heart attack aged 71 the year before, but there might be something about that specific job- the majority of Presidents of the Second Austrian Republic have died in office, though Klestil was the first one since 1974.

      Otherwise there are Prime Ministers Yitzhak Rabin of Israel (1995) and Olof Palme of Sweden (1986), both assassinated. And various Caribbean PMs have died in office, most recently David Thompson of Barbados.

      The last Head of Government of an unequivocally first world country who I can find who died in office of natural causes was PM Masayoshi Ohira of Japan in 1980.

      There have also been a few cases more recently where serving heads of government have suffered a fatal stroke while in office but been removed from office on the grounds of incapacity before they actually died (so they didn’t technically die in office)- Keizo Obuchi of Japan in 2000, and Ariel Sharon of Israel in 2006.

      • matkoniecz says:

        If Poland is first world, there was President Lech Kaczynski who died in a plane crash in 2005- I’m not sure how much power the Polish President exercises on a day-to-day basis. That crash did have a huge effect on Poland, but many other important figures in the Polish government also died.

        Impact from change who acts as a president was basically none. Effects were huge, but for different reasons.

        Dying president was mostly symbolic (for multiple reasons, starting from fact they were visiting site of a Katyń Massacre).

        I’m not sure how much power the Polish President exercises on a day-to-day basis.

        Most of powers are ceremonial + official/unofficial diplomacy. President may also veto legislation (overridden by 3/5 of parliament) and issue pardons.

      • HeelBearCub says:

        Johnson is Head of Government not Head of State.

        You know, I know the difference, and yet the default phrase in America is almost always “Head of State”. Fingers outpaced my brain there.

        An interesting list. Rabin had come to mind, and that one very well have been consequential in exactly the way that the assassin intended.

        • Garrett says:

          This is because the President is both Head of State and Head of Government. And Head of State sounds more impressive.

    • Radu Floricica says:

      Reminds me of this (linked somewhere in this thread):

      2-11 days after exposure (day 5 on average) flu like symptoms start. Common are fever, headache, dry cough, myalgias(back pain), nausea without vomiting, abdominal discomfort with some diarrhea, loss of smell, anorexia, fatigue.
      Day 5 of symptoms- increased SOB, and bilateral viral pneumonia from direct viral damage to lung parenchyma.
      Day 10- Cytokine storm leading to acute ARDS and multiorgan failure. You can literally watch it happen in a matter of hours.
      81% mild symptoms, 14% severe symptoms requiring hospitalization, 5% critical.

    • Fitzroy says:

      Not sure what kind of effect that would have in the UK

      Very little. One of the other holders of the Great Offices of State (Chancellor of the Exchequer, Foreign Secretary, Home Secretary) would almost certainly take over as interim Prime Minister whilst the Conservative party organised a rapid leadership election. I suspect that leadership election would be sewn up within the parliamentary party and not require going to the membership.

      My money would be on Dominic Raab, the Foreign Secretary – Sunak is too green and nobody trusts Patel.

      • Lambert says:

        In the longer term, it’ll definitely have an effect. The Government currently has a large majority and Boris seems to have consolidated a lot of power within that gov’t.

        I see Boris’ manifesto going much the same way as Caesar’s plans to invade Parthia.

    • matkoniecz says:

      We don’t lose heads of “1st world” states while they are in office all that often anymore.

      Are we counting Vatican as “1st world state”? Though that would be kind of cheating for this selection..

      • keaswaran says:

        Even there, within my lifetime, the dying-in-office rate of the Vatican is much lower than it historically is. There have been two Popes whose term ended during my lifetime, and one of those ended without his death. I believe that in historical terms, far less than 5% of Papacies end by methods other than the death of the Pope.

  27. scherzando says:

    Does anyone have opinions on GiveDirectly’s US-based and international COVID-19 relief programs and how they likely compare to other options for donations in their respective (geographical and subject-matter) areas? My priors are that cash distribution is likely to be a good way for individuals and charities to address the economic side of this situation, and that GiveDirectly is likely to implement this well, but I don’t feel super well-informed on the particulars and am interested in other people’s assessments.

  28. mtl1882 says:

    I generally agree, especially that we are placing too much emphasis on ventilators. It’s a little hard to tell with the current data, which hopefully will become more complete in the next few weeks, including how various national strategies compare. I am concerned especially about impacts on our medication stockpiles, which I think we should be spending more time securing than ventilators (this may not be an either/or, but it seems like little focus is happening on non-ventilator stuff) and I’m really concerned the the effects of the shut down on supply chains and finances will risk shortages. Being on a ventilator generally requires sedation, so we need to make sure we have drugs for that. More than that, the only thing worse than not getting a ventilator is not getting one (or being too frail for one) and not having any painkillers and sedatives and all that. Palliative care is a real concern here. The attitude that we can defeat death and end this thing relatively quickly makes me worry we’re pinning unrealistic hopes on technical solutions and missing more basic things like these medications, which reducing suffering for both those who survive and those who don’t.

    Certainly, we want to have ventilators available for those for whom it is most likely to help, but at that point things are already pretty grim. Most hospitalized Covid patients need less intensive forms of care. And we have a lot of people who need treatment for other very serious things that we can’t put off forever. We should be splitting up our healthcare facilitates to handle Covid and non Covid patients separately, making reasonable plans for triage/palliative care, and also setting aside facilities and services for people with conditions that make them vulnerable (which seems to be a much bigger factor than age itself), as they will have to be careful for the foreseeable future. We also want to keep them away from hospitals if at all possible, and have special treatment wards for them. I suspect healthcare workers, police, etc. are the main vectors of spread right now because they can’t help but get infected and can’t just all go into quarantine at once. We’re essentially letting them develop herd immunity first. I don’t think there’s any hope of containment, and we need to admit that and be realistic–be smart in taking mitigation efforts that address specific populations, issues, and locations. There are a lot of practical actions we can take, and problems that need solving. This is all assuming we don’t get any really scary news, like lots of lung damage in survivors or high rates of reinfection, in the next month, which should give us a lot of data. But I think we need to reopen to some considerable extent, with some obvious caveats including strong recommendation to wear a mask, no big gatherings, little air travel, acquiring stockpiles of necessary equipment, and sheltering of the vulnerable, or I’m very afraid of what might happen. But I think the shutdown was a pretty reasonable choice so far while we tried to slow things down and assess them, though I don’t think the government’s policies have been adequate or that the time has been used particularly wisely.

  29. AlexanderTheGrand says:

    For bullet 2, they seem to conveniently discard the “treatment” part of their sentence pretty quickly. As discussed lower in this OT, you have to be VERY sure a preventative vaccine doesn’t do more harm than good since you’re giving it to currently healthy people (1/1000 people having bad side effects, and you’ve done more harm than good). On the other hand, the bar for treatment is much lower because we know the outcome can be serious without intervention. Point being, the 18 months argument doesn’t apply to the dozens of clinical trials going on right now around the world. If we find a treatment that halves the number of serious cases in 3 months, we’ll be happy that we slowed the disease’s progress.

    It also seems to ignore the multiple countries which have gotten their outbreaks under control, as examples of lockdown procedures which did have an end in sight.

    The focus on ventilators may be a fair point, but his two bullet points are definitely not an exhaustive list of ways that shelter-in-place can help.

  30. mefrem says:

    I would like any constructive input from the community about maintaining a possibly months-long job hunt for a junior position (in software—data science specifically). I’ve savings enough to weather months of the searching, but am burning out from lack of clarity about how to spend my time most efficiently across meaningful networking, project/portfolio development, and self-study. And generally trying to stay sane/sharp/helpful.

    • zardoz says:

      I don’t know much about data science, but if you have a computer science degree, all you really need to do to get a junior position as a software developer is send out resumes. You need to tailor each resume to the company that you’re applying to. Include related project work that you did at the university. Add a cover letter about why you’re interested in their company specifically. Set a target like applying to a new company every day and just keep doing that until people respond.

      This isn’t art school. You don’t need a portfolio or github contributions to get hired– or at least, not at the junior level, which is the level you said you were interested in. To be totally honest, self-study is probably not a good use of time either unless the thing you’re studying is the interview questions.

      One more thing, though. If you don’t have the academic qualifications for what you’re trying to do, most companies won’t take a chance on you. I don’t know how it works in data science, but for a traditional programming job no degree would be an instant nope from any big company, and most small ones as well. If you have literally no data science background you probably should take some kind of course, even an online one, to fix that before you try to apply for those kind of jobs.

      • Wasserschwein says:

        My experience regarding programming jobs and academic qualifications is the opposite of yours. I would think this differs a lot from one country to another.

        • Kaitian says:

          Where I am, it really depends on the company. Larger companies will throw your resume in the trash if you don’t have a stem degree, but many smaller ones will not care. It also depends on the product – there are some jobs where you actually need good knowledge of advanced mathematics or circuitry, and while it’s probably possible to self study those things, the average self taught programmer just learns how to code.

          If you can, you could always start with an internship or freelance jobs, then you can prove “experience” to the company you want to hire you. That’s what many people I know did.

        • zardoz says:

          It has changed over time as well. You hear a lot of stories about people dropping out of undergraduate programs to join (or found) a software company in the 1970s or 1980s. Probably a good move then, but not so much today.

          Get an undergrad degree, kids. You don’t need a graduate one for programming but you should have an undergrad degree if you’re just getting started.

          P.S. A lot of old-timers are kind of grandfathered in because they joined the industry when the rules were different. Although at least one of them has told me stories of being hassled by the a BigCo HR department over this very issue.

          P.P.S. You might still be able to get work building websites or something without a degree. It’s not a good career path.

          • JayT says:

            I think this is good advice if you are an average programmer. If you are exceptional, I don’t think the degree matters all that much. Some of the best people I’ve known in the industry left school early for work.

          • HeelBearCub says:

            The problem being that way more people think they are exceptional programmers than actually are exceptional programmers. Just like everything else.

            In addition, I have hunch that “leaving school early for work” might mean “was recruited to a permanent job from their internship”. If that’s the case, then that avenue is only open if you can get the internship, which is vastly easier if you are pursuing a degree.

          • JayT says:

            Well, part of it is that they actually can get that job without a degree in the first place. I know some people that went to an internship and never went back to school and I know some people that were programming before college and got work based off of that. If one of those two things happens to you, then you are probably an exceptional programmer, and I would recommend that you take the job instead of going to or back to school. If it doesn’t work out, you can always go back, but if it does work out, it’s very hard to get those four years of salary back.

          • zardoz says:

            If you’re an intern and the company gives you a chance to skip completing your degree and work for them full-time, you probably shouldn’t take it.

            A career in software typically lasts many decades. Lacking a degree will be a handicap to you long after your first job is forgotten.

            You will also miss out on the chance to intern at a different company, that you might like even better than the first one, the next summer or two. In game show terms, why wouldn’t you want to open more doors and see more possible prizes you could claim?

            You always have the chance to apply to the original company that gave you an offer, once you graduate. If they liked you as an intern, that will certainly count in your favor, even a year or two later.

            The only exception to this rule is if the fist company you interned at is a startup company that you think will make you a lot of money in a very short amount of time. Then maybe it’s reasonable to drop out and work for them. Of course, you’re probably wrong about your assessment (most startups fail). But it’s at least the kind of Hail Mary move that might, possibly work. And if you succeed, you’ll attain some minor place in Silicon Valley mythology, and help motivate future generations of kids to make a bunch of bad decisions.

          • T3t says:

            @zardoz (regarding the comment on internships, in case nesting breaks): I disagree (conditional on the reader living in the US). It is true that most well-established tech companies are not in the habit of poaching undergrads from their degree programs. If you find yourself working for an employer who does make you such an offer, I would strongly advise simply running the numbers on the opportunity cost of forgoing x years of income as a software engineer, including compound growth from investing in the market with your newfound savings.

            As for the lost chance to intern at different companies… suffice it to say that you can always get an actual job at those companies if you are that curious; yes, it is more expensive in terms of time (3-month stints on a resume don’t look great unless you’re doing contract work) but it is still eminently possible. And, again, it is probably not worth it. Even ignoring pay, in most situations you will learn more relevant skills, faster and better, in the industry than in university.

            However, while a startup is more likely to offer you such an arrangement, I would say the trade-off there gets worse the earlier-stage that startup is. Early-stage startups (generally) come with lower base pay in exchange for a greater equity stake, but equity in early-stage startups is a sucker’s bet unless you’re literally a co-founder (or somehow wrangle RSUs instead of options, and even then you’d need a significant % to stand a real chance of overcoming dilution/preference/etc in the event of a successful exit). Which isn’t to say you shouldn’t do it – even “lower base pay” is often quite a bit of money (especially compared to, um, paying an institution instead).

            In my estimation the value of a CS degree falls off a cliff once you can list full-time employment on your resume. The best employers along more or less any metric you consider relevant (pay, work-life balance, culture, mission, etc) do not have strict filters that require candidates have traditional 4-year degrees. The lost “opportunity” to work for the remainder that do is not likely to be worth the 6+ figures of income + compounded market growth that getting a head-start on your career would be worth.

          • Edward Scizorhands says:

            Working at a start-up is not just working for a lottery ticket. It’s working for a lottery ticket that the lottery commission can revoke at any time unless you follow a set of rules that you don’t know about, by virtue of you being a newbie.

            If you want to leave the company, you will be lacking one of the best credentials you could have (especially early in your career). A company that cares about your long-term development will want you to get your degree.

            I know many people who don’t have a college degree and have done extremely well in their careers. Every single one of them with kids, with zero exceptions, has made sure that their kids are on a college track and are going to get the college degree.

      • mefrem says:

        My undergrad was in economics, and I’m finishing a 9-month bootcamp in software development and data science. So I don’t have projects from university, just from the last year. You can see them at my site. Even my site is tailored to distinguish myself by showcasing my intellectual curiosity, ethics, projects, etc.

        If were to summarize your advice in light of my question, it would be to maximize time spent on job searching, networking, and tailoring your applications to the specific company and role. Is that fair?

        • yodelyak says:

          Link is wrong. You meant:

        • zardoz says:

          What has always worked for me is applying to a lot of companies. It’s a numbers game in the end.

          Unfortunately you’re applying at a bad time, because a lot of companies have hiring freezes, and the ones that don’t probably have unofficial “let’s slow down on hiring, guys” policies in place. In a situation like this you might have to choose between accepting a job that you think is beneath you, or going back to college to double down on upskilling. It’s a classic “new grad during a depression” choice (I am told the CS PhD application process in 2001 and 2002 was very, very competitive!)

          Networking helps, but you have to actually know some people to do it. I wouldn’t consider cold-calling or cold-emailing people to be networking (although you should absolutely do that), but just a normal part of applying to a lot of places. Some people like to network at conferences (although I’ve never had much luck with that), but, well… you know. Not now.

    • Bugmaster says:

      From what I’ve seen, even if the coronavirus restrictions do get lifted completely in a couple of months — which I doubt — the economic depression is here to stay; for years if not decades. With every day that passes, your chances of finding a better job drop another notch, so I’d suggest finding one ASAP.

    • T3t says:

      If you are in the US, particularly in a tech hub (or willing to relocate to one, or planning on working remotely for a company located in one), ignore the rest of the advice in the replies to your comment. It is well-meant but inaccurate either due to differences across generations or geography.

      Your #1 priority should be optimizing for landing and passing interviews. Figure out where your bottleneck is and attack it directly.

      If you are not getting interviews, it is likely that you are sending something that is a strong negative signal rather than sending insufficient positive signals. Your website is quite good, especially considering that you are taking the Data Science track rather than Full Stack. I wouldn’t worry too much about it (unless someone here has a good idea for a different font).

      If you are not passing non-technical phone screens, you are not “playing the game” correctly (i.e. signalling the right things in terms of interest/knowledge of the company you’re talking to, giving socially acceptable answers to questions like “why are you interested in this job?”, etc), or you are talking to places that don’t really know what they’re doing (I’d say you don’t want to work at those places, and given an abundance of options you probably shouldn’t, but I recognize that we’re in a different situation now than we were a few months ago).

      If you are not passing technical phone screens (or online coding challenges), then you need to directly practice those skills. Treat this as an optimization problem. Your goal is to learn with sufficient depth the ~10 most commonly applicable patterns/structures/algorithms required to solve problems asked in interviews in your domain, ideally well enough that you can carry on a conversation explaining what you are doing and why in real-time. There are websites that explicitly allow you to practice those skills – for software interviews, I’m familiar with and pramp. They may or may not be useful for Data Science interviews, though I’d be surprised if there weren’t comparable sites.

      If you are not passing “on-site” interviews (such as they are, nowdays), more or less the same advice as from the previous section applies, though with obvious caveats that different companies care about different things and some may have their own quirks (i.e. Amazon generally has a round dedicated to drilling you on their “leadership principles” and how you exemplify them).

      You do not need to worry about having a 4-year degree in Computer Science (or, in your case, a Masters/PhD in CS/Math/Stats/Econ/whatever). Yes, there are places that will filter you out as a result. Those places are not Google, Facebook, Amazon, Microsoft, Apple, etc, nor are they most well-paying Silicon Valley startups. I would argue that, probabilistically, any institution that has this as a strict requirement will be a considerably worse place to work (in terms of culture, career growth, pay, all the good stuff) than places that do not. Also, after your first paid full-time job in the industry, a huge percentage of companies that tossed your resume in the trash the first time around will be climbing over themselves to talk to you. Paid full-time work is considerably stronger social proof that you are capable of performing the duties expected of you than a 4-year degree.

      You do not need to worry about the case where you are an average programmer (or data scientist). You are not an average.

      I do agree that this is largely a numbers game, but there are ways to cut the numbers down significantly. If you want to discuss this further or have any specific questions, I’d be more than happy to continue this conversation. My email (rot13) is orggrefpnyr@cebgbaznvy.pbz. You can also find me on the Los Angeles LW/SSC Discord Server:

  31. Le Maistre Chat says:

    Ancient Egyptian ushabti figures expressed the hope of automating labor away.

    Illumine the Osiris [name of the deceased], whose word is truth. Hail, Shabti Figure! If the Osiris [name of the deceased] be decreed to do any of the work which is to be done in Khert-Neter, let everything which standeth in the way be removed from him- whether it be to plough the fields, or to fill the channels with water, or to carry sand from the East to the West. The Shabti Figure replieth: “I will do it, verily I am here when thou callest”.

    The Book of the Dead, ch. 6

    • FLWAB says:

      This is interesting, and gives me some neat ideas for a Dungeons and Dragons encounter: a mummy served by an army of inanimate object Ashanti.

  32. BlueGold says:

    Do an IRB.

    The general standard is if you plan on creating generalizable knowledge and if the risks would be any greater than those experienced in daily life. If you’d ever plan to present the research, either on the blog at a conference or in a paper then it would be generalizable. If the vitamin in question had any documented risks associated then that would be more than what people might encounter in daily life. When does a study switch from decentralized experiment to official? Usually when grant money gets involved, but in this case the best bet would be as soon as you hit those two points. The blog is already generalizable knowledge so really it boils down to risks. Are you exposing participants to potentially more risk than they experience in daily life? If the answer is yes, probably, possibly, maybe, or anything other than no, you should do an IRB.

    As far as how much trouble, it would depend entirely on your goal and what you did with it. There are legal guidelines around ethical research. Odds are skipping an IRB will not run you into them, especially if you are not using grant money. However, skipping an IRB sends a fairly clear signal to journals, research institutions, and everyone else doing research that you are not interested in adhering to ethical standards. This is a bad place to be. At minimum your results will be suspect and nonpunishable and you will probably need to sit through IRB training. At maximum your institution might feel the need to censure or fire you.

    If you don’t mind those things or if it is plausible you didn’t realize the risks until the study began then those penalties might not matter/apply. It’s also worth mentioning IRBs at universities generally have a streamlined process for getting post hoc approval because a mini-study done by a student for class is not subject to IRB. But as soon as that student decides to try publishing they need IRB approval so schools generally have a process in place to handle that. So if you realize there are potential risks halfway through you can and should still submit to IRB at that point.

    Getting an IRB together at a place that doesn’t already have one might be a bit of a struggle but I would suggest going to the nearest reputable research institution/university and asking their IRB coordinator if you could submit your work for review. Everyone on the IRB board is typically a volunteer so they shouldn’t really be barring you just because you aren’t a member of their particular institution. Most places will have an online form to complete and it’s highly likely you could do all of it via email.

    Ultimately the requirement for an IRB is part of a larger obligation that you adhere to ethical research practices. As a member of any scientific community, ethical research helps you maintain legitimacy, integrity of results and process, and ensures a standard of moral responsibility via oversight for those in a position of power over research subjects, particularly when those research subjects are human.

    tl;dr Do an IRB.

    • Radu Floricica says:

      Is it possible to outsource an IRB? I mean, I’d probably donate $100 for that, half out of spite for bureaucracy. But for this particular case, you might find volunteers.

      I think the frustration and waste of time are a much greater source of friction than converting them to money and paying that amount. Plus, it might actually be good business: doing this centralized could make things cheaper, more scalable, AND much more professional.

      • sclmlw says:

        The term of art in the business is ‘central IRB’, and there are a few. You’ll pay a lot more for it than most blogs are into, but they’re also a lot quicker and less of a hassle than IRBs attached to institutions. They’re basically an IRB-for-hire open to anyone who needs their services. A few options are Advarra, Schulman, and Western IRB (WIRB).

      • BlueGold says:

        Technically, at almost any major institution, everyone except the office staff/coordinator on an IRB is almost always a volunteer (further, those paid people typically aren’t on the actual review board itself, they’re just the ones that file and monitor all the paperwork). A university is certainly within it’s rights to deny a non-member access to their IRB board. But at the same time that’d be pretty crappy given it represents extremely minimal costs to actually do the IRB. I imagine if Scott just went to a professor vaguely interested in what he does and asked if it would be possible they would be willing to try and make it work.

        It’s also not that frustrating or that much a waste of time. Having done a number of IRBs all using human subjects (the harder and more tedious kind) most of them amount to filling out a form and emailing it to the right person. Generally you don’t even need to follow up until the study is done, if it lasts longer than 6 months, or if you’re bringing new people onto the research team. People get antsy about it when you seem like you’re breaking one of the conventions of an IRB, not when you do one in the first place.

        Unless Scott wants to do something that involves significant risks he should be in the same boat. Fill out the form, do some online training, and maybe find a professor or other person of credentials and repute to sign on as the study supervisor. His example would even be eligible for expedited review, which means it wouldn’t even go in front of a full IRB panel and would pass basically by default.

        Really, a properly done IRB isn’t a waste of time. It typically takes about a week for the IRB office to turn around most requests, and expedited reviews can be as quick as a couple of days. Medical stuff will generally be on the longer side, but that’s because the risks are typically greater.

        • FLWAB says:

          It’s also not that frustrating or that much a waste of time. Having done a number of IRBs all using human subjects (the harder and more tedious kind) most of them amount to filling out a form and emailing it to the right person. Generally you don’t even need to follow up until the study is done, if it lasts longer than 6 months, or if you’re bringing new people onto the research team.

          The last time Scott tried to do an IRB it did not match your experience. I can understand why he would want to avoid doing it again. The study consisted of giving inpatients the same screening quiz they normally give them, but checking to see if the screening was accurate. In the end after two years of work the study had to be scrapped due to not meeting IRB requirments. His final thoughts:

          I feel like a study that realistically could have been done by one person in a couple of hours got dragged out into hundreds of hours of paperwork hell for an entire team of miserable doctors. I think its scientific integrity was screwed up by stupid requirements like the one about breaking blinding, and the patients involved were put through unnecessary trouble by being forced to sign endless consent forms screaming to them about nonexistent risks.

          I feel like I was dragged almost to the point of needing to be in a psychiatric hospital myself, while my colleagues who just used the bipolar screening test – without making the mistake of trying to check if it works – continue to do so without anybody questioning them or giving them the slightest bit of aggravation.

          • BlueGold says:

            That’s fair, but most of his post reads like someone with no training getting frustrated by a combination of a lack of resources and lack of available training.

            For example, he brings up needing pens to sign consent forms. This is a reasonable precaution because pencils can be erased. The easy way around it for someone in a position like his (which is unique and a hassle, but literally only an applicable hassle in this extremely narrow area of psychiatric patients that are banned the use of pens) is to simply use electronic consent forms. Easy fix would be to just have people fill out the form online or just sit them in front of a computer. Notably, electronic records are also easier to store in an IRB compliant manner. You’d literally just have to agree not to leave the device they’re stored on unsupervised in an unsecured location (e.g., leaving it sitting in your unlocked car while you aren’t driving it) and to have a password.

            Similarly, he complains about the risk assessment and minimization paragraphs. This is bog standard and a normal part of the protocol. I have to write these when administering something as basic as a survey. Yes, they don’t apply to a lot of research. But the point is not that they don’t apply it’s to show that you’ve sat down and thought about this question, even when you don’t think it matters. You might be wrong. And if you’re wrong people get hurt, so you have an ethical obligation to think about this in advance irregardless of how silly you might think it is.

            Also worth mentioning, although it’s been said before in this thread and the thread on that post, his institution looks like it was simply terribly run and set up for research. The fact that it took such a long time to find the expedited form for example is ridiculous. Most IRB offices vastly prefer expedited forms and a competent coordinator would have pushed him toward this option. Similarly, a PI with prior experience would have done the same. There are a lot of other, similar, examples in the post like the difficulty of finding a secure room, something most institutions have specifically set aside for researchers. This is unfortunate, but not an indictment of the IRB so much as an indictment of the hospital itself.

            Obviously this is a hurdle for inexperienced researchers which is unfortunate. But at the same time, this is exactly why graduate students do theses and dissertations, to get experience navigating this process under the auspices of those who have done so already. Missing out on the training and expecting to be very good at something isn’t particularly rational, even if this is an area where you might not think you need to be trained.

          • The Nybbler says:

            The bottom line is that all the red tape stopped a reasonably determined person from doing a study in their own area of expertise. That you say there was another way through some of these brick walls doesn’t change that, and anyway there’s no guarantee there weren’t more brick walls beyond those.

            If you want the tidy world of IRBs where every i is dotted and every t crossed and signed in triplicate by all involved, you have to accept that you’re going to reduce productive output by significant amounts, possibly orders of magnitude. Pointing out that if the people stymied had only tried THAT MUCH HARDER to get past the barriers you imposed they could have done it doesn’t change that.

          • HeelBearCub says:

            @The Nybbler:
            No, Scott tried to research at a place that wasn’t a research institution. It’s really as simple as that.

            Domain expertise does actually matter.

          • The Nybbler says:

            So there’s an additional barrier: not only must you got through an IRB, it has to be an IRB at a research institution. If you’re not in a research institution, you can’t do research. I don’t think that contradicts my point. As for “domain expertise”, the “domain” appears to be “paperwork”.

          • HeelBearCub says:

            Would you expect your traffic-ticket lawyer to be the most competent to defend you in a murder lawsuit? No? Why not? They are a lawyer after all.

          • matkoniecz says:

            People here have problem with fact that you need a lawyer to run a scientific study.

          • BlueGold says:

            You don’t really need a lawyer, you need someone with experience going through the process. Again, a PI with experience would immediately have defaulted to the expedited form and Scott would never have even seen the longer version or been scared off by the questions in it. One of many examples of how someone without a law degree but some basic experience in this area could have fixed most of the issues they ran into.

            It’s also not really red tape, that’s a poor comparison. This would be like sitting down with a mechanic that has never touched a small engine to work on a car, you wouldn’t call it red tape that they did a poor/sub-optimal job it would just be how the world works. To echo the other poster, domain knowledge matters. Someone that works on cars every day will do better than the same mechanic that works on airplanes. Similarly, the mechanic will succeed more if they have appropriate tools and an actual work area. Yes, the IRB requires sometimes odd paperwork. But to a mechanic that has never opened up the hood of a car, some of the engine’s parts and such probably seem pointless or odd or deliberately hard to use/fix.

            Nor is this about the IRB being at a research institution. Why would I expect to find fishing experts living 10000 miles inland? For the same reason, why would I expect to find the best IRB coordinators at an institution that does not do research? Yes, better resources exist as research institutes. One of those resources is domain knowledge.

            Similarly, it would be a mistake to conflate Scott’s experience in his own area with experience in the IRB. It’s tempting to think of them as the same thing, but they aren’t. All the knowledge of psychiatry, sociology, geology, or any other field in the world has nothing to do with knowledge of how to address IRB concerns.

          • albatross11 says:

            So the question is whether or not the benefits of the IRB system as it is make up for the lost research–the stuff that isn’t done because the researcher or institution doesn’t have the right expertise to get through the process, or the institution’s IRB is a clusterf–k and nothing gets through there. I am certainly no expert, but I strongly suspect something a lot more lightweight than the current system could just as well protect against future Tuskeegee experiments and the like.

            ETA: One thing I hope we get out of the COVID-19 pandemic is a renewed willingness to look at our safety culture w.r.t. medicine and research. The scientist who detected community spread of COVID-19 while the FDA was dicking around ineffectually with its test kit was violating both laws regarding medical testing and rules regarding IRB approval. I’m pretty glad she went ahead and did it anyway. That’s not the normal situation in research, but neither is Tuskeegee/Thalidomide/etc.

          • The Nybbler says:

            Tuskeegee practically justifies everything IRBs do and thalidomide practically justifies everything the FDA does and even if you come up with real and egregious examples of the downsides, the responses here prove there’s always plenty of people who will tell you that “you’re just holding it wrong”.

          • Radu Floricica says:

            We just got here a very light version of “doing your taxes” – basically filling out a PDF file. I said “fuck it” and went to an accountant – since it was the first year, it took about an hour and a half. This year I expect it to take half an hour by phone. I don’t even want to imagine the clusterfuck of trying to figure it out alone.

            Some things are not worth doing without expert help. I mean, technically you can, but when somebody else can vastly improve the experience, why even consider not using them?

    • Douglas Knight says:

      Yes, exactly. Doing on an IRB is endorsing institutional ethics.

  33. Clutzy says:

    The part “very smart reader” leaves out is that hospitals often become breeding grounds for things of this ilk, and they already have vulnerable populations. Ideally, C19 patients should all have to use a special segregated hospital. And not just a ward. They share too many things. A whole physical location .

  34. jolhoeft says:

    To what extent do evolutionary pressures apply to the coronavirus? It seems to
    me that we have created a strong selection pressure on the virus to be less
    deadly. Areas with deadly outbreaks are getting locked down, reducing the
    virus’s ability to reproduce. A less deadly virus (say at a bad cold level),
    would trigger no lockdown and be able to spread through the population with
    little hindrance. This is among the space of viruses that trigger the same
    immune response. A mutation that triggered a different response would be a
    different “species”.

    Could we optimize the lockdowns for breeding a mild version that conveyed
    immunity to the deadly versions? I’d expect this to be a regular topic in
    epidemiology, but I’ve not yet seen discussion of it.

    • kronopath says:

      I can’t speak to this for coronavirus, but there’s a good TED talk about this which covers this for diseases like malaria.

    • fibio says:

      The general rule of thumb is that diseases become less deadly over time. This is both from a pressure not to kill off their host and a pressure on the host to increase their resistance. However, this is an average trend and there’s absolutely no reason that the random walk won’t spike up into high lethality before falling back down.

      Further, viral evolution tends to happen both slower and faster than we as the host would prefer. Ebola is still super lethal and there’s no sign that’s going away any time soon, meanwhile flu comes up with a new strain every twelve months. The current lock-down is merely a blip to the selective pressures on the virus, which is far more concerned about the immune system than anything at the macro-scale. Short of a vaccination campaign, it’s hard to imagine any policy can effect virus behavior in the short term.

      • matkoniecz says:

        I also thought so, but I never really verified it.

        And according to it may be untrue.

        I still have not verified it.

      • Emma_B says:

        “The general rule of thumb is that diseases become less deadly over time. ”
        The basic rule of thumb is that natural selection maximizes the number of copies of itself that a virus can produce when it is active in a host, i.e. before it has either killed the host or been eliminated by its immune system. This number depends both on the lifetime of the host, but also on the number of copies the virus can make per unit of time. If there is a correlation between the two, which seems common, then natural selection will lead to a certain virulence value, not necessarily a very low virulence.

        • fibio says:

          I’m not sure you can conflate virulence and lethality like that. Killing a host is bad for an infection. It dooms any viruses that haven’t gotten off the boat already, save for a few pathogens that remain transmissible even after death. There are many other factors that constrain a pathogen’s evolution, but it does impose some selective pressure on how things progress.

        • albatross11 says:

          Yeah, what the virus is selected for depends on how it’s transmitted.

          If making you so sick you just writhe in misery while mosquitos bite you works for transmission better than leaving you healthy, then the virus will be selected for the writhing in misery bit. If the virus transmits via your diarrhea and vomit contaminating the local evnironment or water supply, then it will be selected for maxing out your output, even if that means your electrolyte balance goes all wrong and you die.

          Most viruses are cleared by the immune system within a couple weeks of infection, including COVID-19. The virus is selected for whatever maxes out its transmission probability during the time before your immune system clears it, in the sense that the copies of the virus that infect the most new hosts will be the ones that become the majority of the ones circulating.

          • ChrisA says:

            This lady from Cambridge University says the virus is cleared within a few days, quote “nine days is nonsense. You don’t excrete a virus that long.” The article is well worth reading.

          • HeelBearCub says:

            That article doesn’t seem like it’s in line with other things we know.

            We already know that there is a long lead time between infection, first symptoms and ultimately severe symptoms. Something like 10 days to 2 weeks.

            If it were the case that people couldn’t be infectious 9 and 10 days after first becoming symptomatic, the hospitals wouldn’t be such a vector for spreading the contagion. Hospital staff wouldn’t be getting sick. We wouldn’t be worrying about aerosolization risks of respiratory therapy.

            At the very least the medical establishments treatment of all of those things suggest that this article isn’t in the mainstream as far as interpretation of the risks of infection.

      • Rick Jones says:

        SARS cov2 has changed 15 nucleotides since it was first isolated. 15 out of 30k! It doesn’t seem to be under much selective pressure right now.

    • albatross11 says:

      There’s been speculation that strong quarantines of sick people selects for strains that are aymptomatic, but I haven’t seen any evidence of this anywhere.

      Alternatively, one worrying possibility is that in a healthcare setting, the one strain that’s best at aerosol transmission is the one all the doctors and nurses bring home to the community.

  35. Aapje says:

    OMG, will these Dutch fixed expressions never run out?

    ‘Een appeltje met iemand te schillen hebben’ = Having to peel a little apple with someone

    Having a bone to pick with someone.

    ‘Spaans benauwd’ = Out of breath like a Spaniard

    Being fearful/anxious.

    ‘De lakens uitdelen’ = Distributing the blankets

    Being in charge.

    ‘Met de mond vol tanden’ = Having a mouth full of teeth

    Being speechless.

    ‘(praten over) koetjes en kalfjes’ = (talking about) cows and calves


    ‘Op je tandvlees lopen’ = Walking on your gums

    By the skin of your teeth. Supposedly, gums used to refer to the welt of a shoe, the connection between the sole and the upper. So if you were walking on the welt, the shoe was so worn that it was about to fall apart.

    ‘Komkommertijd’ = cucumber time

    The holiday period when nothing much happens. Cucumbers have a long growing season and don’t tolerate frost, so in Northern Europe they would fairly consistently be harvested in August/September, a common holiday period.

    ‘Wat heb ik nou aan mijn fiets hangen?’ = ‘What do I have hanging on my bike?’

    What is this weird thing (that’s going on)?

    ‘We zullen dat varkentje wel even wassen!’ = We will wash the piglet!

    We will solve that challenging problem right now.

    ‘Nu breekt mijn klomp!’ = Now my wooden shoe breaks!

    Something truly crazy just happened.

    • Reasoner says:

      Dutch people are the best.

    • DarkTigger says:

      ‘Een appeltje met iemand te schillen hebben’ = Having to peel a little apple with someone

      Having a bone to pick with someone.

      In Germany we say “Ein Hühnchen mit jemanden zu rupfen haben” = To pluck a chicken with somebody.
      All three languages have different idioms here, but all refering to food.

      ‘Op je tandvlees lopen’ = Walking on your gums

      By the skin of your teeth. Supposedly, gums used to refer to the welt of a shoe, the connection between the sole and the upper. So if you were walking on the welt, the shoe was so worn that it was about to fall apart.

      I doubt this. We use the same expression in germany “Auf dem Zahnfleisch gehen.” and I’m quite sure that no one refers to the welt of your shooes as “Zahnfleisch”.

      • Aapje says:

        I doubt this. We use the same expression in germany “Auf dem Zahnfleisch gehen.” and I’m quite sure that no one refers to the welt of your shooes as “Zahnfleisch”.

        That’s the etymological explanation I found from a book by F.A. Stoett from 1923, but most of these explanations seem rather speculative.

        Note that it’s seems quite possible that this expression migrated from Dutch to German and thereby lost its original meaning (just like no modern Dutch person will know that the welt was referred to as gums).

        This page claims that the first Dutch written reference was from 1858, but that the German references all date from the 1900’s. That would fit with a migration of the expression to German, decades after Dutch people used it.

        • Del Cotter says:

          The English version comes from the Geneva Bible of the 1500s, Book of Job, and is a direct translation from Hebrew. Nobody knows what the Hebrew expression meant, but it comes after a description of dangerous weight loss.

          • Aapje says:

            That may be a completely separate expression, though.

          • Del Cotter says:

            I see, yes.

            There is an expression “to be on your uppers” i.e. no sole left, but it does not imply your troubles are over yet.

      • Lambert says:

        ›Zahnfleisch‹ is the worst word in any language.

        Tooth-meat. It’s just a concept I don’t want to think about.

        • Del Cotter says:

          On the other hand, my Slovakian advisor introduced me to the word ›Sitzfleisch‹, “sitting meat”; he accused me of lacking it, which was fair. I later learned the American equivalent expression, “butt glue”.

          • Aapje says:

            We have the same word in Dutch (‘zitvlees’). It refers to the physical and mental ability to sit still.

            It can also be used (even more) figuratively, referring to occupying a job/position for a remarkably long time.

  36. Cheese says:

    “A very smart person” should probably know that there is reasonably good data on clinical characteristics of the disease and hospitalisations out of China and Italy.

    We know from those studies that 80% of detected cases are mild, ~20% are severe to critical. 5-10% require mechanical ventilation. Survival rates are hard to calculate overall, but we think we have about a 40-70% death rate for those requiring mechanical ventilation. Note that more than half of the severe cases are not requiring intubation. It also seems (early days though) in a health system that is able to stay on top of surge capacity (Germany, Australia, SEAsia), we’re getting 50%+ survival in ICU (not all ICU patients will be vented). It’s too early to start making statements on USA and UK clinical data because the one thing we do know about this is that the wean time from ventilation is really long.

    I’m not unsympathetic to arguments about economic cost and how badly this will impact society, I just really wish people would actually consider that there is a lot of information already out there that might inform their analysis.

    For example:

    “especially to ICU beds and ventilators (what percent of people over age 75 survive after going on a ventilator?), and working on getting people other treatments (oxygen, etc.) at home.”

    Vent triage and sending people home with mild-moderate pneumonia is already what is happening most of the worst affected countries today.

  37. matkoniecz says:

    80 or 90 percent of patients who end up on ventilators ultimately die

    How many people who need ventilators die without them? (I suspect that answer is nearly 100%)

    • HowardHolmes says:

      How many people who need ventilators die without them? (I suspect that answer is nearly 100%)

      If one defines “need” as “will die without ventilation,” then “duh!” If one defines “need” as “doctor decides it might help” then we have no way of knowing. It could be that everyone who survives would have survived anyway.

  38. L. Iguana says:

    As a person who has written an IRB, it’s a pain in the butt but is probably worth it. I think if you lacked the IRB it would be hard to get the study published and thus entered into the body of scientific knowledge. I think there’s an SSC post about how every culture has their process to decide that some thing or another is now real medicine – the IRB is part of ours. It might seem like overkill for just taking a vitamin, but doing it shows that the effort went in to double check that there isn’t some subset of people who have a terrible reaction to it. It would be a pity to get an interesting result that is never picked up and used to push knowledge forward.

    One potentially easier way to get an IRB is to see if there are any researchers already working in the area you want to study and ask if they would let you do your study as a subset of their ongoing research. You still might need to file some paperwork but it would be easier than starting from scratch.

    A final thought – you might not want to randomize on last name! Names for certain ethnicities tend to cluster around specific letters. A class I took recently had people split into small groups by name and we had all the people with African and Middle Eastern heritage clustered in the As and then all the Zhangs and Zhous in another cluster at the end of the alphabet.

    • Lambert says:

      Also the classic survey result ‘99% of Britons eat porrige for breakfast’.
      They opened the phonebook somewhere in the ‘Mc’ section and asked everyone on that page.

      • albatross11 says:

        I have the impression (I’ve only dealt with this peripherally, but know people who deal with it all the time) that IRBs vary *widely* in how painful they are to work with–some being stupidly picky over all kinds of dumb details, others being careful to make sure you’re not doing something unethical, but not so inclined to send you on endless paperwork-filling-out exercises.

    • silver_swift says:

      I think if you lacked the IRB it would be hard to get the study published and thus entered into the body of scientific knowledge.

      That assumes the purpose of the study is to get published and thus entered into the body of scientific knowledge.

      If you just want to do an informal study that tells you, personally, whether might have an impact on then it might be worthwhile to bypass as much red tape as you can get away with. If the results turn out to be big enough that you want to start thinking about publishing it, you can always do a second more formal study later (and you want to do that anyway, as “people subscribed to the SSC mailing list” is probably not a very uniform sampling of the population).

    • thisheavenlyconjugation says:

      I’m not sure “published in a peer-reviewed journal” is equivalent to “entered into the body of scientific knowledge” anymore. I assume you can get on bioRxiv without IRB approval.

    • Murphy says:

      Scott has a post about some previous bad experiences with that.

      the main problem seemed to be that his actual institution was badly set up for research and little institutional experience with it. As a result he kept encountering people who were incompetently diligent.

      I suspect if he’d gone through or partnered with any institution which routinely did research then it would have been trivial and fairly straightforward.

      Agreed re randomisation, names are awful.

  39. Meter says:

    I am thinking if it would be possible to combine samples of two of more people on an antibody detection test without impacting too much the false negatives. Does anyone know if there have been studies done on this?

    Being able to test two people with a single test, and being reasonably sure that either both are already immune and can return to normal life, or you can keep them in isolation knowing that there is a 2/3 chance that you are keeping an immune person from working, would mean you only need half the tests to be done. And so on for bigger groups of people.

    Obviously, the effectiveness of grouping also depends on the percentage of population that is now immune.

    As test production scales up, you could recombine test groups in smart ways to still need less than 1 test per person.

    • Chalid says:

      Yes, this has been done before for other diseases, apparently first by the US Army for syphilis tests in WWII, which were at the time expensive. I’m not sure why it’s not being done now.

    • Vitor says:

      Here’s a link of people claiming a method:

      I’m short on time right now, so I haven’t vetted this or looked into the details myself. honestly, I don’t even know what practical difficulties such a method would need to solve.

      On the theory side, there are obvious similarities to error-correcting codes and other information theory constructs. maximizing test yield is thus some kind of combinatorial optimization problem. No idea if someone has actually done that kind of theory work specifically related to pandemics.

      • Meter says:

        Thank you very much. This is exactly what I was asking about: pool testing. From your link, they have tested this on a sample of 50 people, and it seems mixing did not affect detection.

      • gdmclellan says:

        wrt the optimization problem, seems like basic query complexity.

        A test protocol that carries out t tests on a population of n people gives rise to a decision tree with up to 2^t distinct possible outputs. If there are v infected people in the population then there are (n choose v) which is Ω(n^v) possible infected subsets that the protocol could return, so the minimal amount of tests t which will detect all cases is Ω(v log n).

        If you can test up to n people with each test, then you can achieve that lower bound with a divide and conquer approach: split the population in half arbitrarily, then test each half of the population, then recurse on each half that tested positive.

        In practice pool testing probably doesn’t scale up beyond some pool size p reliably, so I guess you’d split the population into n/p pools and use the divide and conquer approach on each pool. If there are an expected v cases per pool then you expect to use O((nv/p)log p) tests.

        Edit: if there are v cases total, you can get O(n/p + v log p) tests by partitioning into n/p pools of p, then using the divide and conquer protocol on each pool that returns a positive test.

        • gdmclellan says:

          Oh, the lower bound was pretty crude — taking Ω(n^v) for (n choose v) only really works for small v. Ω(v log n) can be beaten for v of similar magnitude to n, e.g. obviously n tests always works. Feel free to push (n choose v) through a logarithm to get a finer lower bound. (And maybe there are correspondingly better protocols than I offered for cases of moderate-to-large outbreaks?)

  40. meltedcheesefondue says:

    I responded to Bean’s post arguing for a low risk for nuclear war, and thought the answer would be of interest to people here too:

    I don’t think Toby necessarily disagrees with you. You seem to be saying “the damage from nuclear war is overblown and not an X-risk” while Toby is saying “the damage from nuclear war is possibly overblown and likely not an X-risk”.

    I did an overview of the nuclear war/nuclear winter research a few years back, so I can summarise the main points we’re thinking of at the FHI (I don’t know if these are what Toby is relying on):

    First of all, there a big difference between dropping one nuke on a city while the rest of the country remains functional, and dropping one nuke on a city while the rest of the country is overwhelmed with mass casualties. Look at how poorly countries are coping with a comparatively tiny impact – covid-19 – and scale that up massively, have it all happen within a day, and wipe out most of the business, governmental, and commercial leadership.

    Secondly, the stockpile argument ignores what happens before and after the strike. It is very plausible that, as tensions mount, countries will re-activate and even extend their nuclear arsenals. So there are many scenarios where the arsenals will be far above 1,500. Also, the countries may maintain the ability to wage war even after a nuclear strike (the more “survivable” the first strike is, the more likely this is). Nuclear weapons can continue to be manufactured in many scenarios, as well as the required missiles. If the war drags on, more strikes will occur.

    And finally, nuclear winter. This is the great unknown. It all depends on how “well” cities burn, and how high black soot goes in the atmosphere. We’ve had two “natural experiments”: Hiroshima had a firestorm, Nagasaki did not. And modern nuclear weapons are very different from little boy and fat man, and modern cities are pretty different from WW2 Japanese cities. We’ve had some kinda relevant data – eg the mild outcome Saddam Hussein setting fire to the Kuwaiti oil wells is weak evidence pointing towards safety – but nothing strong. And ethics boards object to us nuking a city to get more info.

    So, if you said “nuclear winter is not very likely”, I agree with you, a thousand times. If you say “nuclear winter is very likely not to happen”, I disagree with you, a thousand times. We just don’t know, and nuclear winter is at least somewhat plausible given the evidence we have. It would overconfident to put its plausibility below, say, 10%, which is certainly enough to worry about.

  41. johan_larson says:

    Some good news about COVID-19 here in Canada. The province of British Columbia, just north of Washington State, has achieved linearity: the number of new cases per day has been flat at 50-60 for several weeks now. Of course, that means the total number of cases is still increasing, but at least it isn’t accelerating.

    Unfortunately that’s not true yet for the largest provinces, Ontario and Quebec.

    • Edward Scizorhands says:

      Not sure if someone brought this up already, but:

      In Northern Italy, 60 volunteers who thought they’d never suffered COVID-19 gave blood. 40 of them tested positive for antibodies [maybe antigens?] to the virus.

      I’m skeptical because it’s an unknown (to me) foreign news source being translated for me, and Northern Italy has been a horror show so it might not be something we want to replicate elsewhere. But still some hope we might escape this.

      • Loriot says:

        A county in Colorado is doing large-scale antibody tests for COVID-19.

        They’re finding that less than 1% have been infected.

        The theory that most (or many) people have already had coronavirus is a wishful-thinking fantasy.

        • albatross11 says:

          I would not be shocked to see very different results from, say, NYC or Seattle or SF. But yeah, so far, the evidence for “we’ve all already had it and didn’t notice” is not looking so good.

        • John Schilling says:

          That’s a rural county fairly isolated from the current centers of the outbreak; I woudn’t put too much weight on it.

        • Chalid says:

          I think the better evidence is South Korea. If the virus was vastly more widespread than we think, then either Korea would not have been able to contain it or Korea would have have observed a much lower IFR.

        • Edward Scizorhands says:

          I understand there’s a meme of “we all got this a month or two ago, stop worrying.” That meme is dumb and dangerous, and I get why my comment pattern-matched to it. But that was not the intent of my comment.

          Northern Italy has had the virus run amok. It might actually be true that they had it recently and have achieved, whether through luck or stupidity, decent herd immunity.

          • Rebecca Friedman says:


            I glanced at the original; most of it is paywalled, but the first paragraph isn’t, and a lot of the data is there. Castiglione d’Adda, the town the people in question were from, is in Lombardy. According to a governmental report*, as of March 12th Lombardy had a 63% infection rate. Italy wasn’t testing a lot of asymptomatic people, and I looked at that and realized that if the asymptomatic rate was around 50%, all of Lombardy had it.

            Presumably an overstatement; the asymptomatic rate in Lombardy must have been lower than I expected (possibly as a result of higher exposure, if higher exposure does lead to worse cases), or else they tested more asymptomatic people. The article only says 40 out of 60 tested positive, which is frankly fewer than I would have naively expected.

            But yeah, it seems perfectly plausible to me. Note 40/60 is specifically from Castiglione d’Adda; it isn’t clear from the first paragraph if that’s the only place samples were collected and tested from, or just the most striking.

            * if anyone is curious, though rather out of date by now.)

          • Douglas Knight says:

            According to a governmental report*, as of March 12th Lombardy had a 63% infection rate.

            Try reading that report again.

          • Rebecca Friedman says:

            You’re right and I’m an idiot.

            That’s percentage of cases that are in Lombardy, not percentage of Lombardy’s population infected.

            Thank you. That’s much better.

      • somatochlora says:

        Castiglione d’Adda has a population of 4659 per Wikipedia.
        As of April 1, the death toll for the town was 80. Can’t find a way to send this link such that it is both unpaywalled and translated, but here:

        This town has had a death rate of 1.7% so far. Not sure why this story is being spread as if it says something about parts of the world that have not seen significant deaths.

      • zzzzort says:

        The slowdown in cases seems to match really well with the onset of the lockdown, which implies they haven’t yet achieved herd immunity, although maybe they were on their way. Still, something like a 0.1% of the population of Lombardy died of coronavirus, so that’s still a pretty bad worst case scenario.

      • LesHapablap says:

        The imperial college people estimated that on March 28 Italy had 5MM-20MM infected (of population of 60MM)

        So they shouldn’t be far off herd immunity today

  42. theredsheep says:

    One day in old Constantinople, a pair of master surgeons get into an argument about the correct way to create a eunuch. One insists that it’s done by severing the penis, while the other says properly you ought to remove the testicles. Medieval and ancient sources are ambiguous and contradictory, and they wind up going back and forth for an hour. At last, one of them decides he’s sick of the argument and tries to conciliate. “After all,” he says, “penis, testicles–the end result is much the same from our point of view, isn’t it? For practical purposes.”

    “Oh, please,” the other scoffs. “There’s a vas deferens between the two!”

    • Bobobob says:

      80-year-old man in the car with the radio on. On the news, the announcer says, “We have reports of a driver on I-40 heading in the wrong direction.”

      One driver?” the man exclaims. “There are hundreds of them!”

    • GearRatio says:

      I wrote almost this exact joke for my urologist when I got my sperms permanently quarantined.

  43. Murphy says:

    the UK’s ICNARC have released data on demographics of covid cases in UK hospitals comparing COVID to influenza cases in the last few years.

    Some of it may be skewed by the locations with the most cases but if it really is hitting people of asian and african descent harder in the UK then it’s not good news for a bunch of countries.

    There’s a non-trivial chance that it’s related to the different haplotypes of the genes coding for the ACE2 receptor.

    The East Asian populations have much higher AFs in the eQTL variants associated with higher ACE2 expression in tissues (Fig. 1c), which may suggest different susceptibility or response to 2019-nCoV/SARS-CoV-2 from different populations under the similar conditions.

    • J Mann says:

      My wife was looking at a story on the number of African-Americans hit in Michigan. I assume the effect was likely that Detroit’s urban areas got hit harder and earlier than the suburbs, or maybe health cofounders related to wealth disparities, but if there’s a genetic component, we’re going to get conspiracy theories for years longer than currently predicted.

    • EchoChaos says:

      This could also explain the data in Germany. It’s not that Germany did way, way better, just that Germans are less susceptible.

      Someone had a differential data in Switzerland by canton based on native language and Germanic Switzerland is doing the best of all the Swiss cantons.

      If true then we should expect the American West to do quite well, the American south to do well among whites and very poorly among blacks.

      • Murphy says:

        I’d be cautious about making strong predictions like that. The haplotypes likely involved seem to be super-common in all populations so the difference would be more similar to some populations having 2/3rds as many serious cases so hospitals would struggle anywhere.

        If hospitals have 10X as many serious cases as they can handle it’s not much better if in some regions it’s 6X

        • EchoChaos says:

          That was random rambling hypothesizing, not any sort of serious epidemiology.

          Please do not alter policy based on said random internet post.

          But given where we are seeing spread now and where is getting hit VERY hard versus relatively hard, this seems at least plausible.

          If Southern American whites (a notoriously physically unfit and relatively old subpopulation) do surprisingly well and the American West continues to have basically flat curves, then this will be confirmed.

          This doesn’t mean that Germans, German-Americans and their related subpopulations shouldn’t continue to socially distance or anything.

          • Joshua Hedlund says:

            It’s still very early, but the higher death count of the less-populated yet more-white Palm Beach county compared to the higher-populated yet less- white Miami-Dade county in Florida would not seem to support that hypothesis. (My hunch is Age is a simple factor swamping others, but haven’t looked into details, caveats of sample sizes anyway, etc)

          • EchoChaos says:

            @Joshua Hedlund

            Still early enough to not be sure, and isn’t Palm Beach mostly Northeasterners with a winter home? In that case, they would be substantially less Germanic than the rest of the South (Borderers) and West (Borderers and Puritans).

            Remember that it’s specifically Germanness, not whiteness that we’re talking about here. The Northeast is heavily Italian/Irish/Polish/etc.

            If WASPs are getting hit less hard (e.g. Rhode Island doing better than neighboring Connecticut), that would argue for this.

          • m.alex.matt says:

            EchoChaos: Two things,

            1. Please, for the love of God, do not take categories from Albion’s Seed as a genuine guide to the genetic background of different populations in different areas. The whole thesis of the work is that founding cultures survive ethnic change, anyway, but even then founding cultural categories are very broad-brushes that are not going to easily map onto any kind of population genetics.

            2. To the extent they are a guide to anything, the Borderers are not significantly more Germanic than their later immigrating Irish cousins. They’re heavily Celtic enough to make the distinction fairly worthless. Both the England/Scotland border area and the island of Ireland have seen enough admixture over the millennia that I find it hard to believe that both populations aren’t very mixed. Especially those of Irish heritage who have lived in New England for more than a century and are thoroughly mixed with local Yankee lineages.

            Oh, also: The West isn’t particularly ‘Puritan’, except the Mormons themselves. If anything, it’s more straight German or, in areas, Scandinavian.

          • EchoChaos says:


            1. Please, for the love of God, do not take categories from Albion’s Seed as a genuine guide to the genetic background of different populations in different areas.

            I am not, and as I said, this is WILDLY armchair-ish.

            However, I will push back and say that the Scots-Irish are substantially less Celtic and more Germanic than are the Scottish or the Irish. They are mostly genetically an English strain. This can be seen by 23 and Me tests from the South.

            There are notable genetic differences between the English and the rest of the British Isles that persist to this day. 1500 years just isn’t that long in human population terms when there isn’t mass-scale migration.

            The West isn’t particularly ‘Puritan’, except the Mormons themselves. If anything, it’s more straight German or, in areas, Scandinavian.

            Yeah, which is why I said the American West would do so well in my original post. I should’ve included “German” in my previous post, you’re right.

          • Le Maistre Chat says:

            However, I will push back and say that the Scots-Irish are substantially less Celtic and more Germanic than are the Scottish or the Irish.

            Which Scottish? Scots is a Germanic language, while the Highlanders are Gaelic. Culturally the former were stereotyped as stolid, boring Presbyterians when the latter were still rising up for the Stuarts.

          • EchoChaos says:

            @Le Maistre Chat

            Which Scottish? Scots is a Germanic language, while the Highlanders are Gaelic. Culturally the former were stereotyped as stolid, boring Presbyterians when the latter were still rising up for the Stuarts.

            I assumed @m.alex.matt was talking about Highland Scots, because he compared them to the Irish, which are indeed a more Celtic populace.

        • Gerry Quinn says:

          What you say is not untrue and there are other confounders (maybe the climate will have a big effect on some countries; population density and social behaviour certainly will.)

          But the ratio differences between covid and viral pneumonia (2017) in that imgur were quite large. (Could still have tangential effects built into them too, of course.)

      • Lurker says:

        This could also explain the data in Germany. It’s not that Germany did way, way better, just that Germans are less susceptible.

        Shouldn’t the regions bordering on Germany be doing quite well in this case? One of the hardest hit parts (Edit: maybe the worst, actually) of France borders on Germany and if you go back in history far enough actually was part of Germany once. There are a lot of people “living across the border” in that area (meaning that they socialize across the border, have relationships, children, work, etc), so I’d expect a random sample on the French side of the border to be genetically indistinct from a random sample on the German side of things.
        I really don’t think this theory stands up to a common sense check, but maybe I missed something.

        [Disclaimer: I’m German, so I’d actually really, really wouldn’t mind this being true, since most of the people I’m worried about are German to, but I just don’t think it fits. I think we’ve got better numbers through a mixture of methods and luck (which still might run out), not genetics]

        • EchoChaos says:

          Shouldn’t the regions bordering on Germany be doing quite well in this case?

          They are in Switzerland, which is why I brought this up.

          One of the hardest hit parts (Edit: maybe the worst, actually) of France borders on Germany and if you go back in history far enough actually was part of Germany once.

          This is the sort of on the ground info that I am interested in, thanks! That does indeed punch a pretty hard hole in it. Is that region of Germany also hit hard?

          I’m German, so I’d actually really, really wouldn’t mind this being true, since most of the people I’m worried about are German to, but I just don’t think it fits. I think we’ve got better numbers through a mixture of methods and luck (which still might run out), not genetics

          I’m old-blood Anglo-American, so it would maybe help me, but since I have African-American family (and AAs are by far the worst hit so far), I’m definitely not doing this out of any sort of “gotcha”, I am genuinely curious.

          • Lurker says:

            They are in Switzerland, which is why I brought this up.

            I suspect that’s due to geography/the shape of the Alps – in some parts it’s definitely easier to reach/cross the borders than to go to another part of Switzerland, but I don’t know how pervasive that is. Anybody Swiss/living in Switzerland here who can chime in?

            Is that region of Germany also hit hard?

            Sort of? The states with the most cases (in absolute numbers) are also the ones with the highest population and one of them borders on France.
            Here are some maps: (I hope the link goes through…)
            It’s all in German, but it’s mostly maps/pictures, so I hope it’s ok.
            The upper map is cases per 100 000 people, the lower map is colored in in absolute number of cases.
            If you look at the lower map, you basically also get a rough map of population numbers.
            The parts of Germany that have the darkest color on those maps, are doing well enough to accept critical cases from the overloaded French hospitals (currently, at least), so:
            hit hard? yes. But a lot of the early cases came from people coming back from (ski-)vacations, who’re usually fairly young and fit and a lot of the community spread came from carnival(Fasching) and church events – all of which are more likely in the southern (and western) part of Germany. East Germany has pretty low population density (aside from a couple big cities) and tends to be poorer and less religious. Add to that the longer distance to the ski resorts that caused a lot of the early cases, and it makes a lot of sense for them to have fewer cases in absolute numbers as well as per capita.

            I’m definitely not doing this out of any sort of “gotcha”, I am genuinely curious.

            Oh, no! I didn’t think you were. I’m actually wondering about the difference in death rates as well, but I threw out the “could be genetic” line of reasoning pretty early for the reasons above.
            But please keep asking/trying to poke holes in this. I’d rather get as close as possible to the truth than be right. I’m happy to go look for more in depth information, as well.
            What’s been bothering me a lot about this entire crisis is the lack of data 🙁

          • EchoChaos says:


            So I dug into the French cases more, and I am no longer sure that Grand Est is a good counterargument. Yes, part of it is/used to be Germany, but it also contains several cities with substantial African minority immigrants, especially Strasbourg.

            This is all “speculative internet research”, but since France doesn’t collect demographic data on deaths, that’s all I have to go on.

            I totally agree with how frustrating the lack of data is.

            I don’t think we can get rid of a genetic conclusion yet, the differential between blacks and whites in the United States is so absurdly large that something has to be there. It could be lifestyle, of course, but we have inner city poor whites here too. White cities like Pittsburgh aren’t getting hammered like black cities like Detroit are.

          • Lurker says:

            I don’t think we can get rid of a genetic conclusion yet, the differential between blacks and whites in the United States is so absurdly large that something has to be there.

            No, I agree that we can’t get rid of the “genetics make a difference” theory. I just think that “German ancestry* vs. other white people ancestry make a difference” is one that doesn’t work very well.

            In general, my personal pet theory is that this is one of those rare situations where small social/behavioural differences between cultures have a really disproportionate impact, but I’m not really sure how to go about falsifying this.

            *I have no idea how to phrase this properly, but I hope the point is still clear. I’m also really, really skeptical, if it’s even possible to be genetically German/French/Italian/etc. Anything with higher resolution than Europe/Africa/Asia seems to be assuming a lot.
            (Disclaimer: Physicist here, I think biology is interesting, but I’m definitely a layman and it’s not something I’ve dug deep on.)

          • EchoChaos says:


            I have no idea how to phrase this properly, but I hope the point is still clear. I’m also really, really skeptical, if it’s even possible to be genetically German/French/Italian/etc. Anything with higher resolution than Europe/Africa/Asia seems to be assuming a lot.

            This is one where it seems pretty clear that “it depends on what you mean” is a big caveat.

            Obviously they are more closely related than other places, but there are notable genetic differences between them, which is why we can tell them apart visually.

            It’s a 20th Century thing that there is substantially inter-country intermarriage and travel. Lots of places in Europe were genetically relatively isolated from each other even when geographically close.

            It’s not at the level of the Sahara desert, obviously, but the other thing is that we simply haven’t had enough generations to have any sort of substantial genetic change of our current mixed mode.

          • ana53294 says:

            There is a certain German phenotype, as well as an Italian one, Greek one, etc. For each of those countries, people who aren’t of mixed origin can be visually distinguished from people of a different country.

            Of course, the more mixed an area is, the harder it gets to see where people are from. But if people care enough about it, they learn to detect drops of some phenotype they’re looking for. Like people who are able to see how much Jewish or African ancestry a particular white-looking person has. Some people detect a difference between somebody who is a 1/16th African, rest white and somebody who is pure white.

            I never cared particularly about those differences, but I am able to tell where in Spain somebody is from, as long as they aren’t mixed.

            Phenotypic heritable differences in appearance are presumably due to genetic differences. Whether those slight differences are enough to make differences in resistance to a disease, I don’t know. Humans are very good at distinguishing faces, so maybe the differences we see are not sufficiently big.

            So while Italian/French/German/Russian are not races (I’m not even sure how much we can talk about biological human races when talking about blacks/whites), they certainly each have a certain number of characteristic features that distinguishes them from others.

          • Lurker says:

            Well, I know people who claim to be able to do that, but I’ve never been able to at all – I can guess at the continent that recent ancestors have probably come from and that’s about it.
            Typical mind fallacy, I suppose 🙂

          • EchoChaos says:


            I lived in Korea for two years as a kid on an army base, and I can reliably tell the difference between Chinese, Japanese and Koreans, but I’m hopeless with Southeast Asians.

            Whites I can tell broadly (e.g. an Italian from a Swede), but not with the accuracy that my friends who grew up on army bases in Europe.

            I think it’s a childhood learned thing.

    • Del Cotter says:

      I’m pretty sure it’s location. That will be London, higher density, skewing younger and dumber, skewing less conscientious about the people around them. They’re filling the parks in good weather and using the crowded transport network.

      I’m surprised they’re willing to break down by ethnicity and not location: they ought to know how misleading that can be. “Black”+”South Asian” is basically being a proxy for “Lives in London”. I could be wrong, but that’s how I’m reading it.

      • Murphy says:

        they did both, not including that was my sin:

        Looking at the demographics, it looks like part of it is explained by London demographics but it’s a bit more pronounced than just demographics but the numbers are small enough for the difference to possibly be noise.

    • keaswaran says:

      It seems to me that the natural prior should be that social variations among people will lead to greater variation of infection rate than physical variations, because that’s how it works for most diseases. (Sickle cell and Tays-Sachs are notable as special exceptions; for most other diseases the bigger question is who you have sex with or how often you share blood or air or go into mosquito-infested jungles.)

      It also seems to me that the natural prior should be that racial groups differ more in their social behavior than in their physical features. (People of different races tend to have different levels of income, different specific types of jobs, different media channels they watch, different religious institutions they go to, and different personal social behaviors even down to the level of how far apart they stand when having a conversation.)

      Putting both of these together, it seems likely that more of the racial variance around infection rates will be explained by these social factors associated with race than with physical factors.

      • EchoChaos says:

        This is a genuine question, but are the Swiss more culturally similar to other Swiss or to their co-lingual neighbors across the border?

        My understanding is that the impact on French-speaking Switzerland is more similar to France than to Germanic speaking Switzerland.

        This seems a natural test case.

        • HeelBearCub says:

          The Swiss (at least the 40 years ago Swiss, but I wouldn’t have expected this to change) have a fairly unique culture as a whole, but each of the regions also have their own specific culture. Basel and Geneva are different kinds of places, but the Swiss train conductors all universally tried to time their signal to the engineer so that the train began visibly moving precisely when the sweep second hand of the station clock reached the exact moment of scheduled departure.

          I would imagine that a confounder here is that I believe it’s easier to get from Geneva to the French towns just over the border than it is to get to other regions of Switzerland.

    • Jaskologist says:

      We had a link recently indicating pretty significant differences in susceptibility based on a person’s blood type (B is more at risk, O is less at risk). That alone would be enough to make a big difference between ethnicities, since blood type proportions vary a lot.

  44. Hysteria says:

    Here in Brazil things are going pretty badly regarding Covid19.

    Not even two weeks of quarantine and people “are sick” of it already; store owners figured out that they can stay open and keep all workers working if they “control the amount of people inside of the store”, thus evading the government orders against gatherings.
    End result is that people will be crowded in long queues on the sidewalks while a couple clients go inside to do their business. (Sidewalks being public and thus not the store’s responsibility)

    People also got sick of not being able to go outside to enjoy parks/smoke/buy street food so they’re just going out and doing it anyway; parks in particular were very crowded on this weekend, with families playing around and old people chilling under the sun.

    I mean, what could go wrong?

    • Nick says:

      Not counting on it, but I really hope name checks out. 🙁

      • Hysteria says:


        Brazil as a country prepared pretty readily for the pandemic; PPE production got ramped up; the government decided on quarantine measures comparatively earlier than other countries and there was (and is) extensive media campaigns to make people stay home and follow safe behaviors.

        But none of that matters if people at large just decide to do whatever. The average brazillian’s trust on media (or ability to understand what’s going on) is pretty low, so everyone from store owners to the poor worker is pushing the limits of what they can and can’t do “under quarantine laws” and engaging on riskier behaviors just to restore the “normalcy” of day to day life.

        We are also not testing nearly enough cases because we lack the lab equipment to do so, and the government has extremely narrow requirements to allow testing. (Most people get a CT scam and if it shows a possible infection on the lungs, they’re sent home for 14 days, not everyone gets the official test and becomes a statistic.)

        • Robin says:

          Thank you for this first-hand information. Over here, the newspapers only write that the Brazilian president does not believe in Covid-19, says that it is just a grippezinha, Brazil must not stand still, and does everything to undermine what regional governors try to do, and the favela gangs force a lockdown.
          Good to hear that this is not the entire truth and the government is doing something, after all.

          About following the quarantine, or being sick of that, it’s similar in Ecuador, especially in Guayaquil it is bad. And people are still protesting against the closing down of the markets. Sure, many poor people need to make their living every day and cannot afford to lock down. And then there are those who don’t care.

    • Radu Floricica says:

      Here big chain stores were pretty fast to adapt. There are queues outside most of them, but they put markers on the ground every 5 feet, and people seem to respect them. Same for the queue before paying. Not sure how mom and pop stores are doing, because most of them closed down years ago. Yey for globalization, I guess.

      • Del Cotter says:

        Only 1.5 metres? In this neighbourhood the stores are marking every 2 metres (6.5 feet).

    • Gerry Quinn says:

      Deaths are scaling up which may concentrate minds a bit. But the psychological phenomenon is probably there in most Western countries. Police in Ireland and the UK have been having to clamp down a bit, especially people going to parks etc.

    • Viliam says:

      End result is that people will be crowded in long queues on the sidewalks while a couple clients go inside to do their business.

      Staying on fresh air, even close to many people, seems safer than being in a closed space with the same amount of people. Think about the concentration of viruses per cubic meter (outside, the wind can blow them far away and up in the sky), and the places where they can land (who cares if the tree leaves are covered with COVID-19).

      • albatross11 says:


        It sure seems to me like being outside in a not-crowded place is better in every way for avoiding spread than being inside at the same density of people. The air circulation is better, there’s likely sunlight to eventually inactivate viruses and a breeze to avoid high concentrations of virus in air, etc.

  45. bean says:

    Biweekly Naval Gazing links post:

    The big activity this time has been on the merchant ship front, where I’ve looked first at the old general cargo ships, and then at the container ships which replaced them, massively reducing shipping costs and enabling today’s global economy.

    Reader Suvorov has looked into the checkered history of Southern privateers during the American Civil War.

    Lastly, my wife recently paid a visit to the SS Anne, and shared her experiences.

    • ChrisA says:

      Container ships and containers in general are another good example of a technology that seems to have been invented far later than it should have been. I would imagine they would have been really useful during the second world war for instance. Any thoughts on why they were so late?

      • Bobobob says:

        This reminds me of a book I once saw on display in Barnes & Noble that made me think the whole “X: Why the invention of X was indispensable to the modern world” craze had run its course:

        The Box: How the Shipping Container Made the World Smaller and the World Economy Bigger

        On the other hand, the second edition has almost 400 reviews on Amazon, so the topic may be of wider interest than I first imagined.

        • Evan Þ says:

          I’ve read that book; it’s very good. Apparently containerization was a huge impetus to the modern globalized economy, by drastically decreasing shipping costs of just about everything.

        • keaswaran says:

          It’s a great book. I read another book about shipping containers at around the same time (Ninety Percent of Everything, by Rose George) and I have trouble remembering which facts I got from each book independently and which were in both. But I think it’s much more foundational to the world you think you know than you realize.

          (One of the unexpected points – why do you think every urban area had a renaissance of its waterfront in the last couple decades? One part of it is all the clean water acts that made the rivers into pleasant things to look at rather than sources of pollution, but another part is that container ships need much bigger land areas to load and unload, so that shipping, and the industry that locates next to it, left the cities, so that industrial pollution of the 50s and 60s could give way to industrial abandonment and beginnings of water cleanup in the 70s and 80s, and then artist followed by residential occupation in the 90s and 00s.)

          • Bobobob says:

            The fact that there are not one, but two, books about shipping containers does in fact convince me that they are more important to the world than I realized.

      • Del Cotter says:

        IIRC containerisation kind of did get its start in WWII, particularly the part where you keep the ships leaving on schedule and do not delay departure for the sake of filling hold space. It turned out to be more important that they keep transporting, than that they be “efficiently” full. If convenient, you could throw in junk you wouldn’t normally prioritise, on the off chance they could use it at the other end. Late deliveries could just go on the next ship, because you could rely on their being a next ship, since they were shuttling regularly.

      • matthewravery says:

        How is the technology of a shipping container fundamentally different from the technology of a wooden barrel? Volume and material, sure, but weren’t people shipping barrels of (relatively) consistent size for centuries? Merchant ships were built to ship barrels of goods. Systems were developed for efficiently loading and unloading barrels. What am I missing? (Probably a lot! Please enlighten me.)

        • bean says:

          One big issue is the tare weight. An ISO container can hold 5 to 10 times its weight in cargo. A barrel usually can’t, unless it’s cargo is high-density. So packing, say, clothes in a barrel isn’t going to work well, particularly as barrels also stack very poorly. They’re also not an atypical size for break-bulk cargo, so you don’t really gain much in efficiency.

          The other issue is one of scale. It only costs a little bit more to move a container on and off a ship than it does a barrel. You need a couple of guys to hook it up, and one to run the crane. But each container holds a lot more, so the total amount of work required is vastly reduced.

          • Garrett says:

            Containers can also be locked and sealed which makes it less convenient for longshoremen to pilfer.

        • Del Cotter says:

          Palletization, that was the word I was missing. Here’s a history site:

          I suspect barrels are a nightmare we civilians don’t appreciate from outside. At a guess, they’re inconvenient for powered vehicles to lift, hard for humans to lift, all too easy to start rolling, easy to be injured around, easy to come loose in a ship. I’m re-reading The Pride of Chanur and “canisters on the dock” hefted and stacked by the spaceship crew has not aged well.

          • Kindly says:

            Barrels are round precisely because we want them to roll, since they’re hard for humans to lift. This is less relevant when there are cranes.

          • Del Cotter says:

            No, it’s for structural integrity; a rectangular hoop would not be able to serve its function. Rolling them is handy (until they crush you) but as an affordance, roll-ability is more what Stephen Jay Gould called a “spandrel”, something that just happens to exist because the essential features bring it into being, while not being an essential feature itself.

            The existence of containers you can’t roll never did them any harm. You roll a barrel because you can, but it’s not why you made a barrel a barrel.

          • zardoz says:

            Barrels are a really old technology that goes back to ancient Rome, or possibly even earlier. They’re round because they’re basically a bunch of planks held in by metal hoops, right? You can’t really make a square one with ancient Roman tech (or medieval tech for that matter).

          • Del Cotter says:

            The oldest ones were made with wooden hoops. The Wikipedia page describes one of the uses of second-hand wine barrels as lining for wells, which is the source of many surviving ancient barrels, since old barrels would usually otherwise be burned or broken up (same problem as with old papyrus documents we wish had survived). That page gives as an example the one now about 900 yards from me, from Silchester.

            Clay amphorae for shipping (and clay tablets for writing) don’t have the problem of being valuable fuel sources when their primary use is over, so they are much more common (don’t rot either)

        • zzzzort says:

          The scale and concomitant movement to a capital intensive crane instead of individual longshoremen moving barrels one at a time is the major factor.

          There were a lot of smaller advantages, like being able to load one shipping container on a semi, and having the containers be weathertight enough that they can be shipped outside of a hold. Containers stack really well, both because they’re nice rectangles but also because they’re strong enough to support it.

        • Lambert says:

          It’s much harder for dwarves to smuggle themselves in shipping containers.

      • bean says:

        I’d say it’s a mix of cost and lock-in. Labor costs rose a lot post-WWII, which made the system more viable. And container ships can’t use ports which aren’t equipped for it, which is a serious problem for independent shipowners. If you’re a big enough customer, you can make that work, but a lot of them were always kind of financially marginal, particularly with the depression. Add in the post-WWI shipping glut, and you have a case where nobody is willing to pay to set up the system.

        There were some steps towards containers during WWII, with extensive use of pallets and the like. But the vast majority of the shipping stock had old cargo-handling gear, and they were spending a bunch of time offloading cargo in places that didn’t have good facilities. And the specter of a ship showing up with containers and being able to unload them would not please logistics planners.

      • cassander says:

        the container needs a few conditions things before it’s useful. Most importantly reliable, high speed cranes and other container moving infrastructure combined with labor costs high enough to make the investment in them worthwhile. I’m not an expert, but I’d guess that electric motors for cranes powerful enough to lift full containers quickly weren’t a thing until the mid-20th century.

        • CatCube says:

          I think the electric motors capable of handling them existed, but they were very expensive and you economized on the number of motors.

          I look at a dam we constructed in the 1930s, Bonneville, and they used two cranes to handle all 18 spillway gates. This required much more labor for riggers and crane operators, but minimized the (relatively) very, very expensive capital costs for the hoisting equipment.

          Twenty years later, when building further dams in the late 40s to the 60s, we gave each gate its own hoist so now you don’t have to rig at all, just push a button. Much less efficient in the use of equipment, since each motor is now being used for a tiny fraction of the time, but much more personnel-efficient.

          We went back in the 70s to give each gate at Bonneville its own hoist, so they could be operated from the control room, though this was as much for fish passage purposes.

          • ChrisA says:

            I remember seeing pictures of cranes lifting stuff out of holds of ships, including cars etc from the 1930s. It looked like they were rigging each item specially, and it must have taken a long time to unload each ship. They must have had cranes that could lift large loads though since that is necessary in both container and non-container systems. My guess is the coordination issue, one port doing this wouldn’t be enough for the entire system to move to containerisation.

    • Procrastinating Prepper says:

      The SS Anne? Did she fight Gary and obtain the Cut HM? /s

  46. sjcatchpole says:

    Out of curiosity, has anyone looked at whether not having had a tonsillectomy plays a role in defense against Covid-19? Twigged by:
    Tonsil and adenoid removal associated with respiratory, allergic and infectious disease
    Date: June 7, 2018
    Source: University of Melbourne
    Summary: Removing tonsils and adenoids in childhood increases the long-term risk of respiratory, allergic and infectious diseases, according to researchers who have examined — for the first time — the long-term effects of the operations.
    linked here

    and various data on regional and temporal differences in the rate of removal including

    Large international differences in (adeno)tonsillectomy rates. and

    [Epidemiology of tonsillectomy and/or adenoidectomy in Italy]. “We observed a wide geographical variability of regional tonsillectomy rates, standardised by age and sex. They ranged from 3.5 x 10.000 (CI 3.1-4.0) in Basilicata to 19.0 (CI 18.6-19.5) in Piemonte. We found an inverse correlation (r = -0.50) between regional tonsillectomy rates for tonsillitis and minimum temperatures recorded in capitals of the regions.”

    Compared with this data on Covid-19 deaths in Italy – where Piedmont is #3 on the list with 1,168 deaths, and Basilicata is very last on the list with 13 (notwithstanding other obvious potential factors like proximity to the original outbreak)

    • George3d6 says:

      I would assume you can de-confound this data from anything else.

      Deconfounding the effects of tonsils themselves is hard since the people that have them removed are going to have been a less-healthy demographic by default to being with (though for many it might have corrected as they reached adulthood).

      Now deconfounding that from other traits that encourage a region to “prescribe a lot of tonsillectomies” and just the random noise that goes in CFR in a given region… yeah, no, that’s impossible.

      The only way to answer this question is to look at individual patient data, but because we live in a Bizzaro-land none of the Western countries (or any country really) will publish in-depth patient histories, even when said patients already died, even when it can save millions of life… so, eh.

      • sjcatchpole says:

        Based on other reading (admittedly starting from the tonsillectomy Wikipedia page) it seems like rates are somewhat arbitrary by place and over time, and determined somewhat by health care funding structures (if you pay per surgery rather than fixed salary) and different clinical standard (it was really common in the uk and parts of the us 80 years ago, less so by the 60s and had fallen out of favour by 40 years ago – so in line with the death curve for covid-19)

        Agree without case data this would be hard but was curious if anyone knows whether people are making that sort of data available (or analysing it)

    • keaswaran says:

      Basilicata is one of the southernmost regions and Piemonte is one of the northernmost regions. There’s lots of things that run on a north-to-south gradient in Italy. What we think of as one country, Italy, is actually in many ways more naturally thought of as a project of Piemonte occupying the south (and it was doing so under the rulership of a family from Savoie, much of which is now in France or Switzerland).

      If you could find variance within the south on these variables, and variance within the north on these variables, and correlations within each region, then maybe there would be good evidence. But if it’s all just correlation that results from two separate correlations with the north-south gradient, then it doesn’t tell us much.

      • Le Maistre Chat says:

        What we think of as one country, Italy, is actually in many ways more naturally thought of as a project of Piemonte occupying the south (and it was doing so under the rulership of a family from Savoie, much of which is now in France or Switzerland).

        Italy is a good example of how weird nationalism can be. It had an organic element (the language everyone had to learn in national primary school was the language of Dante, not something made up the week before), but it absolutely was a mid-19th century expansionist project by the King of Sardinia and Savoie/Piedmont (a Southeast Corner of Francophonie).
        During the French Revolutionary and Napoleonic era, the Rightist philosopher Joseph de Maistre was a Francophone Savoyard noble whose family had been ennobled from the ranks of cloth merchants by the king of that oddly-shaped monarchy.

      • sjcatchpole says:

        This was just the first set of tonsil data that had some solid geographic spread – agree Italy might have plenty of other data.

        Based on other reading (admittedly starting from the tonsillectomy Wikipedia page) it seems like rates are somewhat arbitrary by place and over time (even within countries), and determined somewhat by health care funding structures (if you pay per surgery rather than fixed salary) and different clinical standard (it was really common in the uk and parts of the us 80 years ago, less so by the 60s and had fallen out of favour by 40 years ago – so in line with the death curve for covid-19)

        Just seemed like there is a big unknown which is why deaths seem to vary so much by age cohort, and to be geographically patchy (although lots of drivers for this likely related to how epidemics spread)

        Would love to see US regional data for the tonsillectomy – one article was saying that back in the 20s and 30s in New York it was more common than not to have them out…

    • sjcatchpole says:

      Another point on this – it is something that old people have in common that younger people don’t in the US:

      Year US Tonsillectomy Rate per 10,000 population (under-15s)
      1965 165.5
      1970 124.7
      1975 86.5
      1980 57.4
      1985 37.8

  47. benquo says:

    IRBs are a way for institutions to cover their asses. If you’re not relying on an institution that this would expose to perceived reputational risk, the realistic worst case scenario is that randos who are already inclined to get mad at you seize on the lack of IRB approval as a grievance. A/B testing is NOT regulated by anything like an IRB, even though it’s psychological testing for the explicit purpose of separating the targets from their money, without informed consent.

    Although you might be mistaken about whether your plan does in fact rely on an institution that this would expose to perceived reputational risk.

    Not all institutions that seem like public utilities behave as such. Email & office software providers will probably let you do nearly anything they don’t think is spam, but if you make an app that seems medical at all it may be difficult getting it approved without an academic or related institutional endorsement, which may then require an IRB. (A friend just had trouble getting their open-source contact-tracing app into an iOS testing environment for just this reason. No trouble with Android, though.)

    As pointed out elsewhere, academic journals and reviewers are, unsurprisingly, extremely credentialist, and thus exactly the kinds of institution you’d expect to care about IRB approval.

  48. Well... says:

    The meetups section of the sidebar makes it seem like they’re all cancelled, but presumably many of them are still going on via some kind of virtual platform (Google hangouts, Zoom, etc.). This presents an opportunity for people to attend meetups in other areas they normally wouldn’t be able to get to. May I request that organizers post dates/times/links to their virtual meetups?

  49. Loriot says:

    It occurred to me that I should probably start thinking about what to do if my roommate catches CV. We’re both working from home, so the risk is low, but I figure it’s better to think about these things now rather than having to figure it out on the spot should the worst happen.

    I share a two bedroom apartment with a roommate. My understanding is that if one of us catches it, it will be basically impossible for the other to not catch it as well, though I suppose it doesn’t hurt to try. I’m also wondering about things like how to react in an emergency.

    What sort of things should I be thinking about to prepare? P.S. We live in California, in the south bay.

    • Edward Scizorhands says:

      What facilities do you share? Kitchen, bathroom? If one of you caught it, do you both agree that the person who caught it would try to isolate themselves, and in exchange the unsick person would bring food to the sick person?

      You could use colored tape to mark paths on the floor for each person to walk, and places on each door for each person to touch.

      Do you have a thermometer that you could use to monitor your temperature daily?

      When I was prepping[1], I bought a shower curtain that I could use to try to seal off a room, as a makeshift filter in addition to a door, and that if this all turned out to be nothing would still be useful as a shower curtain.

      [1] I was very late to the game as far as preppers go, but way ahead of the game as far as the normal public.

      • Loriot says:

        We do share a bathroom and kitchen. I’m not sure he’d want to eat anything I might cook (he followed the carnivore diet and worries a lot about nutrition). I don’t have a thermometer.

        • Edward Scizorhands says:

          Can you agree to get food the other would eat? That’s your call on if you could come to an agreement, but both of you will seriously benefit, from either side, whichever gets sick. Even if it means the sick one orders out and the healthy one drops it in front of their door.

          If you can, come up with schedules for using common areas, so you know you can avoid each other. If nothing else, then top-of-hour/bottom-of-hour, or even/odd hours. The sick one should clean thoroughly after using it. If you can’t come up with schedules, text each other when leaving your room.

    • Gerry Quinn says:

      It may be that many people can’t catch it or get it asymptomatically, so there’s that at least. In fact I think the advice here – at least of a week or two ago – had you holing up longer than him, because he’s further on in the infective stage.

      A mask for him while he’s around the house, and him using facilities last (are chamber pots still spoken of?) and eating only pre-prepared food in his room. While you go heavy on the bleach and soap and never ever touch your face. That’s all I can think of!

      If he gets to an emergency state, he’ll have to be admitted to a hospital. I guess it would be good to keep a history of dates and symptoms.

      (Speaking of pre-prepared food, I guess buy or prepare and freeze some, including stuff of a kind that a sick person might like.)

    • JayT says:

      I think the only thing you can really do in the case that your roommate catches it, is leave. Even then, there’s a really good chance you already have it from the period before he was symptomatic. If I remember right, you’re pretty young, which means you’re probably a fairly low risk person anyways. Unless you have other health reasons, I think you probably just shouldn’t worry too much.

    • Robin says:

      The trouble is, one of you might catch it, but have very mild or no symptoms at all, but still be infective.

      And just a few OTs back, they were saying that whoever catches it outside will have a much smaller virus load and better chance of a mild outcome than the one who gets infected at home and gets a higher virus load.

      But keeping all this in mind while not even showing symptoms will make your daily life pretty unpleasant, when you still need to share a kitchen and a bathtub (Ha! Loriot reference!).

      Best is if anyone suspects having symptoms, start keeping distances, washing hands, maybe wearing masks around each other? A more positive thought: Some people get very bedridden and unable to get up, so you might be lucky to have a roommate at all who can look after you.

    • robirahman says:

      I had coronavirus already and have recovered, and my roommate didn’t catch it from me. (Or if he did, his case was asymptomatic.) I would suggest avoiding being in shared spaces concurrently, and airing them out after you walk through them if possible.

      I had also stockpiled two weeks worth of food and multivitamins, and that turned out to have been a great decision. I was stuck inside for two weeks, but at least did not suffer much for want of items available only outside.

  50. matthewravery says:

    This board has basically failed to address the most pressing issue of our day*: Tiger King

    Critical questions that need to be addressed:

    1) Who is the most awful person in this show? Despite everything, I kinda want to go with Antle.

    2) Did Baskin kill her husband? Plenty of memes around this, and a lot of folks close to him seem sure she did it. She certainly was ready to pounce (heh) when his disappeared, but I heard a lot of motivated reasoning behind, e.g., the ex-wife and kids. Plus, I think it was obvious that he was into so scummy stuff. I put the likelihood that she did it around 70%.

    3) Should Joe be in jail? I think he obviously should, though the murder for hire charge seems like a transparent set-up by Jeff Lowe and company.

    4) What’s the deal with Carole’s new husband?? Who is that guy? Where did he come from?! Biggest unexplored mystery in the show, IMO.

    *Or maybe I’ve missed it in all the COVID-19 discussion

    • BlazingGuy says:

      1) Agree that Antle is probably the worst, but a lot of people gave him a run for his money.

      2) Cui bono? Wouldn’t he have tried to take some of the money with him to Costa Rica?

      3) I mean probably? I agree that he was set up by Jeff and Alan, but it sounds like he actually *did* give Alan money to go kill Carol (even if Alan never planned on doing that.) AIUI that would not rise to the level of entrapment, if Jeff/Alan had been police officers and not sketchy weirdos. I don’t really know about the ethics/legality of euthanizing tigers that you bred yourself.

      4) No idea but I heard that Netflix is going to drop a new episode this week, maybe the people will finally get the answers we deserve!

    • Conrad Honcho says:

      Uh, spoiler warning maybe?

      1. How do you figure Antle is more awful than Joe? Sure, he’s got a sex cult, but so did Joe, and, while I think joining Antle’s sex cult is a terrible idea, nobody forced the girls to do it. They seem happy. If it’s about cub euthanasia that’s not proven, and Joe shot a couple of his tigers, too.

      While Joe is the most awful, Carole gives me the creeps more than any of them. Her act is so phony, and she started this mess going after everybody, and she is no better. She was going after the competition, exploiting people, exploiting the tigers, while pretending like she’s all noble. At least the rest of them admit how awful they are.

      2. Yes. I mean, obviously. 100%. I don’t care if Yudkowsky sets fire to my garage in the middle of the night.

      3. Yes. And I doubt he’ll get out early because you kind of need to show some contrition, and as far as I can tell he thinks this is all a travesty of justice.

      And I don’t buy the murder-for-hire thing being a set-up. There’s absolutely no reason for Jeff Lowe to set him up. He already had the zoo and had it in his name for years. There’s no reason for him to set up Joe. And given Joe’s extensive public rants about wanting to kill Carole, there’s no reason to think “gosh, he might have hated her but he would never do anything like that!” Of course he would.

      4. I don’t know, but I also found him creepy. And the biggest unexplored mystery was why Joe could buy his husband trucks and guns and toys but not a damn bridge for his meth-mouth. Edit: Or a shirt.

      • matthewravery says:

        Maybe it’s because I don’t know Antle’s background. Both Joe and Carole seem like they came from pretty fucked up circumstances and sorta stumbled into where they ended up. With Joe especially, nothing seemed premeditated. It was all just him reacting to things. He didn’t set out to do this thing with the cats, it just happened to be the first consistent way he found to survive (and thrive?) in the world. Carole I read in a similar way. She stumbled around until she found a grift that worked and then couldn’t do anything but double down because she didn’t know anything else to do. I also think both Joe and Carole are full believers in their own bullshit.

        With Antle, I got the sense that he knew exactly what he was doing every step of the way. Less self-deluded, more calculating. Also the soul patch.

        • Conrad Honcho says:

          I see where you’re coming from, but in the episode about Antle, didn’t they show that he grew up with the elephant, and studied with the…yogi or whatever the spiritual leader was? It seems to me like that was who he is, but then he basically figured, “I’m really awesome at all of this stuff, I should also have a harem.”

          ETA: #1 thing I learned from this show: tigers will get you mad laid.

          • Winja says:

            Lotsa people drinking their own bathwater and peeing in the tub in that show.

          • Winja says:

            ETA: #1 thing I learned from this show: tigers will get you mad laid.

            Was anyone else surprised at how cheap it is to buy a tiger cub?

            $2,000 for your very own tiger seems crazy cheap. I’d have expected a tiger to be more like $5,000-$10,000.

          • Conrad Honcho says:

            Yeah but then they cost like $10k/year just in meat. Unless you get expired meat from Wal-Mart, that is.

          • Winja says:

            True. Food costs, like you say, are ridiculous. Also, I would presume that you’d also have to have ample space, and you’d probably have to take them to a specialist vet.

    • Winja says:

      I would go with Lowe possibly being the worst.

      The difference between Jeff Lowe and Joe Exotic is the difference between Steve Martin and Michael Kaine in Dirty Rotten Scoundrels.

    • Anteros says:

      I’m slightly embarrassed to admit that I watched Tiger King (My wife was watching it, and it sort of crept up on me, like an unpleasant rash) but hey ho..

      1) I’d almost find it easier to answer ‘Were there any non-awful characters?’ Just about – the young lady who had her arm torn off by a tiger seemed surprisingly sane and grounded.
      Funnily enough, I quite warmed to Antle – despite the obvious character flaws.

      2) No idea, but of course she quite clearly could have done.

      3) Yes, for a few months or so.

      4) Good question, but also ‘Isn’t it a remarkable coincidence that he’s just as much of a fruit cake as everyone else on the show (torn-off-arm-lady excepted)?’

      I don’t want to provoke the ire of all you wonderful USanians, but both my wife and myself were muttering ‘Only in America…’ as the characters and their behaviors became more bizarre by the episode.

      ETA How come the only person on the show who could walk normally was the guy with no legs??

      • BlazingGuy says:

        This was not at all clear from the show, but apparently Saff is a trans man who hasn’t gone by ‘Kelsey’ in years. I guess a lot of people were unhappy w/ their portrayal on the show, but for my money that’s the only one that’s remotely justified.

        I’d also submit that Reinke isn’t awful. Maybe there is something about losing a limb that gives people a better sense of perspective?

        • Anteros says:

          Thanks for the correction.

          And yes, I agree that Reinke wasn’t awful. In fact he was so non-awful, I’m not sure why he was on the show..

        • matthewravery says:

          I actually thought most of the folks who worked at Joe’s zoo (pre-Jeff Lowe, and not including the reality show guy) seemed okay. Like, I buy that most of them had tough lives and were working there because it was one of the few paths open to them. And they legit seemed to care about the cats and got pissed at Joe when he stopped caring.

          But a lot of that could come down to framing by the filmmaker.

      • Winja says:

        Agree that the employee who lost her arm seemed decent.

        My vote for most reasonable person on the show is Joe’s Campaign Manager.

    • Jeff Lowe is the most detestable person in my opinion. He doesn’t seem to have any kind of sense of morality.

      Honestly I can’t hate Joe Exotic. Yeah, he tried to kill Carole Baskins but can you really blame him? She ruined his life out of spite. That would be enough to make anyone want someone dead.

      • Winja says:

        I want to like Joe Exotic because he seems like the archetype of the sort of larger than life person that can only exist in a libertarianish-type America; someone who’s totally outside of the norm and manages to take their outsider status and build something real with it.

        He’s the sort of protagonist that I want to root for, but I can’t just simply because he’s so ridiculously unsafe; firing guns around other people, engaging in sketchy behavior to run his business, distributing meth, etc.

        On the other hand, Carol Baskin is the sort of despicable person I can’t stand. The archetypal nosy bitch who can’t abide by anyone doing things she doesn’t like, even if they’re the sort of things she’s doing. She’s a total Dolores Umbridge.

        Jeff Lowe is a despicable grifter who’s perfectly happy to take advantage of other people for his own gain; he has no goal outside of satisfying his own avarice.

        I want to like Antle for the same reasons I want to like Joe Exotic, but ultimately, I think he and Lowe are somewhat similar in that they want to use the pretext of dangerous animals and wealth as a way to get laid. Lowe just packages it up in Vegas glitz and Antle dresses it up in Eastern mysticism.

        Of all of the animal dealers, the least despicable one seemed to be the ex-drug dealer guy who showed up for about one episode.

        • Anteros says:

          She’s a total Dolores Umbridge.

          Spot on!

          • Winja says:


          • gbdub says:

            Was Umbridge ever so obviously hypocritical though? Nosy, cruel, condescending, sure. But the infuruating thing about Baskin is the hypocrisy. She got her start breeding big cats and teaching other people to have cats as pets. She runs a zoo where she ruthlessly exploits volunteer labor with a creepy Skinner box shirt system, monetizes the hell out of everything, spends all that money on he legal feuds, all while riding her high horse and unleashing self righteous twitter mobs on her rivals.

            At least she doesn’t breed cats anymore? But oh yeah, she offed her rich hubby and dumped his family.

          • Conrad Honcho says:

            And her zoo looked less than pleasant for the animals. Antle looked like he had the nicest facilities. The one thing Carole’s doing better than the others is “no cub petting/breeding,” but that’s not a financial hit for her. She could just as well have realized if she does cub petting she can get a few bucks from the rubes who come to her zoo, but if she comes out as an anti-cub petting crusader on the internet she can get lots of bucks from rubes all over the world. I’m not sure that counts as virtue.

          • Winja says:

            I don’t get the “no petting” thing.

            I mean, I could see how it might be a bad deal for a cub that you’re planning to raise to release into the wild, but under a controlled situation with proper supervision, it seems to me like something Carole Baskin was just using as a wedge issue than something that’s really damaging in and of itself.

            Frankly, I think her hatred for Joe was a much higher priority than her love of big cats.

          • Conrad Honcho says:

            I think the real problem is the perverse incentives. Cub petting looks like it makes some serious money. But then you kind of want cubs so you can make money off them. Which can only be used for petting for about six months. Then they turn into a $10,000/year expense that can only be looked at, just like the other 100 tigers you’ve got. So maybe Doc Antel takes six-month-old cubs out in the back woods and snuffs them out, while breeding more cubs he can make bank off of for six more months.

          • Winja says:

            I could definitely see that being the case. After all, how many people in the US are going to buy one of those tiger cubs to keep as a pet after that 6 month period where you can do the whole petting zoo thing?

    • Konstantin says:

      One thing I would have liked to see is an experienced zookeeper explaining exactly how you humanely take care of tigers and what everyone is doing wrong. I know you probably shouldn’t feed them a diet of diseased cows and expired Walmart meat, but it would be nice to have an expert voice explain why. Do any of these people employ licensed veterinarians, or anyone with formal training and experience working in accredited zoos?

    • gbdub says:

      Didn’t Scott write a whole post about Buddhist sex cults? Surprised this hasn’t come up yet, since it features one!

      1) Antle is probably the worst, but we don’t see enough of him on the show so seems comparatively sane. I mean he lures girls into a creepy sex cult and runs the big cat version of a puppy mill (hidden behind a slick exterior).

      Carole is worse than Jeff. Jeff is a sleaze bag and a con artist, but he’s amoral. Carole is a hypocrite and probably a murderer.

      2) Yes. If not yes, she definitely knows a lot more than she’s letting on about what happened.

      3) Yes. The particular way he got nailed was a setup, but he was going increasingly insane and was going to something that wound up with him jailed or dead eventually. And it es clear that whatever care he may have originally had for the animals, it was long gone by the time of his downfall.

      4) Agreed that his origin was oddly unexplored given how important he was… it seems like there is more to him than just Carole’s sidekick

  51. Ethical considerations with research are a professional ethics concern and would not typically overlap with actual laws, except insofar as a research grant carries with it certain legal obligations. And of course it would be entirely possible to design a study that violated actual laws. As an easy example, any RCT involving marijuana would still federal law.

    I know of no legal basis on which your voluntary vitamin study would violate any state or federal law. That said, there are a lot of laws in this country (and around the world, since you have an international audience), so if you do this enough times, it’s entirely possible you could inadvertently run afoul of some obscure law. To definitively answer your question, you’d have to research the laws of every state and every country where you have people who are solicited in the study with specificity. That, of course, would be a colossal waste of time.

    Most research does not run into this problem because it is conducted within a single jurisdiction and under the guidance of one or two universities. Sadly, the frictionlessness on which the internet is built does not lend itself to simple answers with respect to complex legal questions.

    I suspect if I were sufficiently motivated, I could find a few ways in which you’re violating GDPR with this site. It’s just a question of whether you’re doing this enough to get on the radar of a motivated regulatory authority. The answer is almost certainly not. My suggestion would be to use common sense and cross your fingers that you don’t get in the crosshairs of some motivated prosecutor or regulatory body.

    • I suspect if I were sufficiently motivated, I could find a few ways in which you’re violating GDPR with this site.

      I don’t doubt this is true, but I do find it an odd example to pick, since this is harder than most people might initially think. Allow me to hijack your comment to dispel some GDPR myths! 😀 (I’m sorry, this topic is like a flare gun, it summons me.)

      He’s not selling anything to European citizens; he’s not offering a free service explicitly to European citizens, either; that European citizens use his site despite the aforementioned doesn’t actually imply his site falls under GDPR. But assuming for a moment that it did, since the site is (effectively) only run by one person (Scott), this rather limits what GDPR would even expect of him. HTTPS and the promise to keep his WordPress installation and plug-ins up to date might be quite enough.

      (That said, IANAL, though I did do GDPR compliance for a living for a few months, and I’m one of the bizarre people that actually read the whole thing.)

      • gph says:

        I’m curious, what would happen if he wasn’t in compliance and whatever organization enforces the GDPR decided they wanted to take action? Would they attempt to enforce fines in some way? Would they block the site in Europe?

        • pas says:

          Short answer is nothing. Because SSC is not an EU business.

          GDPR is not a criminal thing, the EU doesn’t really deal with criminal stuff anyway (member states do and some of them do sometimes block sites usually via DNS or BGP that they deem illegal – for example online casinos without license in some specific member state).

          Enforcement has two routes, one is when someone in a member state reports a site (actually they report the business – as the GDPR extends to mailing lists and even paper based stuff too, obviously, but talking about sites is easier), or if the data protection agency in that member state notices a problem without a report, and the other way is when an EU agency starts a proceeding (and this is when we get the news about billion euro fines).

          As far as I know there’s no blocking just fines in the GDPR itself. But member states usually have their own version too.

          • gph says:

            Thanks for the response, that’s about what I expected. Though it does make me wonder why some newspaper websites based in the US would block European traffic due to concerns about the GDPR? I guess they probably are owned by multinational media conglomerates that do business within European jurisdiction.

        • Edward Scizorhands says:

          GDPR insists they have domain over things that have no connection to the EU.

          In practice, they aren’t going to fine Scott, and if they did Scott could use the SPEECH Act to tell them to STFU.

          • acymetric says:

            How would the EU even go about enforcing a Fine against Scott, a person who lives in the United States who owns a site that (I assume) is hosted in the United States?

            What mechanism even exists that would allow them to enforce anything here?

          • Edward Scizorhands says:

            That’s why “in practice, they aren’t going to fine Scott.” They would rather ignore him than reveal their impotence: that they cannot enforce a fine against some guy.

          • acymetric says:

            That’s why “in practice, they aren’t going to fine Scott.”

            That isn’t “in practice”, it’s “by policy and jurisdiction”. “In practice, it won’t be enforced” makes sense for people driving 53 mph in a 50 mph zone, not here.

            “In practice” implies that something could be enforced they just don’t enforce it. That isn’t the case here.

  52. Edward Scizorhands says:

    It’s early to see if this is true, and the NYT is biased, but Germany says they are getting good results by putting people on ventilators very early.

    “When I have an early diagnosis and can treat patients early — for example put them on a ventilator before they deteriorate — the chance of survival is much higher,” Professor Kräusslich said.

    • albatross11 says:

      That is a strategy that relies on having a lot of ventilators and a lot of people trained to manage patients on ventilators. I think it won’t be able to continue if they get a huge surge.

      • Scott says:

        Sure, but it’s still evidence that ventilators actually do something to better clinical outcomes.

  53. Conrad Honcho says:

    If you’re stuck at home with the kids, have a Switch and are looking for couch co-op games to play, there’s some good sales right now.

    Trine Ultimate Collection (4 players) – On sale 19.99

    The Swords of Ditto (2 players) – On sale $7.49

    No Heroes Here (4 Players) $14.99

    Super One More Jump (4 players) – On sale $0.99

    Conduct Together (4 players) – On sale $0.01

    For The King (3 players) – On sale $12.49

    Death Squared (4 players) – On sale $2.24

    Full Metal Furies (4 players) – On sale $6.79

    Crawl (4 players) $14.99

    I bought Trine, Swords of Ditto, For the King, and Conduct Together (literally only a penny).

    I also highly recommend Castle Crashers, Lovers in a Dangerous Spacetime, and Guns, Gore and Cannoli 2 (1 is also good but 2 is great) but none of those are on sale right now. Still all pretty cheap though.

    On Steam Greedfall and Disco Elysium finally went on sale, so I picked those up. Also Hellblade and I finally bought Cities Skylines. Just finished Jedi: Fallen Order, which was really, really enjoyable. It was just…a video game, instead of a platform for EA to sell me microtransactions. Please make more Star Wars games like that.

    What are you playing? Besides Plague, Inc, obv.

    • Eltargrim says:

      Just got Hollow Knight in the Humble Covid Bundle, so giving that a shot. My metroidvania experience is pretty slim (Just the Metroid Prime games), but I’m having a good time so far. Getting the dash cape has been helpful, was having a harder time with the moss knights without the extra mobility.

      • Three Year Lurker says:

        The pogo bounce where you swing your nail down as you come down on something is probably the most important skill. Good for getting in a couple hits when the knight swings at you.

        • Eltargrim says:

          I’ve noticed a few sections where the pogo bounce seems to be a method of mobility, rather than a simple attack; in particular I’m thinking of the path between the Greenway and the Fog Cloud (sorry if I’m butchering the names, been playing only for a couple of hours). I’m not great at it, but sounds like I should practice!

          • Three Year Lurker says:

            Yeah, it’s important in many places. You can even cross some spikes with it (but not spiky vines). There are some things that are impossible without it. I don’t want to spoil anything, so I won’t say what.

            It’s also possible to parry some attacks by hitting their attack on the right frame. I can never time it right on purpose, so I don’t try.

            If you ever need to travel a long distance without taking damage, travel the path twice without using a bench. Enemies only respawn when you rest so you can go through once clearing enemies, then come back and have a much easier time avoiding damage. (this will make sense when you reach the point you need it)

      • mitv150 says:

        I found that skipping over a difficult portion and coming back later almost always led to it being much easier the second time around due to either new charms or new abilities.

      • Conrad Honcho says:

        Now I don’t want to do anything illegal, but I would kill a man, in front of his own mama, to get just the release date for Silksong.

      • Peffern says:

        Hollow Knight for game of the decade 2010-2019. Enjoy it. It’s great.

      • Apogee says:

        Me too! I played it back when it first launched but this is the first time I’ve gone for a serious try at the DLC, which gets pretty ridiculous.

    • Matt M says:

      I’m generally a few years behind on gaming trends, but lately I’ve been reaching even farther back in the well than that.

      I just recently completed Alan Wake, which had been sitting in my library for years following when I picked it up for like $1 when they put it on fire sale because music licensing issues were about to cause them to have to yank it off the market entirely. Decent game, but not great. Not really my favorite genre so maybe others would enjoy it more.

      And I just started Saints Row 4, which I’m enjoying quite a bit. I’m a pretty big fan of the sandbox/action genre and for several years, this was basically the little brother franchise to GTA, but I never really got into it. I’m liking it enough to regret having not played it when it was current. I’m somewhat interested to go back and revisit the earlier games in the series, but I always struggle with that because I just hate going “back in time” as it regards features and convenience. I do hope they re-start this franchise again someday though. “GTA but with the absurdity cranked up to 11” is a formula that is working very well for me.

      • acymetric says:

        And I just started Saints Row 4, which I’m enjoying quite a bit. I’m a pretty big fan of the sandbox/action genre and for several years, this was basically the little brother franchise to GTA, but I never really got into it.

        Saints Row was like the little brother and the class clown to GTA. Took itself even less seriously, which allowed for a lot of great fun. I haven’t played since II (I may have played III a little) but I really enjoyed that series.

        • Loriot says:

          I haven’t played them, but my impression is that it started as a GTA parody, but each successive game doubled down on the wackiness factor, to the point where the games end up with you becoming president, fighting off an alien invasion, and then recusing people from hell while they sing opera songs.

          • Matt M says:

            4 (the only one I’ve played) starts with you as President and all your friends as cabinet, and then immediately proceeds into a crazy alien invasion.

      • I discovered the existence of Everquest Project 99 ( and have been enjoying some nostalgic MMO action. Turns out that the game, which was a quantum leap twenty years ago in several ways, looks even more rich and gorgeous on my new medium-gaming-caliber (for 2020) graphics card and monitor.

        As for the gameplay, I’m loving it now just as much as I did then. Nothing but compliments for the volunteer team who’s behind Project 99.

        The Project 99 population has risen these past couple of weeks, just enough to provide plenty of fellow lowbies/tourists for me to group with. (I was never a hardcore raiding guy even back in the day.) So I’m pretty happy with it as an immersive sheltering-in-place diversion.

        Anybody wants to try it out, email me and I’ll help you shortcut the install.
        And if you join the Green server, look up Chinetter the dwarf cleric or Spotswood the wood elf ranger.
        “You’re In Our World Now”

      • Nick says:

        My only familiarity with Saints Row is Jesse Cox’s playthrough of 3.

      • Faza (TCM) says:

        There’s a pretty big leap in premise and feel between Saints Row 2 and Saints Row the Third. SR2 still feels like a GTA clone, whilst the Third gives up all semblance of sanity or seriousness – and is all the better for it.

        If you’ve got a bit of spare cash (I think ten to fifteen bucks on, you might want to put SR4 on hold and finish SR3 first. SR4 has major throwbacks to previous parts of the series (but SR3 more than anything) and it feels so much better if you have the background. Plus, SR3 is an excellent game in its own right.

      • Matt M says:

        I enjoyed that video and wholly agree with just about everything said in it.

    • Randy M says:

      There’s a free demo for Mario and Sonic at the Olympics which I played when I was with my kids.
      I finally bought Darkest Dungeon this weekend. Scratches an itch I had for a strategic, team management fantasy game.

    • tossrock says:

      Divinity: Original Sin II is great for couch co-op, which I’ve been playing a lot of with my partner. Really satisfying turn-based combat gameplay, with a Diablo-like loot system and real RPG style dialogs. Would recommend highly.

    • matkoniecz says:

      What are you playing?

      I really recommend Factorio if one likes base-building/optimization/construction games. Has free demo. Game is well optimized, with big attention given to fixing bugs and performance issues – next versions keep getting faster, not slower. Massive amount of content, no grinding. It has OK price (and never was on sale and developers announced that sale will never happen).

      As bonus is really interesting blog by its authors about how game is developed.

      Besides Plague, Inc, obv.

      Plague, Inc seems to be typical in-game-purchase game, so I assumed that it is not worth playing. Even in case of bonus irony.

      Additionally, it looks like a game mechanics that are supposed to be modelling reality but utterly unrealistic to the point that will just irritate me. Apparently plague in this game is a telepathic Zerg that evolves worldwide at once with the same mutation.

      • acymetric says:

        There isn’t really any reason to ever purchase anything in-app for Plague, Inc. It’s just a fun (free) time-killer.

        As far as modeling reality, yes a free mobile game isn’t a perfect real-world model of disease spread/evolution…but most other games aren’t perfect real-world models of anything either.

      • Nick says:

        Apparently plague in this game is a telepathic Zerg that evolves worldwide at once with the same mutation.

        That’s absurd, to be sure, but I don’t think there are better options for a playable game. Suppose you choose everything at start instead. Given a plague doesn’t have any agency, you’re effectively locked out of affecting the results of the simulation once it starts playing, which is not very fun. It may be neat to watch things play out, but it’s not a game.

        (It’s not impossible to imagine a game where you’re locked out after turn one. Suppose it’s a two player game where one player kits out his plague and another player chooses how the world will respond and what resources it will have. That would be very different, but at least it would be a game.

        It seems hard to make it fun, though. Suppose you give both players a lot of variety. The plague player should just go every time for something weird that the world is unlikely to be specially prepared for. But the defensive player, without information, will just play the likeliest strategy to win; perhaps a very conservative strategy that invests a lot in cure research and fast acting quarantines. Alternately, suppose you don’t give the players a lot of variety. Then where does the strategy come in?)

    • acymetric says:

      Currently working my way through FF7 (still on disc 1). The mods available for the PC make it a pretty nice experience.

      Have Baldur’s gate, Elder Scrolls Online, and the two Knights of the Republic games queued up…keeping my eye out for other games on my wishlist to go on sale.

      I also got Crusader Kings II for free, but I haven’t been able to focus and get into it enough to understand what is happening.

      • Randy M says:

        Currently working my way through FF7 (still on disc 1). The mods available for the PC make it a pretty nice experience.

        Interesting timing, I hear they just released the first episode of the remake (PS4).

        • acymetric says:

          I don’t have a console (and I don’t like paying 60+ for games) so I probably won’t be picking that one up any time soon.

          • Edward Scizorhands says:

            We bought Untitled Goose Game on the Switch and it is so much fun for 20 bucks.

          • Conrad Honcho says:

            I want UGG, but I refuse to pay $20 for a 3 hour game. If it goes on sale I’d pick it up.

          • Randy M says:

            Fair enough, I’m in a similar place as far as that one goes, but I saw a review for it this morning that was pretty positive.
            The end of disc one was, what, the Cid and the Rocket? Or the haunted mansion and Vincent?

          • acymetric says:

            No clue, I’ll let you know when I get there 😉

          • Randy M says:

            Well, that makes sense. Do you have Cid, Vincent, or Yuffi yet?

            The mods available for the PC make it a pretty nice experience.

            I missed this at first. Are these content mods, or more of a UI thing?

          • acymetric says:

            UI stuff. Higher resolution for textures and the menus, updated character/enemy models that look a little less dated. Nothing that changes the story or the way the game plays (although there are mods for that, I’m not using any). Basically just visual improvements.

            Oh, and newer arrangements of the music that sound better than the old MIDI stuff (the MIDI especially seems to not come off well on Windows for some reason, apparently). Still the same music, but someone arranging it with real orchestral sounds and such instead of just base old school MIDI.

          • Matt M says:

            In the original game, the end of disc one was… uh… I guess it counts as spoilers still so I won’t say it, but it’s the spoiler from this game that everyone knows…

          • Matt M says:

            Oh, and newer arrangements of the music that sound better than the old MIDI stuff (the MIDI especially seems to not come off well on Windows for some reason, apparently).

            IIRC, the soundtrack between the original PS1 version and the original PC version were actually different. I think the PS1 didn’t actually use MIDIs but something a little more advanced/complex. But the PC version went full MIDI and a lot of the songs were sloppily remade as midis. I could be remembering that wrong though…

        • theredsheep says:

          I think it gets released on the tenth, so three days from now. To be avidly snapped up by a grouchy, demanding, quarantined audience. Poor Squeenix.

          • Conrad Honcho says:

            I hear good things, but there is zero chance I’m playing until the full game is released. Episodic games can bite me.

      • Conrad Honcho says:

        If you want something new in that genre, I’m really enjoying Greedfall so far. It’s on sale on Steam (or was last week).

    • Apogee says:

      I’ve been getting into One Step from Eden lately. I don’t usually like deckbuilding games for a variety of reasons, but I’m having fun with this one. I think I’ve seen people talk about Slay the Spire in the discord before, and this has a pretty similar setup albeit with another layer of mechanical execution added on.

      I share a house with a bunch of other nerds and we’re all on lockdown together, so we’ve also been playing a bunch of board games both on and off Tabletop Simulator. One of them even has a VR rig that we got set up in the living room and I’ve been having a blast with it.

      • Randy M says:

        From what I’ve seen there’s a lot of twitch to it Eden.
        Slay the Spire recently added the fourth character to the switch version, so I’ve been learning that again, much fun.

    • Nick says:

      Let’s talk Minecraft. 1.16 is on its way, bringing a lot of new content to the Nether, including new biomes and mobs. I’ve been playing a bit of 1.15, and I like most of the content they’ve added in the last year or so, following the long sleep. The new villages are really nice, and the ocean stuff is neat. So are the new textures. I also like many littler additions like hay bales and campfires; they’re both so multipurpose!

      I mostly play modded, though. I have a 1.12 pack I put together with the following mods, chosen basically to make the game more fun and pretty to explore: Ambient Sounds, Better Foliage, Biomes O Plenty, Dynamic Trees, Sound Physics, Underground Biomes, YUNG’s Better Caves. Plus Optifine and shaders, and a few others for quality of life. It looks gorgeous, at least going by the usual standards.

      I also have a 1.10 pack with a secret building project on it. It’s just an instance of Feed the Beast Beyond with about half the mods disabled and a few of my own selection added, because I was too lazy to gather the files myself. I also made some of my own textures, because I wanted a higher resolution. Texture packs can’t hope to keep up with the modding community, so you either make due with a mixed pack—ugly, especially if you want higher resolution—or you make your own—impractical, but at least you can be selective. In my case I could afford to be selective.

      • VoiceOfTheVoid says:

        Currently enjoying a Realms server with my friends, we’ve been working on a town and starting wars over the forest of bamboo that has grown from the single shoot I found in a sunken ship.

        Also, figured out how to run a modded minecraft server on Amazon EWS’s free tier, so that I could play this gamemode where you fall sideways with my friends. If you want to avoid having to use the honor system and give yourself another anchor from creative mode every time you die, simply give yourself a repeating command block, install it in the bedrock near spawn, set it to no redstone needed, and give it the following command:
        /replaceitem entity @a slot.inventory.26 mysttmtgravitymod:gravityanchor 1 2

    • acymetric says:

      Since we’re talking about games, what do people think about No Man’s Sky? I bought it when it came out, only to find that my system at the time couldn’t run it (I had a slightly old Nvidia card) so it sat “on the shelf” for several years. My new laptop, it turns out, can run it. Is it worth trying to dive into?

      • Conrad Honcho says:

        I have not played it, because the reviews when it came out were “this was so promising what the hell did you people do?” A year later apparently they fixed it and it’s a good game now.

        So, I’ve never played it, but people on the internet say it’s good now, and you should probably give it a try and report back.

    • DeWitt says:

      I got Bannerlord on release, and have been playing it since then. The game is about as janky as you’d expect, but Taleworlds has been patching it literally every day since then, and the multiplayer is really quite enjoyable already

    • Just bought Super Smash Bros. Ultimate, which should keep me entertained for a good while. I’m starting with Adventure Mode and will see where that takes me.

    • Iago the Yerfdog says:

      It’s not on sale, but I got the Megaman Zero/ZX Legacy Collection and have been really enjoying it. Close to beating the first Zero game (using Save-Assist but not Casual Story mode), then I plan on going to ZX since I’ve watched LPs of the Zero games back when I assumed I’d never get a chance to play them but I purposefully avoided watching a playthrough of ZX or ZX Advent.

  54. dsc says:

    Question for people with a medicine/physiology background (which I definitely do not possess):
    Are ventilators more effective for people from high elevations? It seems to me this could be the case since their bodies are acclimated to having less oxygen, so I could imagine that when placed on a ventilator their bodies would be better able to make use of that sudden large influx of oxygen than people from lower elevations. Is this the case? Why or why not?

    • theredsheep says:

      RT student, haven’t got to vents yet, but if you need a vent at all you’re in pretty terrible shape, regardless of your starting point. IIRC–don’t quote me on this–people from elevation do better by having higher red blood cell counts, but since your RBCs are always being replaced that would probably go back to normal after a while at a low-elevation hospital. Not sure. In any case, people go on vents for different reasons; if you’re on there because you can’t protect your airway, for example, that’s a different issue, and if your lungs are filled with fluid the extra RBCs won’t do much good (extra RBCs also increase your risk of embolism, which is why we don’t all have extra). Being from elevation might encourage you to be in better shape and avoid getting on a vent in the first place, IDK.

      Also there is definitely such a thing as too much oxygen, for various reasons, not all of them relevant to this scenario.

    • rahien.din says:

      The rate-limiting step in oxygen absorption is gas diffusion across the biological membrane, and this must occur before the red blood cells can take up the oxygen.

      People who live at higher altitudes have more oxygen-carrying capacity in their blood chiefly because they have more red blood cells.

      In ARDS (such as due to COVID-19), the problem, in large part, is disease of the gas-exchange membrane.

      Since the adaptation/acclimitization is downstream of the rate-limiting step and downstream of the problem, there shouldn’t be any physiological difference between high- and low-to-moderate amplitude persons. They may be able to transport more oxygen via the blood, but their problem is getting oxygen to the blood.

      That said, in some circumstances you may have to operate your ventilator differently. If you are at high altitude and don’t adjust your ventilator pressure settings, your ventilator might not work as effectively. So there is a failure mode wherein high altitude could contribute to ineffective ventilation.

      • theredsheep says:

        Do you know if pressure at elevation would make air entrainment work differently? That just occurred to me.

        • rahien.din says:

          You mean in concrete production?

          • theredsheep says:

            Don’t know what you mean by concrete production. I mean in the Venturi system used for mixing air and O2. It depends on the width of an aperture to mix the air with O2 from the wall outlet. But if that’s calculated based on a given pressure from the wall, and the atmospheric pressure is different, I’d think that could skew the output. No?

          • Lambert says:

            I’d hope that there’s some kind of closed-loop (homeostatic) control system monitoring the air pressure and O2 partial pressure and adjusting valves accordingly. (Either automatically or with an RT twiddling the knobs till they get the right numbers)

            Absent this, I don’t think a venturi would behave the same. But I’m too tired to apply Bernoulli’s principle right now.

          • theredsheep says:

            Well, there’s not usually a need to dynamically adjust a Venturi system, since atmospheric pressure generally doesn’t change–but by the same token, I imagine that, if it makes a difference, hospitals at elevation have specialized Venturi masks/valves/etc. I’ve never heard either way.

          • Lambert says:

            So are they entraining air from the environment into the patients lungs?
            Without filters or anything?

          • theredsheep says:

            Filtering is a whole separate matter. But if you want a patient to have a given amount of oxygen per liter of gas inhaled into the lung (FiO2, fractional inspiration of O2), the simplest way to achieve this is to take room air, with a known O2 concentration of 21%, and add more oxygen. And the simplest way to do that is by letting in air through an aperture–you can do this with a cheap set of disposable plastic parts, no need for power beyond the pressure of the O2. I imagine ventilators use that same process, just with fancier parts. Although, come to think of it, I do believe they use medical-grade air from the wall outlet, so that’s your answer. I imagine there’s a mechanized Venturi process in the middle of the machine, anyway.

            ETA: And that’s my answer, too–atmospheric pressure is irrelevant if they’re using wall outlets for both. I do wonder about the entrainment systems we use outside fo vents, though, like in venturi masks.

          • Garrett says:

            > venturi masks

            Those have the property of being run at atmospheric pressure in-practice. All ventilation is handled by the patient so there’s no need to have changing pressure to force air into the lungs.

            They come with a series of plastic adaptors with slightly different opening sizes and require different flow rates for different fractions of oxygen which are embossed on the side in barely readable type.

          • theredsheep says:

            Yes, I’ve handled venturi masks. My question is, since they depend on using a variable-size aperture to allow a certain amount of room air to mix with the O2 flow, could that room air being at a different atmospheric pressure make a difference in their function? IDK enough to say either way; I haven’t heard that they use different masks in Denver, but then why would I?

  55. Well... says:

    I’ve noticed that the gas I keep in a 2-gallon gas can in my shed for my mower sits over the winter and still powers my mower when I fill up in the spring. In fact, if I fill up that gas can it will frequently last me a whole year; the gas in it doesn’t appear to go bad.

    This suggests it might be possible, even advisable while gas is cheap right now, to fill up several much larger gas cans, and then use it later when gas gets more expensive.

    I’ve heard people say you can’t do this because the gas degrades when stored over extended periods of time. Is it that my mower can for whatever reason tolerate the somewhat-degraded gas? Or is it that those people are wrong, or are imagining a scenario where the gas is like a decade old or something, rather than 6-12 months?

    • broblawsky says:

      The gas doesn’t become non-combustible, but it does become more hydrated and oxidized over time. That means it forms more gummy, polymeric deposits in your engine when burned. That can happen in ~6 months. Your mower engine is probably much less finely machined that your car engine, so the deposits have less of an impact, and because you run it less often, you don’t notice the changes in performance. If your canister is fully airtight, you can avoid this, but it’s hard to get large, airtight containers.

      • Well... says:

        That makes sense, thanks.

        Just curious, why would a large plastic gas can be less airtight than a small one? Does air get in through microscopic cracks in the seams? The actual openings to pour through are usually the same (presumably standard) size for both.

        • A1987dM says:

          The actual openings to pour through are usually the same (presumably standard) size for both.

          If the total amount of air leaking is the same for both, then the larger one will have less air leaking per gallon of gas.

        • broblawsky says:

          Pressure vessels just generally get worse at containing pressure as they get bigger. It’s harder to cast/mold them seamlessly. The square-cube rule is on your size, but practical mechanical engineering questions aren’t, and in this case, they win.

    • mitv150 says:

      I think the problem here is “much larger gas cans.”

      My car’s gas tank holds 17 gallons. That’s a considerable volume and it weighs about 100 lbs.

      Gas has dropped about $0.60/gallon, meaning that I can save $10 a tank.

      So, for each $10 saved, I need to fill, transport, and store 100 lbs of gasoline? This doesn’t seem economical or safe, short of owning some type of large tank that can be towed.

    • Econymous says:

      Ethanol is hydrophilic. Water in engines is bad news.
      I’ve heard from many sources that ethanol-free gas is pretty stable, but that gas with ethanol is really bad for engines after it has sat more than a month, particularly in a gas can, which I’ve never found to be 100% airtight. It might run but be doing permanent damage to your equipment.

    • Edward Scizorhands says:

      EMH complaint: if hedging like this was worth it, professionals who can manage millions of barrels of stuff would do it already.

      • Matt M says:

        Yeah, there’s a wide variety of people who are much better equipped to profit from doing this sort of thing than you are. They include oil companies themselves, terminal companies, container ship companies, and more I’m probably not thinking of.

      • matkoniecz says:

        Note that in this case transaction costs are lower. Product will not be resold for profit, but just used later.

        Storage costs are still likely going to make it a bad idea, but EMH complaint does not really apply here. Yes, market reaction will reduce profitability but this alone is not enough to make it pointless.

        In the same way as buying more of so