Pointing out what you think is an error is fine, as is saying you'd prefer if I relied less on Twitter, but you can do it politely. Mistakes will occasionally happen, that's the speed premium.
In addition, your comment here assumes that Everybody Knows all that and what is and isn't reliable, and why, and implies that anyone who disagrees is an idiot, and that's also Not Ok.
If other commenters agree that Lin is wrong on the substance, I will update and edit, but since it doesn't have huge practical implications and I don't trust your confidence for now I'm not going to look into it further and rely on the other eyes on this.
If others share this sense that this particular wording ('making my life worse') is having this particular effect I encourage them to speak up. If they do, I'll consider choosing different wording that means the same thing. If I don't hear collaboration on this, I'm going to assume it's fine.
In either case, I'm confident you got the message I was trying to send.
Hey, I'd like to add a second vote for fewer twitter links (minus the original vitriol). The reason is because I usually read your blogs at school and the school wifi blocks twitter for...reasons. Really enjoying the blog though--nice to see another both magic the gathering player and rationalist.
If the issue is twitter being blocked then that's an argument for using more direct links instead, which seems reasonable, but Studies Show (e.g. the stats) that the links are more to show you have them, and rarely get checked. Still, good note. Different mechanism than above, and isn't an arguments against e.g. quoting Twitter.
Trusting anything you see on Twitter by default is obviously a terrible heuristic, on that we agree. One must use such things responsibly, and know from context how much trust to have in any given statement, while keeping in mind the correct rate of getting it wrong is usually not zero.
The idea that it would be a trivial additional amount of time/work to 'validate these facts' is in my experience very wrong, even setting aside the fact that you have chosen sources you think are trustworthy and that validate facts, and that others would strongly disagree with those choices you have made. It would double the amount of time at least, and is basically saying *at least* 'hire someone to fact check like a newspaper would do.'
Instead, I do my best to calibrate and convey my level of trust in a given claim (and if it's not high enough, not use the source). It's all we can do.
The idea that these posts could be produced without Twitter, and that I would be better off avoiding it, is simply wrong. It wouldn't only take longer, it's not possible. There's a reason no one else is doing a similar thing (at least that I am aware of).
1) Your Local Epedimiologist this week stated "There is zero evidence that masks harm, physically or psychologically, kids."
2) There is no evidence that I am subscribed to YLE any longer.
3) Annecdotally, I've seen that masks-blocking-non-verbal-communicatoin is a pretty huge burden on people with Autism, and for boys more than girls. Most that I'm interacting with are middle-school aged and their parents. No one has ever studied this, so of course there is No Evidence on this either.
1) Yeah, I have no idea how people type that with a straight face. Definitely not if they have kids.
2) I'd say there's strong evidence against that hypothesis, but there's also some evidence in favor, which is that you seem to have previously been reading it, and implied that you previously subscribed. So there! Bayes ruled.
3) Interesting note, thank you. That makes sense, I could see it going either way (e.g. maybe they need the extra info, or maybe they don't know what to do with the extra info).
1) "It Is Difficult to Get a Man to Understand Something When His Salary Depends Upon His Not Understanding It"
2) Damn, got me.
3) Like many autism features, some people have it and some people don't. I know one little girl who has never had any trouble interpreting facial expressions, and masks don't bother her beyond the typical annoyances. Her brother on the other hand really is in perminant midnight. I have an entire TED talk in me about how autism isn't a spectrum, its a basket of goods.
Isn't DeSantis' point that while Omicron might be the dominant strain, Delta didn't just vanish like throwing a light switch? (The latest numbers I've seen say about 1.6% of Covid cases are still Delta in the US.) So why deny people unlucky enough to catch Delta instead of Omicron a treatment that works? Especially since Omicron's lesser severity means Delta will be over-represented in hospitalized cases relative to general prevalence.
To clarify, suppose that either Delta or Omicron were not a coronavirus. Would this policy make sense? Would you take away the rabies vaccine because we were in an influenza epidemic?
Probably not. So why take away a treatment some of the Covid victims could profit from, just because most wouldn't?
That arguments holds water if and only if you can identify the remaining Delta cases as Delta cases in time to give them the right treatments, and solve the logistics issues involved, and avoid giving it to Omicron cases. If that's true, then DeSantis has a (small and rapidly shrinking) point. But the new nowcast is 0.1% I think, not 1.6%, and even at 1.6% I doubt they'd bother testing for Delta.
Does it matter if it is the right treatment though? If you are in a scenario where it is life or death for a patient and you can't take the time to determine if they have omnicron or delta, give them the monoclonal antibodies just in case. The side effects of monoclonal antibody treatment are extremely minimal [1], so why not just give it even if it may not be the correct treatment (ie, they actually have Omnicron but you administer the monoclonal antibodies for Delta). If you have Delta, great, your chances of survival just increased dramatically, if you don't, you may have minor side effects. Obviously, this is not ideal, and you could evaluate what a patient has prior to prescribing treatment, but that is not the scenario presented.
Edit: Note, I did a cursory search for monoclonal antibody treatment adverse effects, and this seemed to be the best source. So I really am curious, as I just hadn't heard much about the safety profile, and never have had to seriously consider taking them. Perhaps there is a significant risk. And I should add the hypothetical argument goes against one of my core beliefs (which can probably rightfully be called out for being foolish) which is I personally strive to avoid medical interventions unless absolutely necessary, so the idea of shotgunning extra treatment just in case is anathema to me, but perhaps not for most people/ particularly scared people in a hospital with covid.
[1]Brobst B, Borger J. Benefits And Risks Of Administering Monoclonal Antibody Therapy For Coronavirus (COVID-19) [Updated 2021 Dec 3]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2022 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK574507/
On being post covid: I'm planning on waiting another week and a half and then restarting NYC rationalist weekly meetups - I want it to be very unambiguously over, and there are people who are more covid hawkish than me. Hopefully by then it'll be clear enough that we can agree to start being officially post-covid.
On predictions: mostly agree with your updates, but I'd buy airline mask requirements to 40% on the basis of "airlines love keeping dumb regulations" and keep the probability of omicron optimized boosters at 10% (I was at 8% before, so updating upwards but by less than you, and mostly as a response to your update rather than a direct response to the 5 month thing).
I don't see the airlines as motivational here, they will drop the mask requirement as soon as they're told they're allowed to do that. It's more that the FAA/TSA and related things love regulating airlines and think they're a special magisterium. I think if you're 40% now you should have been at 50%+ earlier. Which is reasonable.
And yeah, I agree we should restart meetups soon and 2/8 seems like a reasonable time for it. I notice I'd be willing to host then if there was need.
I'm confused about your discussion of schools and kids/students. I know that _some_ parents of school-age kids _are_ concerned about the kids themselves being infected and falling ill (or otherwise being harmed) – I regularly interact with one in particular.
But the, IMO, charitable interpretation around the general concerns of schools and students isn't that the kids themselves are at any significant risk but that everyone else they interact with outside of school is. The particular parent I mentioned has told me that their child has been sick several times already this winter (not from COVID, or at least negative on all tests), and all the anecdotal evidence I've ever observed has been that kids, and 'schools' (i.e. anywhere large numbers of kids interact with each other for extended periods), are probably the most common means of transmission for all kinds of contagious diseases.
I commented on an earlier post that the same parent's kid participated in a school music performance. That the school did this didn't and doesn't make any sense to me – in combination with all of the other precautions, e.g. mandatory masking at all times, even outside, and all classroom windows being open in the winter. I have greater appreciation of the thinking of the 'bubble person' you mentioned – they at least seem to have a 'physical model' and to be seriously cautious about being infected, whereas (most) everyone else, including the particular parent I mentioned, seems to be deluding themselves that whatever official precautions are mandated _must_ be (almost) perfectly reasonable.
Motte/bailey. You better believe these people constantly talk about the dangers *to children* rather than the secondhand dangers to adults, because they know if they made the real and valid arguments you suggest, they'd lose, and also they've been misled and don't understand the kids are actually safe. Also the whole 'no physical model' thing is real.
There's definitely a valid 'I don't want my kid getting my parents sick and we all live together' argument for some parents. I don't think it's strong enough to carry the day in the past and it's definitely not now with Omicron, I don't think it's close. If the few people who live in such situations wanted to form their own class with special rules, I'd think that was at least not insane, but that would be a different proposal.
Charity is good but it's important to not confuse what arguments are actually getting made with the ones that are valid.
I'm so confused about what arguments are actually getting made that I'm not sure describing even what I referenced as a "motte/bailey" isn't the same kind of 'confused charity' that I agree my own attempt was (if much more mild). If one is arguing 'from authority', why bother 'retreating' at all? (What would the 'motte' be anyways in those cases?)
I think I 'reach for charity' in part to counter what seems like (otherwise) likely ubiquitous 'distributed hypocrisy'. Among the individuals I follow closely enough to track actual 'arguments' (e.g. you), charity isn't needed anyways. I am also _extremely_ sympathetic to the significant demands needed to try to think about any of this clearly. I'm relatively sure most people are mostly trying to _avoid blame_ than take calculated risks, let alone anything more sensible.
I don't think you're doing the m/b thing, I'm saying mask advocates sometimes do it, but I agree on reflection it's less common/central than I thought when I wrote the comment. m/b is "Masks protect kids because children are our future and we must protect them" vs. "Masks protect kids because kids can then infect adults" or "Any additional infection results in exponential growth" which in turn is a stronger argument than that.
The charity of 'these people aren't thinking clearly' I'm totally on board for, to extent it counts as charity. Parents who think their children are in serious danger deserve our sympathy.
I think it's useful to focus on how the phrase "herd immunity" is used. I remember from times before Covid that we talked about herd immunity in the context of the protection afforded to the vulnerable people (immunocompromised individuals, the frail and the infirm) by presence of levels of natural or artificially-acquired (i.e. vaccinated) immunity among the population as a whole, sufficient to prevent uncontrolled, outbreak-type spread. Herd immunity doesn't mean a complete immunity of all people in the community, only enough immunity to prevent epidemic, uncontrollable spread. For pathogens such as whooping cough, where vaccine immunity is long-lasting, the preferred means of achieving herd immunity is vaccinating enough children to prevent outbreaks. For pathogens such as coronaviruses, which are commonly endemic and tend to produce frequent but mild reinfections (as it may be eventually the case with Covid-19) it means doing nothing in particular about infections among the healthy and young but it may mean special precautions around the Covid-vulnerable persons, which is exactly what the Declaration advocates. In other words, the Declaration is actually conventional epidemiological wisdom, applied to a new pathogen that seems to follow a trajectory similar to multiple previous pathogens of this kind (e.g. the Russian Flu of 1889). Almost everybody eventually gets infected and re-infected, multiple times, by most major strains of common viruses (rhinovirus, RSV, coronaviruses) and trying to prevent that just is not reasonable. But it is reasonable to protect the sick elderly and to develop better vaccines.
This is just old-fashioned, epidemiological common sense. That it is seen as a dangerous and abominable heterodoxy is a testament to how far off the beaten path we are now.
That was pretty much my take. Since people can't be prevented from getting the virus, your next best option is to just allow people to get semi-annual natural boosters via minor infections. That way all the infections are minor due to limited immunity die off, and transmission to those who really need to avoid contracting it is limited because minor infections are less likely to be able to transmit a more infectious viral load. Essentially, getting minor infections one or two times a year has the same effect as getting the booster vaccines.
I don't know if herd immunity is the right term for that sort of thing, but other than that question it makes sense.
Part of the reason for the weird stabilisation of case numbers is that cases in children are rising rapidly while falling rapidly in older adults. This is somewhat predictable from the timings of school holidays which have delayed the peak in children. The stabilisation in middle aged adults could be increased cases in parents, but could be something else too.
It is hard to make sense of the comment by Sunetra Gupta. It does remind me of https://www.nitag-resource.org/sites/default/files/e361266bc0e478ac95c2bcf4bc698b40186f2206_2.pdf which is an official statement by the UK's JCVI (who effectively set vaccine policy) explaining that chickenpox vaccine will not be provided to children as infections in children are needed to boost the immunity of adults (paragraph 10). Similar logic could be applied to conclude that having a continual high number of SARS-Cov-2 infections will reduce levels of severe Covid. Maybe Gupta meant "the low risk of severe disease among the vulnerable arises from the maintenance of immunity through constant reinfection", which might be true.
Was hoping to get your take on BA.2 omicron variant. I'm picking up bits and pieces of info finding it hard to synthesize into a story. The usual "no evidence of any increased virulence" vs "the sky is falling again". This is the kind of thing I look towards you to research for me.
Mike Makowsky should be a lot more charitable towards Peterson regarding the value of models. Mathematical economics models have done very poorly over the past century, bordering on possibly being a huge dead end. The replication crisis is maybe not so bad as in psych or sociology, but it is not going well. It has been a few years since I talked with Mike, but I am pretty sure he knows this. It is true that it is not impossible to find true things, but that is a LONG way from saying that economic modeling has produced much replicable truth. I know a lot has happened since 2008, or 2001, or the last time someone said "we have tamed the business cycle" with a straight face, but advanced modeling techniques haven't produced much truth.
Peterson definitely seems to me to be doing the move where (1) you're trying to build a model. (2) the model can't include everything (3) therefore your model is useless, and also (4) all models are always useless, it proves too much and is a superweapon against anyone claiming to know anything about anything. If you instead want to say (5) models have a bad track record and we should be skeptical, that's different, but that's definitely NOT what was being said here.
I haven't seen more than the 49 seconds included there in Mike's Twitter, so I am not able to really address Peterson's argument. It seemed to me he was saying climate models are by definition a model of everything, so not just temperatures here or there or rainfall, "weather isn't cllimate" etc., but everything, but since the modelers leave out lots of variables by necessity the models are not going to work, presumably because the choices of which variables to include lets you optimize for whatever outcome you want the model to show. That's my guess from 49 seconds.
However, if he is making an argument of the form "Models are a LOT less useful than the scientists that make a career publishing them claim", which is what Mike seems to be arguing to be "scientific nihilism", then Peterson has a point. Mike moves from scientific nihilism as believing we can't know anything to defining it as not thinking econometric techniques lead to real knowledge and useful models. That is quite a move. Add in the assertion that people just don't like using the mathematical methods because they are not smart or hard working enough to understand the math, and the bit about how being a nihilist that doesn't like math can be a useful career gambit instead of being a proper mathematical economist, and yea, I have heard this dozens of times before at economics conferences. Usually by DSGE modelers when asked "So... why do you guys keep doing this when your models are consistently unable to generate useful predictions?"
TL;DR Peterson might go off the rails later in the talk, I don't know. Makowsky definitely does, however, equating scientific nihilism with not thinking econometric modeling is a valuable way to spend your time, ignoring the fact that most econometric models give bad results, i.e. lie to you about how the world works. Scientific nihilism might be a cornerstone of bad economics, but believing you know far more than you do about how the world works and just what levers to pull to make things happen based on models is pretty much the rest of the edifice.
Anecdotes are not data, but: what I'm hearing from colleagues in India is that official numbers are under-reporting the physical reality.
One observation was that when your whole family comes down with Covid during the Omicron surge, it's obvious that's what's happened. Perhaps you confirm with an at-home test, or perhaps you don't have access to one; either way, you don't get officially tested, and so your infections aren't counted as cases.
I agree with both you and Anon that it's likely India is underreporting by a lot, but the underreporting should be somewhat proportional - if they have X reported cases now and X reported cases in a week, it's odd for that to be true while they went from Y actual cases to 4Y actual cases during that time, as it implies a 75% reduction in rate of reporting cases, which only makes sense if there's a cap but we know that the cap is a lot higher than this because of last time.
During the Delta surge, there was more concern about confirming cases, in part because one family member would show potential symptoms and want to know if they should isolate from the rest of the household. With Omicron, it seems like it's more common that several members of a household get it all at once — at which point they've given up on containing it within the household and there's less reason to test.
A different person mentioned being scared of Delta, but viewing Omicron as mild. They didn't link that to testing, so maybe I'm reading too much into it, but I wonder if that's a factor — if you're not worried about what you've got, you might be less likely to go get a test.
In other words, it seems like changes in attitude towards Covid could have an effect on testing habits, which turns the "under-reporting factor" of the equation into a function of time instead of a constant.
I don't know if this helps you understand the Indian conundrum, but IMO, as someone who lives here, the data is extremely untrustworthy. In many places governments are dictating what the data should say for whatever ridiculous policy they want to justify. Tests are often/typically conducted by people who come home to collect samples, and they will agree to give you the test result you want for travel etc. The whole thing is a shit show.
As someone who also lives here (India). Your point is incorrect. It isn't govt mandate that is showing less cases. The cases are less because people are not opting to take tests when they see symptoms. They know what it is (e.g. one person in the family impacted) and they have also seen that it gets fine in a few days. In my organisation with 400 people, there is 50-50% split of people who had tested for covid and recovered and who recovered from viral/flu and never bothered to test. These are educated / well paid people.
Tests in India are easily available and there is no large scale fudging as you are indicating. There is a lot to blame this country for but faking COVID numbers is the last thing. And faking deaths is even harder. Just remember the chaos and crisis of second wave. This wave is indeed less virulent and people are rightly taking the call to live with it / avoid the obvious test based on their risk reward equation.
Just out of curiosity… I really find your writing helpful to understand the immediate considerations of acute illness of Covid-19 but to provide a really accurate analysis and assessment of risks and what constitutes, for example, a safe time to call it over, shouldn’t you include observations on Long Covid?
Precautions and measures to curb the spread are not simply to reduce deaths and hospitalisations but to reduce the number of healthy people whose lives will be completely turned on their head as they are debilitated by the long term illness.
I do not feel you have ever really factored this in even though it a very real and important factor in any pandemic narrative.
Governments continue to ignore the size of the problem- focussing short term and not long term economic impact it will have with millions debilitated and unable to work.
There are now over 30 studies and trials into treatments for Long Covid so the Pharma world understands the size of the problem- it will be years before many people see the treatments available to them and there’s no certainty of a cure.
I was intending to do The Long Long Covid post this week but life interfered. I absolutely have thought about this question and factor it in, and have said more than once it's the majority of the risk to the young and healthy. But I continue to think that the idea of lots of lives being 'turned on their head' is simply not happening. I do not believe millions will be unable to work.
Nor do I think 'curb the spread' is a thing and I'm increasingly lacking patience for those who say we must do it when there's clearly no way to do it.
Well it depends on what you mean by “lots”. If you mean millions of people then you’re dead wrong because there’s already millions of lives that have been turned on their heads. (We’ve not even seen the long haulers from omicron yet…)
Your writing is great but please try to dig a little into the scale of long covid and show an awareness of it to provide a valuable and comprehensive snapshot.
You can always curb the spread- people in Asia have worn masks since previous outbreaks to do just that. Ventilation systems and masks will help…. And the reason for that is to protect the people (millions) that need protecting. At least until treatment is available to protect those who are susceptible (including the healthy young people and kids who just don’t know what a debilitating illness is coming their way)
If you can’t tell I am a long hauler and if there was more information about the very real risks to healthy people a year ago I may have still caught it but I’d have damn sure been far more careful.
The criticisms of Peterson’s statement are valid, but there is a sliver of a valuable point there too. When activists and scientists discuss climate models as an ironclad consensus from on high, they overstate their case. The problem does suffer from the high dimensionality Peterson alludes to and the typical person on the left would benefit from considering the necessary subjectivity that this introduces into the science.
Admittedly the typical person on the right is more likely to encounter this conversation and requires a different kind of counterbalance, but at least as the conversation goes on Peterson appeals to the various degradations caused by water pollution and the potential to address them feasibly.
Official warning for making my life worse.
Pointing out what you think is an error is fine, as is saying you'd prefer if I relied less on Twitter, but you can do it politely. Mistakes will occasionally happen, that's the speed premium.
In addition, your comment here assumes that Everybody Knows all that and what is and isn't reliable, and why, and implies that anyone who disagrees is an idiot, and that's also Not Ok.
If other commenters agree that Lin is wrong on the substance, I will update and edit, but since it doesn't have huge practical implications and I don't trust your confidence for now I'm not going to look into it further and rely on the other eyes on this.
If others share this sense that this particular wording ('making my life worse') is having this particular effect I encourage them to speak up. If they do, I'll consider choosing different wording that means the same thing. If I don't hear collaboration on this, I'm going to assume it's fine.
In either case, I'm confident you got the message I was trying to send.
Hey, I'd like to add a second vote for fewer twitter links (minus the original vitriol). The reason is because I usually read your blogs at school and the school wifi blocks twitter for...reasons. Really enjoying the blog though--nice to see another both magic the gathering player and rationalist.
If the issue is twitter being blocked then that's an argument for using more direct links instead, which seems reasonable, but Studies Show (e.g. the stats) that the links are more to show you have them, and rarely get checked. Still, good note. Different mechanism than above, and isn't an arguments against e.g. quoting Twitter.
Trusting anything you see on Twitter by default is obviously a terrible heuristic, on that we agree. One must use such things responsibly, and know from context how much trust to have in any given statement, while keeping in mind the correct rate of getting it wrong is usually not zero.
The idea that it would be a trivial additional amount of time/work to 'validate these facts' is in my experience very wrong, even setting aside the fact that you have chosen sources you think are trustworthy and that validate facts, and that others would strongly disagree with those choices you have made. It would double the amount of time at least, and is basically saying *at least* 'hire someone to fact check like a newspaper would do.'
Instead, I do my best to calibrate and convey my level of trust in a given claim (and if it's not high enough, not use the source). It's all we can do.
The idea that these posts could be produced without Twitter, and that I would be better off avoiding it, is simply wrong. It wouldn't only take longer, it's not possible. There's a reason no one else is doing a similar thing (at least that I am aware of).
1) Your Local Epedimiologist this week stated "There is zero evidence that masks harm, physically or psychologically, kids."
2) There is no evidence that I am subscribed to YLE any longer.
3) Annecdotally, I've seen that masks-blocking-non-verbal-communicatoin is a pretty huge burden on people with Autism, and for boys more than girls. Most that I'm interacting with are middle-school aged and their parents. No one has ever studied this, so of course there is No Evidence on this either.
1) Yeah, I have no idea how people type that with a straight face. Definitely not if they have kids.
2) I'd say there's strong evidence against that hypothesis, but there's also some evidence in favor, which is that you seem to have previously been reading it, and implied that you previously subscribed. So there! Bayes ruled.
3) Interesting note, thank you. That makes sense, I could see it going either way (e.g. maybe they need the extra info, or maybe they don't know what to do with the extra info).
1) "It Is Difficult to Get a Man to Understand Something When His Salary Depends Upon His Not Understanding It"
2) Damn, got me.
3) Like many autism features, some people have it and some people don't. I know one little girl who has never had any trouble interpreting facial expressions, and masks don't bother her beyond the typical annoyances. Her brother on the other hand really is in perminant midnight. I have an entire TED talk in me about how autism isn't a spectrum, its a basket of goods.
Isn't DeSantis' point that while Omicron might be the dominant strain, Delta didn't just vanish like throwing a light switch? (The latest numbers I've seen say about 1.6% of Covid cases are still Delta in the US.) So why deny people unlucky enough to catch Delta instead of Omicron a treatment that works? Especially since Omicron's lesser severity means Delta will be over-represented in hospitalized cases relative to general prevalence.
To clarify, suppose that either Delta or Omicron were not a coronavirus. Would this policy make sense? Would you take away the rabies vaccine because we were in an influenza epidemic?
Probably not. So why take away a treatment some of the Covid victims could profit from, just because most wouldn't?
That arguments holds water if and only if you can identify the remaining Delta cases as Delta cases in time to give them the right treatments, and solve the logistics issues involved, and avoid giving it to Omicron cases. If that's true, then DeSantis has a (small and rapidly shrinking) point. But the new nowcast is 0.1% I think, not 1.6%, and even at 1.6% I doubt they'd bother testing for Delta.
Does it matter if it is the right treatment though? If you are in a scenario where it is life or death for a patient and you can't take the time to determine if they have omnicron or delta, give them the monoclonal antibodies just in case. The side effects of monoclonal antibody treatment are extremely minimal [1], so why not just give it even if it may not be the correct treatment (ie, they actually have Omnicron but you administer the monoclonal antibodies for Delta). If you have Delta, great, your chances of survival just increased dramatically, if you don't, you may have minor side effects. Obviously, this is not ideal, and you could evaluate what a patient has prior to prescribing treatment, but that is not the scenario presented.
Edit: Note, I did a cursory search for monoclonal antibody treatment adverse effects, and this seemed to be the best source. So I really am curious, as I just hadn't heard much about the safety profile, and never have had to seriously consider taking them. Perhaps there is a significant risk. And I should add the hypothetical argument goes against one of my core beliefs (which can probably rightfully be called out for being foolish) which is I personally strive to avoid medical interventions unless absolutely necessary, so the idea of shotgunning extra treatment just in case is anathema to me, but perhaps not for most people/ particularly scared people in a hospital with covid.
[1]Brobst B, Borger J. Benefits And Risks Of Administering Monoclonal Antibody Therapy For Coronavirus (COVID-19) [Updated 2021 Dec 3]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2022 Jan-. Available from: https://www.ncbi.nlm.nih.gov/books/NBK574507/
On being post covid: I'm planning on waiting another week and a half and then restarting NYC rationalist weekly meetups - I want it to be very unambiguously over, and there are people who are more covid hawkish than me. Hopefully by then it'll be clear enough that we can agree to start being officially post-covid.
On predictions: mostly agree with your updates, but I'd buy airline mask requirements to 40% on the basis of "airlines love keeping dumb regulations" and keep the probability of omicron optimized boosters at 10% (I was at 8% before, so updating upwards but by less than you, and mostly as a response to your update rather than a direct response to the 5 month thing).
I don't see the airlines as motivational here, they will drop the mask requirement as soon as they're told they're allowed to do that. It's more that the FAA/TSA and related things love regulating airlines and think they're a special magisterium. I think if you're 40% now you should have been at 50%+ earlier. Which is reasonable.
And yeah, I agree we should restart meetups soon and 2/8 seems like a reasonable time for it. I notice I'd be willing to host then if there was need.
I'm confused about your discussion of schools and kids/students. I know that _some_ parents of school-age kids _are_ concerned about the kids themselves being infected and falling ill (or otherwise being harmed) – I regularly interact with one in particular.
But the, IMO, charitable interpretation around the general concerns of schools and students isn't that the kids themselves are at any significant risk but that everyone else they interact with outside of school is. The particular parent I mentioned has told me that their child has been sick several times already this winter (not from COVID, or at least negative on all tests), and all the anecdotal evidence I've ever observed has been that kids, and 'schools' (i.e. anywhere large numbers of kids interact with each other for extended periods), are probably the most common means of transmission for all kinds of contagious diseases.
I commented on an earlier post that the same parent's kid participated in a school music performance. That the school did this didn't and doesn't make any sense to me – in combination with all of the other precautions, e.g. mandatory masking at all times, even outside, and all classroom windows being open in the winter. I have greater appreciation of the thinking of the 'bubble person' you mentioned – they at least seem to have a 'physical model' and to be seriously cautious about being infected, whereas (most) everyone else, including the particular parent I mentioned, seems to be deluding themselves that whatever official precautions are mandated _must_ be (almost) perfectly reasonable.
Motte/bailey. You better believe these people constantly talk about the dangers *to children* rather than the secondhand dangers to adults, because they know if they made the real and valid arguments you suggest, they'd lose, and also they've been misled and don't understand the kids are actually safe. Also the whole 'no physical model' thing is real.
There's definitely a valid 'I don't want my kid getting my parents sick and we all live together' argument for some parents. I don't think it's strong enough to carry the day in the past and it's definitely not now with Omicron, I don't think it's close. If the few people who live in such situations wanted to form their own class with special rules, I'd think that was at least not insane, but that would be a different proposal.
Charity is good but it's important to not confuse what arguments are actually getting made with the ones that are valid.
I'm so confused about what arguments are actually getting made that I'm not sure describing even what I referenced as a "motte/bailey" isn't the same kind of 'confused charity' that I agree my own attempt was (if much more mild). If one is arguing 'from authority', why bother 'retreating' at all? (What would the 'motte' be anyways in those cases?)
I think I 'reach for charity' in part to counter what seems like (otherwise) likely ubiquitous 'distributed hypocrisy'. Among the individuals I follow closely enough to track actual 'arguments' (e.g. you), charity isn't needed anyways. I am also _extremely_ sympathetic to the significant demands needed to try to think about any of this clearly. I'm relatively sure most people are mostly trying to _avoid blame_ than take calculated risks, let alone anything more sensible.
I don't think you're doing the m/b thing, I'm saying mask advocates sometimes do it, but I agree on reflection it's less common/central than I thought when I wrote the comment. m/b is "Masks protect kids because children are our future and we must protect them" vs. "Masks protect kids because kids can then infect adults" or "Any additional infection results in exponential growth" which in turn is a stronger argument than that.
The charity of 'these people aren't thinking clearly' I'm totally on board for, to extent it counts as charity. Parents who think their children are in serious danger deserve our sympathy.
I signed the Great Barrington Declaration today.
I think it's useful to focus on how the phrase "herd immunity" is used. I remember from times before Covid that we talked about herd immunity in the context of the protection afforded to the vulnerable people (immunocompromised individuals, the frail and the infirm) by presence of levels of natural or artificially-acquired (i.e. vaccinated) immunity among the population as a whole, sufficient to prevent uncontrolled, outbreak-type spread. Herd immunity doesn't mean a complete immunity of all people in the community, only enough immunity to prevent epidemic, uncontrollable spread. For pathogens such as whooping cough, where vaccine immunity is long-lasting, the preferred means of achieving herd immunity is vaccinating enough children to prevent outbreaks. For pathogens such as coronaviruses, which are commonly endemic and tend to produce frequent but mild reinfections (as it may be eventually the case with Covid-19) it means doing nothing in particular about infections among the healthy and young but it may mean special precautions around the Covid-vulnerable persons, which is exactly what the Declaration advocates. In other words, the Declaration is actually conventional epidemiological wisdom, applied to a new pathogen that seems to follow a trajectory similar to multiple previous pathogens of this kind (e.g. the Russian Flu of 1889). Almost everybody eventually gets infected and re-infected, multiple times, by most major strains of common viruses (rhinovirus, RSV, coronaviruses) and trying to prevent that just is not reasonable. But it is reasonable to protect the sick elderly and to develop better vaccines.
This is just old-fashioned, epidemiological common sense. That it is seen as a dangerous and abominable heterodoxy is a testament to how far off the beaten path we are now.
That was pretty much my take. Since people can't be prevented from getting the virus, your next best option is to just allow people to get semi-annual natural boosters via minor infections. That way all the infections are minor due to limited immunity die off, and transmission to those who really need to avoid contracting it is limited because minor infections are less likely to be able to transmit a more infectious viral load. Essentially, getting minor infections one or two times a year has the same effect as getting the booster vaccines.
I don't know if herd immunity is the right term for that sort of thing, but other than that question it makes sense.
A couple of thoughts from a UK perspective.
Part of the reason for the weird stabilisation of case numbers is that cases in children are rising rapidly while falling rapidly in older adults. This is somewhat predictable from the timings of school holidays which have delayed the peak in children. The stabilisation in middle aged adults could be increased cases in parents, but could be something else too.
It is hard to make sense of the comment by Sunetra Gupta. It does remind me of https://www.nitag-resource.org/sites/default/files/e361266bc0e478ac95c2bcf4bc698b40186f2206_2.pdf which is an official statement by the UK's JCVI (who effectively set vaccine policy) explaining that chickenpox vaccine will not be provided to children as infections in children are needed to boost the immunity of adults (paragraph 10). Similar logic could be applied to conclude that having a continual high number of SARS-Cov-2 infections will reduce levels of severe Covid. Maybe Gupta meant "the low risk of severe disease among the vulnerable arises from the maintenance of immunity through constant reinfection", which might be true.
Was hoping to get your take on BA.2 omicron variant. I'm picking up bits and pieces of info finding it hard to synthesize into a story. The usual "no evidence of any increased virulence" vs "the sky is falling again". This is the kind of thing I look towards you to research for me.
My lack of mentioning it is because I am on the side that the sky is very much not falling.
Mike Makowsky should be a lot more charitable towards Peterson regarding the value of models. Mathematical economics models have done very poorly over the past century, bordering on possibly being a huge dead end. The replication crisis is maybe not so bad as in psych or sociology, but it is not going well. It has been a few years since I talked with Mike, but I am pretty sure he knows this. It is true that it is not impossible to find true things, but that is a LONG way from saying that economic modeling has produced much replicable truth. I know a lot has happened since 2008, or 2001, or the last time someone said "we have tamed the business cycle" with a straight face, but advanced modeling techniques haven't produced much truth.
Peterson definitely seems to me to be doing the move where (1) you're trying to build a model. (2) the model can't include everything (3) therefore your model is useless, and also (4) all models are always useless, it proves too much and is a superweapon against anyone claiming to know anything about anything. If you instead want to say (5) models have a bad track record and we should be skeptical, that's different, but that's definitely NOT what was being said here.
I haven't seen more than the 49 seconds included there in Mike's Twitter, so I am not able to really address Peterson's argument. It seemed to me he was saying climate models are by definition a model of everything, so not just temperatures here or there or rainfall, "weather isn't cllimate" etc., but everything, but since the modelers leave out lots of variables by necessity the models are not going to work, presumably because the choices of which variables to include lets you optimize for whatever outcome you want the model to show. That's my guess from 49 seconds.
However, if he is making an argument of the form "Models are a LOT less useful than the scientists that make a career publishing them claim", which is what Mike seems to be arguing to be "scientific nihilism", then Peterson has a point. Mike moves from scientific nihilism as believing we can't know anything to defining it as not thinking econometric techniques lead to real knowledge and useful models. That is quite a move. Add in the assertion that people just don't like using the mathematical methods because they are not smart or hard working enough to understand the math, and the bit about how being a nihilist that doesn't like math can be a useful career gambit instead of being a proper mathematical economist, and yea, I have heard this dozens of times before at economics conferences. Usually by DSGE modelers when asked "So... why do you guys keep doing this when your models are consistently unable to generate useful predictions?"
TL;DR Peterson might go off the rails later in the talk, I don't know. Makowsky definitely does, however, equating scientific nihilism with not thinking econometric modeling is a valuable way to spend your time, ignoring the fact that most econometric models give bad results, i.e. lie to you about how the world works. Scientific nihilism might be a cornerstone of bad economics, but believing you know far more than you do about how the world works and just what levers to pull to make things happen based on models is pretty much the rest of the edifice.
Anecdotes are not data, but: what I'm hearing from colleagues in India is that official numbers are under-reporting the physical reality.
One observation was that when your whole family comes down with Covid during the Omicron surge, it's obvious that's what's happened. Perhaps you confirm with an at-home test, or perhaps you don't have access to one; either way, you don't get officially tested, and so your infections aren't counted as cases.
I agree with both you and Anon that it's likely India is underreporting by a lot, but the underreporting should be somewhat proportional - if they have X reported cases now and X reported cases in a week, it's odd for that to be true while they went from Y actual cases to 4Y actual cases during that time, as it implies a 75% reduction in rate of reporting cases, which only makes sense if there's a cap but we know that the cap is a lot higher than this because of last time.
Slighty more annecdotal information:
During the Delta surge, there was more concern about confirming cases, in part because one family member would show potential symptoms and want to know if they should isolate from the rest of the household. With Omicron, it seems like it's more common that several members of a household get it all at once — at which point they've given up on containing it within the household and there's less reason to test.
A different person mentioned being scared of Delta, but viewing Omicron as mild. They didn't link that to testing, so maybe I'm reading too much into it, but I wonder if that's a factor — if you're not worried about what you've got, you might be less likely to go get a test.
In other words, it seems like changes in attitude towards Covid could have an effect on testing habits, which turns the "under-reporting factor" of the equation into a function of time instead of a constant.
I don't know if this helps you understand the Indian conundrum, but IMO, as someone who lives here, the data is extremely untrustworthy. In many places governments are dictating what the data should say for whatever ridiculous policy they want to justify. Tests are often/typically conducted by people who come home to collect samples, and they will agree to give you the test result you want for travel etc. The whole thing is a shit show.
As someone who also lives here (India). Your point is incorrect. It isn't govt mandate that is showing less cases. The cases are less because people are not opting to take tests when they see symptoms. They know what it is (e.g. one person in the family impacted) and they have also seen that it gets fine in a few days. In my organisation with 400 people, there is 50-50% split of people who had tested for covid and recovered and who recovered from viral/flu and never bothered to test. These are educated / well paid people.
Tests in India are easily available and there is no large scale fudging as you are indicating. There is a lot to blame this country for but faking COVID numbers is the last thing. And faking deaths is even harder. Just remember the chaos and crisis of second wave. This wave is indeed less virulent and people are rightly taking the call to live with it / avoid the obvious test based on their risk reward equation.
Just out of curiosity… I really find your writing helpful to understand the immediate considerations of acute illness of Covid-19 but to provide a really accurate analysis and assessment of risks and what constitutes, for example, a safe time to call it over, shouldn’t you include observations on Long Covid?
Precautions and measures to curb the spread are not simply to reduce deaths and hospitalisations but to reduce the number of healthy people whose lives will be completely turned on their head as they are debilitated by the long term illness.
I do not feel you have ever really factored this in even though it a very real and important factor in any pandemic narrative.
Governments continue to ignore the size of the problem- focussing short term and not long term economic impact it will have with millions debilitated and unable to work.
There are now over 30 studies and trials into treatments for Long Covid so the Pharma world understands the size of the problem- it will be years before many people see the treatments available to them and there’s no certainty of a cure.
I was intending to do The Long Long Covid post this week but life interfered. I absolutely have thought about this question and factor it in, and have said more than once it's the majority of the risk to the young and healthy. But I continue to think that the idea of lots of lives being 'turned on their head' is simply not happening. I do not believe millions will be unable to work.
Nor do I think 'curb the spread' is a thing and I'm increasingly lacking patience for those who say we must do it when there's clearly no way to do it.
Well it depends on what you mean by “lots”. If you mean millions of people then you’re dead wrong because there’s already millions of lives that have been turned on their heads. (We’ve not even seen the long haulers from omicron yet…)
Your writing is great but please try to dig a little into the scale of long covid and show an awareness of it to provide a valuable and comprehensive snapshot.
You can always curb the spread- people in Asia have worn masks since previous outbreaks to do just that. Ventilation systems and masks will help…. And the reason for that is to protect the people (millions) that need protecting. At least until treatment is available to protect those who are susceptible (including the healthy young people and kids who just don’t know what a debilitating illness is coming their way)
If you can’t tell I am a long hauler and if there was more information about the very real risks to healthy people a year ago I may have still caught it but I’d have damn sure been far more careful.
Click through rates are typically very low, so if I think the thread is necessary I try to paste in the required parts via snipping tool.
The criticisms of Peterson’s statement are valid, but there is a sliver of a valuable point there too. When activists and scientists discuss climate models as an ironclad consensus from on high, they overstate their case. The problem does suffer from the high dimensionality Peterson alludes to and the typical person on the left would benefit from considering the necessary subjectivity that this introduces into the science.
Admittedly the typical person on the right is more likely to encounter this conversation and requires a different kind of counterbalance, but at least as the conversation goes on Peterson appeals to the various degradations caused by water pollution and the potential to address them feasibly.
Couldn’t they be discussing something like subinfectious exposures (that’s supposed to prevent immunity from from fading). I know this is MDPI, but still: https://mdpi-res.com/d_attachment/viruses/viruses-13-00961/article_deploy/viruses-13-00961.pdf