This is a thread to list important insights and key open questions about the coronavirus and the coronavirus response. The inspiration for this thread is Eliezer’s post below.
I’d like this thread to be a source of claims and ideas that are self-contained and well-explained. This is not a thread to drop one-liners that assume I’ve been following your particular news feed or know what’s happening in your country or that I’ve read a bunch of studies on (say) viral load. There’s a place for such high-context discussion, and it is not this thread.
Please include in your answers either a claim or an open question, along with an explanation or an explicit model under which it makes sense. I will be moving answers to the comments if they don’t meet my subjective quality bar for justification – see the last justified answers thread for examples of what quality answers look like.
The purpose of giving models and data is to allow other people to build on your answer. Everyone can make arbitrary claims, but models and evidence allow for verification and dialogue.
The more concrete the explanation the better. Speculation is fine, uncertain models are fine; sources, explicit models and numbers for variables that other people can play with based on their own beliefs are excellent.
This thread is inspired by a post by Eliezer Yudkowsky which I’ll reproduce below, in which Eliezer lists eight answers that this sort of post would come up with.
These are not justified to the standard of the thread, so you (you!) can get some easy karma by leaving an answer that justifies one of these with the sources/data/explanation needed to argue for it. It includes much of the discussion elsewhere on LW (e.g. by Wei Dai, Zvi, Robin, and others), so it shouldn’t be hard to find the prior discussion.
Eliezer’s post (link):
What do we early-warning cognoscenti now know about Covid-19 that others haven’t currently figured out? What’s the TOC of that blog post? @WilliamAEden @robinhanson
My stab at a TOC:
1: The Dose Hypothesis—the theory that C19 fatalities vary by how high the initial dose, and possibly how it’s administered.
1a: So: Human trials of variolation are hugely urgent.
1b: So: Getting C19 from a roommate might be much worse than getting it on public transportation.
2: Challenge trials of vaccines save net lives.
3: Ventilators no longer look as important because they only save 15% of the patients on them.
4: There’s huge apparent variation in CFR by country, and explaining this, or explaining it away, seems kinda important.
4a: CFRs may be underestimated by up to 3-fold, based on looking at excess death rates year-over-year.
4b: CFRs may be overestimated because of too little testing.
5: There was a huge EMH failure w/r/t C19, and it hasn’t been explained away AFAIK.
6: Most of the economic damage from a real shock like this one is still due to the secondary demand shock, which can be prevented by decisive central bank action.
6a: We know the Fed isn’t currently doing enough here because inflation expectations are dropping, showing the AD shock exceeds the AS shock.
6b: Stock prices take into account the next 15+ years of earnings. The real C19 shock only damages the next 2 years of earnings. A financial recession would damage many more years. Stock prices mainly reflect central bank policy, not C19.
7: Face masks do work, though others seem to have mostly figured this out.
8: The mainstream media’s words on C19 may be best interpreted as not intended to mean things; like the way that MSNBC’s talk about Bloomberg being able to give each American over $1,000,000 can’t have had a concrete model of reality behind it.
Any items I’m missing here?
Claim: The true infection-to-fatality ratio is definitely about 0.5% to 1%, and most probably around 0.7%, with significant long term morbidity in at least several percent of survivors. Notions that this disease is already widespread or that it has flulike mortality and morbidity or most people are asymptomatic are definitively disproven.
This has been independently estimated in this range before, based on normalizing data from the Diamond Princess and areas where testing was thorough
https://www.thelancet.com/journals/laninf/article/PIIS1473-3099(20)30243-7/fulltext
https://www.medrxiv.org/content/10.1101/2020.03.05.20031773v2
There are a few robust new pieces of data supporting this now.
1 - Blanket RNA testing in Austria.
https://www.theguardian.com/world/2020/apr/10/less-than-1-of-austria-infected-with-coronavirus-new-study-shows
Given a 0.3% current acute infection rate and some epidemiological modeling they estimate 1% of their total population has been infected at some point, with a death rate of 0.77%. Maybe a few false negative PCRs, which would lower that number.
2 - Two serology surveys have now happened in Europe. One was in a hard-hit town in Germany, and one was in a hard-hit town in Italy at the epicenter of its outbreak. In both places, they got approximately a 15% seropositive rate. In Germany, we only have information on deaths with positive test results and it comes to 0.35%. In Italy, total excess deaths over this time last year are about 2.5x the confirmed positive deaths and account for 0.1% of the population, giving an infection fatality rate of 0.7%. It is easy to imagine that some deaths did not get positive tests in Germany which along with a less-old population could make up for the difference.
3 - New test data coming out of NYC.
https://www.nejm.org/doi/full/10.1056/NEJMc2009316
Hardly an unbiased sample, but of 200+ pregnant women coming into a hospital to give birth that were blanket-RNA-tested, 15.3% tested positive.
Of this set of positive tests, only 12% of them were symptomatic on admission, and a further 10% developed symptoms over the course of their 2-day-long stays bringing it to a total of 22% symptomatic upon discharge or transfer. Presumably already-symptomatic very-pregnant women were more likely to be in the hospital already.
Doing a little armchair epidemiology. Let’s assume that half of the deaths of currently infected people have happened, due to the lockdown extending the doubling time from three days to more than a week. We get:
~8000 deaths * 2 / (15.3% of 8 million) = 1.3% infection to mortality rate.
If we assume that there were more symptomatic women who didn’t show up to normal birthing due to going to the hospital for COVID symptoms, or that there is a good stock of people who have recovered in the city, we get a lower death rate. If 20% of the total population was ever infected, we get a 1% mortality rate. 30% ever infected, 0.67%.
EDIT: 4 - Apparently there is a similar maternity ward study in Stockholm, revealing 7% positive. There have been 550 deaths there, and a population of 2.3 million. If we again assume half of current cases that will die has died, we get a infection to fatality ratio of 0.68% without further corrections. I suspect they haven’t crushed the doubling time as much as NYC, raising this number, which then can get lowered down again as I did above.
EDIT: 5, a meta analysis of a whole bunch of research comes to exactly my original conclusion, 0.5% to 1% with a central tendency of 0.8%.
https://t.co/51b3bJYg3e?amp=1
UPDATE as of 4/23/2020.
I’m so sick of being right.
Seroprevalence in NYC reported as 21%.
https://twitter.com/NYGovCuomo/status/1253353516803993600
There can be false negatives, but at this positive level false positives are less of an issue than at low levels. Also, they apparently specifically grabbed people out and about at grocery stores, so very sick people may have been excluded, pushing levels down. On the other hand, people shopping might be more likely to pick it up.
Pretty much right on the nose...
https://twitter.com/trvrb/status/1253398329766973441
″ If we then take deaths as of today as 17,200 based on excess deaths (https://nytimes.com/interactive/2020/04/21/world/coronavirus-missing-deaths.html), we’d get an infection-to-fatality ratio of ~1%. ” I suspect the true seropositivity is higher than the measured due to selection effects on the net, which would push this down a bit.
EDIT: Apparently this test also only detect IgG, which is the type of antibody that rises last and can take two weeks or more to be detectable in some people after symptoms develop.
That would be IFR when the sick can be appropriate treated, right? I think it can be >>1% when hospitals are overwhelmed. It also obviously depends on demographics, prevalence of diabetes, etc.
Indeed! Anti-inflammatories, oxygen therapy, anticoagulants, blood pressure management, eventually invasive ventillation (though that seems less effective than was previously thought).
I suspect the United States will have a substantially higher IFR than Europe due to all the obesity and metabolic disease, and that when the ICUs pop it also rises.
New information. The Italian town of Vo was blanket-RNA-tested twice in late February and early March. A total of 3% of the town tested positive, and they were able to lock down this subset and shut down transmission from continuing in the town indicating they caught enough of the asymptomatic-but-transmissive carriers.
https://www.medrxiv.org/content/10.1101/2020.04.17.20053157v1
Here we have a detailed analysis of these positive-testing people.
43% asymptomatic all the way through.
~20% hospitalized. This means almost 40% hospitalization of people who were symptomatic. Critera for hospitalization is an interesting question, as is the age breakdown of the town.
One death, in a town in which a total of 82 people tested positive. That one death was what alerted authorities to the outbreak in the town, so we need to take it as a given rather than taking it as even weak evidence of an over 1% fatality rate.
No cases among hundreds of children, even in houses with symptomatic family members. Extra cases among the elderly (as in 1% of 20 year olds versus 6% of 70 year olds). Small numbers but significant. Unclear if that means they are not getting infected or are not producing long-lasting infection that is detectable or if social structure has something to do with it.
No obvious difference in symptomatic versus asymptomatic across the age distribution, subject to sample size.
More men took more than 2 weeks to clear the virus than women.
Definite confirmed asymptomatic people passing it on, and presymptomatic people passing it on.
3% of the town testing positive via PCR in the beginning of March is compatible with 15% of other towns in the area testing positive via serology a month later, as that would be less than 3 doublings. They find a ‘serial interval’ of only about 7 days in their contact-tracing data from before the lockdown, and 10 days after the lockdown, and a replication number before the lockdown of about 3 and 0.14 during the lockdown. That’s a doubling time of well under a week before the lockdown. The rest of Italy’s lockdown didn’t include such good contact tracing and thus still was probably doubling a few times during that period.
UPDATE as of 4/28/2019.
Others coming to this exact same distribution more rigorously.
https://t.co/51b3bJYg3e?amp=1
Compiling rigorous data, the compatible range is circa 0.5% to 1% with a central tendency of 0.8%.
Yeah, it’s been clear for some time that the IFR is about 0.5% and that about half the cases are asymptomatic. Give or take 30% on each. The reported variations are mainly due to testing or age/health condition bias.
Have you got a source for that ‘about half the cases are asymptomatic’? I was under the impression that far more cases show symptoms eventually, and that the studies showing half of the infections are asymptomatic add the disclaimer ‘so far’, which means very little if the spread is growing exponentially with a doubling time of several days.
The Diamond Princess was 50% asymptomatic after 2 weeks of [in cabin] quarantine. It would be nice to have more follow-up on it, but that’s already much better than most measures of asymptomatic cases.
The end data is that about 20% of the Diamond Princess was asymptomatic all the way through. They were particularly old, though.
In Iceland 50% are asymptomatic upon first test, then some progress. The data in Korea suggests 30% asymptomatic.
Small number statistics of government officials supports this too. 7 total congressmen have tested positive or been presumed positive. One or two were hospitalized, some said “it hit me hard” or “my case is mild”, and 2 out of 7 (Rand Paul and Joe Cunningham) reported either zero symptoms or only loss of the sense of smell.
Where do you get your data on the Diamond Princess? As far as I know, there are no updates on symptoms. Perhaps you get it from this, which is not data, but an inference?
Is there reason to believe the raw numbers are more accurate estimate of the rate than the model prediction? Also, what are the type-1 and type-2 errors of the tests used on the Diamond Princess? I heard some early reports that both of these might be significant, but then never heard anything about them again.
I checked that link above and followed their references to find other datasets, but two of them are in Japanese, one only deals with self-selected patients who showed symptoms, and the last two have small sample size (12 patients, two papers cover the same event).
Update: I have found https://www.eurosurveillance.org/content/10.2807/1560-7917.ES.2020.25.3.2000045, which benchmarks the real-time reverse transcription polymerase chain reaction (RT-PCT) tests. They state zero false positives in a trial with 297 non-COVID-19 samples, although they do retest 4 samples that showed “weak initial reactivity”. Since the non real-time version of RT-PCT is supposed to be even more reliable, this means false positives are presumably not a big deal (even at a pessimistic 4/297 false positive this still means only 41 false positives out of 3063 tests done on the Diamond Princess).
First of all, it is very important to distinguish data from inferences.
Second, the inference is idiotic. It’s probably a calculation error, but it’s just not worth reading to determine what went wrong.
I don’t have a good model to give me any predictions on what reasonable numbers of asymptomatic cases would be, or how truncation influences these numbers. Could you explain why the inference is idiotic, and perhaps give a more reasonable one?
Here are some quotes from the paper. What is the simplest model you can make from them? Forget the word “model”; what conclusions can you draw?
Check all the references from https://www.nature.com/articles/d41586-020-00885-w for some data, as well as worldometer.
I am trying to find the Japanese government webpage with frequent updates as to the state of patients that were evacuated, still updating the ones that are still in the hospital (! Morbidities...)
So, yes, it is simply misquoting the source that I cited.
Is there a good source for many things we know from the Diamond Princess data? Or even just the numbers so far from DP? I’m not sure how to find that data.
I’ve looked into this a lot and I agree strongly with this being a good range.
How do you draw that conclusion?
The effects of measures on the spread take weeks to show up in the data.
If the doubling time hadn’t cratered, the hospitalization rate would’ve remained exponential. At the time of posting it was comparatively flat, and I estimated.
The half came from the fact that it usually takes ~3 weeks to die, that the exponential spread had only stopped a few weeks earlier, and a drawing of a triangle and square representing a rise and flat that I drew a vertical line through.
Claims:
Severe cases should be treated with anticoagulants
Inhaled interferon, antivirals, and other effective treatments are probably much more effective when taken early to prevent the first few replication rounds.
A case is probably followed by a period of immune suppression, and possibly some T-cell immunity amnesia.
-----
The virus may be causing abnormal inflammation and a whole-body, but especially concentrated in the lungs, hyper-coagulable state that is triggering microscopic blood clots in the lungs that are one of the main contributors to morbidity and mortality and ineffectiveness of ventilation. Effective treatment of severe cases should probably include anticoagulants unless there are contraindications, and another effective treatment has been an interleukin inhibiting antibody normally reserved for severe arthritis. See the entire recent twitter diggings of @_ice9.
There have been major reports from clinicians that the lungs of COVID patients are not responding the same way as typical ARDS. The ability to get oxygen in the blood is too low compared to the amount of air they can get into them. Autopsies are revealing lots of small clots, and blood tests are finding the most predictive measurement of outcome is an indicator of blood clot dissolution (D-dimer).
Children with non-severe cases are having anomalously high rates of sores and discoloration on fingers and toes, indicative of diffuse coagulation in small vessels causing mild tissue damage that seems to heal on its own afterwards.
This hyper-coagulable state *might* explain the reports of anomalously low oxygen measurements in people that would ordinarily indicate death or unconscoiusness. They might have small clots in the finger the sensor is on triggering temporary sporadic low blood flow. It also could explain more of the fact that ventilators are less useful than they thought—some people going on them probably didn’t actually need them.
(Before anyone asks, that preprint that was making the rounds suggesting the virus was destroying hemoglobin was STUNNINGLY and EMBARRASSINGLY bad. Complete bullshit, not worth even hate-reading unless you find fantasy biochemistry from a universe in which chemical reactions have energies you normally associate with nuclear reactors and viruses do photosynthesis funny).
-----
Additionally, there are two bits of immunology that explain parts of this virus’s behavior and suggest ways of hurting it. First, the virus evolved in bats in which the interferon response is on an absolute hair trigger, and accordingly in human cells it almost completely escapes the interferon response. This allows it to replicate to absurd viral loads before the immune system notices it, explaining the extreme infectiousness shortly before symptoms develop. Then when the immune system notices it, it goes all out on a huge viral infection, triggering an inflammatory response that is all out of whack and can do a lot of damage. This means that it is vulnerable to inhaled interferon pretreatment (https://www.biorxiv.org/content/10.1101/2020.03.07.982264v1). On top of this, it may be that anything that reduces the replication of the virus in this period before the adaptive immune system mounts a robust response could reduce the probability of progression to severe disease. If antivirals work out or if chloroquine is effective (given the biochemistry I am very hopeful!), they will probably be most effective early via reducing the fraction of patients that progress to severe disease.
Second, there is evidence that the virus is able to enter and destroy (but not replicate within) T-cells using the same receptor it uses everywhere else, triggering immune suppression and altering the inflammatory profile (https://www.nature.com/articles/s41423-020-0424-9). It lacks HIV’s obscene dirty tricks and isn’t actually replicating within them, so this would be a temporary thing until recovery.
-----
That last bit may sound bad, but it is far from unique. When looking for other examples, one should look to the Measles virus. It too basically escapes the interferon response and grows to absurd highly-communicable levels (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC112268/), and it actually infects and replicates within T-cells and B-cells which it then rides throughout the body. This causes people who get the measles to basically forget 70% of their adaptive immune responses from before they were infected, and go through a period of immune suppression afterwards. I expect the loss of immune memory to be smaller in this case because the new bug doesn’t seem to infect B cells or all types of T-cells from what I have seen.
Here’s a paper that situationally agrees with you on anticoagulants… Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019 patients with coagulopathy
449 people. Specifically, they observed no difference in survival between heparin users/non-users overall, but in the very-high-D-dimer subset (or in people with lots of sepsis‐induced coagulopathy), survival seemed to be better with heparin.
This link carries no new information yet, but seems to be a placeholder for a future review paper on this topic.
The “blood coagulation as a major contributor to death” bit generally matches pretty well with some early results where high D-dimer predicted worse rates of mortality fairly reliably, since D-dimer is basically a problematic-blood-clot indicator.
There’s a potential complicating factor for the elderly, which is that many of them are already on anticoagulants (to mitigate stroke-risk). And going on some experiences of my grandparents, it seems to be hard to navigate the risks mitigated with anticoagulants with the risk of bleeding out unless you’re pretty careful. All the same, it looks like a promising line of improvement to treatment of severe COVID-19.
Nitpick: __ice9 has 2 underscores, not 1.
As for why blood clots are a problem in the first place… one of the hypotheses I’ve seen floating around is that it might be tied into complement system malfunction?
Warning that this is pretty speculative...
The complement system is an immune response that uses C-protein complexes to poke holes in membranes to kill cells and fight large infections.
This paper used results from 5 lung autopsies and tried to draw a link between the prolonged procoagulant state in the lungs with excessive activity of the complement system. I could barely follow it beyond that.
I had also heard before that complement system malfunctions were thought to be connected to bad vaccine response for SARS-1.
I don’t feel certainty in this at all. But it comes up semi-consistently, and I don’t have a better theory yet.
I think it’s a horse, not a zebra. ACE2 is expressed on endothelial cells lining blood vessels. If you get bad viremia the inner sheath of blood vessels, especially in heavily infected organs, probably just gets all messed up.
Hi everyone. I’ve been reading Less Wrong and related material off and on for a few years, and I finally made an account and this is my first comment.
I also wondered why it took so long for the market to react in February to the likelihood of a pandemic, especially after cases were increasing in Italy and Iran suggesting it could easily spread worldwide. I’m not a big trader, basically just buying and holding index funds for the long term, and the only thing I did differently was to hold off buying more at the peak—instead waiting for what seemed like an inevitable fall before buying (although I didn’t wait long enough). After reading Wei Dai’s posts about his strategy during this time, it seemed clear that the EMH hadn’t performed well here and I wished I had been more confident in my own analysis beforehand.
Later, it occurred to me that there are relatively few individuals who trade at high enough volumes to actually influence market prices, most of whom work for financial institutions. I’ve never worked in that industry, but I imagine employees are subject to the same social pressures as any company, including a reluctance to act differently than what is culturally acceptable within the organization (especially to take a culturally unacceptable risk which may look foolish initially). Since financial firms generally believe in the EMH, it would be very difficult for an individual employee to act otherwise. A low or mid-level trader is probably authorized to make trades based on any financial news, but they probably can’t just go their manager and say “I think I know better than the entire world that this pandemic will be worse than expected and lower stock prices, and the EMH hasn’t priced it in yet.” Even if a few investors do trade on that assumption, they will initially be overwhelmed by algorithmic trading which would tend to revert prices back to the previous EMH-based equilibrium. The markets wouldn’t actually move until either 1. effects from the pandemic start to change actual financial data, and the trading algorithms begin to account for that, or 2. most of the top leaders of financial institutions become convinced prices should go down, and make it socially acceptable for their employees to sell large volumes of stock or change their trading algorithms accordingly.
Until then, there could be a temporary inadequate equilibrium where nobody who has the power to move market prices has a socially acceptable reason to do so. In this situation, rational individual investors who don’t face these organizational social constraints may be able to outperform the EMH.
Maybe a version of the EMH that takes this into account would be “stock prices accurately reflect all public information that is socially acceptable for high-volume traders to base trades on” or something similar.
Others have written here about similar thoughts on the EMH failure, including Matthew Barnett and Alex Shleizer, but I haven’t seen it proposed explicitly as a result of social pressures or incentives within organizations before, so I thought this could add to the discussion. As I mentioned, I don’t work in the financial industry, so I welcome the comments of those who have more relevant experience in the field.
Hm. This seems worth poking at: if a financial firm believes in the EMH, why would they be making trades at all?
My understanding of the EMH is that an oversimplified version is “you can’t beat the market, any public information is already priced in”. If you believe this version, you should just buy into low-fee index trackers.
A more sophisticated version is: “the market has lots of clever people trying to beat it. If you can beat those players, you can beat the market”. Under that version, it makes sense for a large financial firm to make trades despite believing in the EMH, because a large financial firm is exactly the kind of organization to have the resources to beat those people.
But under that version, it seems like the firm needs to have some way for people to signal “I think I know better than the entire world...” and make trades on that, because that’s the only way trades ever get made.
And my intuition-that-I-can’t-justify is that: if people couldn’t make trades in this specific case due to social reality within the firms being out of sync with actual observable reality, then that’s probably not confined to this one particular case; and we’d see individuals beating the market a lot more often than we do.
(This is very much a case of “I don’t actually know what I’m talking about.”)
Totally agree with your analysis, especially that large firms and individuals within them are quite clever and operating as described in your sophisticated version. They are the very people who put the “efficient” in the efficient market hypothesis! And I agree that every trade is essentially a statement that “I know better than the entire world...”, with real money on the line. Yet the combined actions of these clever people, who have huge incentives to beat the market and are used to trading with confidence that they know better than the entire world, seemingly did not move the market quickly enough, and at least some individuals were able to take advantage of this.
This gets to the crux of the argument—under what circumstances would this disconnect occur, and can others outside the industry recognize when it may be happening? It must be quite rare, otherwise as you mention we would see more evidence of individuals regularly beating the market. It might require several factors to come together at the same time, without which the usual efficiency will be maintained. Some thoughts along these lines, expanding on my earlier comment:
1. Traders might only be willing to bet confidently against the entire world in their field of expertise (analyzing financial data), but not based on data from other fields (pandemic predictions). They may think: there are likely others who know more about this topic than me, and the market hasn’t moved yet (or, the knowledge is priced in already), so how can I justify betting big on my hunch? Especially when I would have to explain it to my manager (or the board of directors) if I was wrong?
2. Related to the last point—loss aversion bias regarding one’s social status with the organization. If someone bets big based on pandemic data, and is right, they will gain some social status and probably a financial reward. But if they’re wrong, they will lose social status and might be at higher risk of being fired. If the potential loss outweighs the potential gain, the safe option would be to not bet on pandemic data, and only trade based on financial data (as would be socially expected within the organization) and face little or no social status penalty.
3. Algorithmic trading was probably based initially only on financial information, which had not turned for the worse yet, so any pandemic prediction-based trades would be overwhelmed by algorithmic trades as far as moving stock prices. This would last until algorithms started taking into account pandemic data and/or the actual financial data got worse due to pandemic effects.
I’m sure others could identify additional relevant factors at play here.
One thing I don’t have any calibration on is how big a trade would have to be to overcome this neutralizing effect of algorithmic trading on overall stock prices. For example, say there is an employee of a large firm who has the authority to risk 1% of the firm’s overall portfolio. If they switched that 1% from a fully long position to fully short, would the market stay lower, or would algorithmic trading revert it to the previous equilibrium? What if the CEO of the firm switched their entire portfolio from long to short? What if the CEO did that, and also announced publicly they had done so and their reasons why? At what point would we expect markets to actually change course?
This quote from Moral Mazes seems relevant to this earlier discussion, and may provide further understanding for why markets were slow to respond to the pandemic (emphasis mine):
In Feb/March, if the relevant financial institutions were going through such a behind-the-scenes process of “establishing the inevitability” of the pandemic before large market-moving decisions could be made, this could explain the apparent delay (and corresponding opportunity for the rational individual investor). One can imagine individuals within these firms feeling each other out—“This pandemic might turn into a big deal, huh?” “Yeah, but the boss hasn’t seemed too concerned yet, let’s give it another few days before we bring it up again”—before the consensus grew large enough where the decision became inevitable.
If this model is accurate, when would we expect to see these kinds of delays (and opportunities) in other situations? Here are some factors that may have contributed:
The early pandemic required integrating a lot of information outside the core areas of expertise of firms and their traders, leading to more uncertainty and a longer delay to reach consensus.
People are bad at extrapolating exponential growth (citation needed), and while some individuals within firms may have realized the implications right away, others may have thought their concerns were way overblown, again prolonging the time to reach consensus.
This was a rare event that had not occurred within anyone’s living memory, so there was no good frame of reference to fall back on, also increasing uncertainty.
I feel like there’s something here worth investigating more closely, although I’m still having trouble understanding it as well as I would like to. For now I’ll note that these three factors also seem very applicable to the current state of AGI development, and so may tie in with previous discussions such as this one.