What do the smartest people think?
There are many different types of superintelligence. Some people understand quantum mechanics. Some people are really good at chess. Some people can read Finnegan’s Wake.
[I’m tempted to say, “Some people understand supply and demand
Scott Alexander is really, really good at analysis. So good that he’s become famous for engaging in this activity, with a cult following all over the world. (In recent years, the new people I meet in Orange County are largely through Scott Alexander meet-up groups.)
Thus it would be interesting to see what would happen if Alexander took a deep dive into the Covid origins debate, and looked at all of the evidence that’s been presented by both the Lab Leak and Zoonosis advocates.
Now he’s done so. First in a post that looked at a highly publicized debate between the two sides, and now in a follow-up post that dealt with a wide range of comments on his first post. (For those who don’t know, his comment sections are very long and full of high quality observations. If the objection is not in his comment section, it’s probably not worth considering.)
In the end, Alexander comes down pretty close to where I am:
For now, I’m still at 90-10 zoonosis.
And like me, he gets frustrated with the game of whack-a-mole played by Lab Leak proponents:
I know this comments post won’t be the end of the story. I know that (just as with every other one of my posts, I’m not blaming origins debaters in particular here) someone’s going to go “Sure, Scott confronted 489 arguments. But he failed to confront the strongest argument against his case – this one obscure article in a Nepalese journal that nobody except me has ever heard of. That means he’s a bad-faith actor strawmanning everyone he disagrees with!” I know that someone will find some detail I’m wrong about and spam it all over Twitter with “Scott didn’t realize that an 91Q mutation is different from a ZY6 mutation, how can you ever trust anything he says?” And I know that next month, someone will come up with another SMOKING GUN! – and if I don’t respond to it immediately they’ll say I’m scared and know I’ve lost and am refusing to admit I’m wrong out of sheer stubbornness, and twist some quote of mine to show I’ve admitted I’ve changed my mind.
You need to devote a considerable period of time to the debate if you truly wish to become informed on this issue. Alexander’s two posts are a good place to start.
I occasionally read some really smart bloggers who view both lab leak and zoonosis as being roughly equally probable. I respect their views, as that likely means they didn’t waste enormous portions of their life taking a deep dive into this debate like I did. Good for them. I suspect that if they read Alexander’s two posts carefully, they’d switch their views toward strongly favoring zoonosis—the evidence is pretty strong in that direction.
A couple years ago, commenters raked me over the coals for refusing to admit what they thought was obvious—that Covid came from a lab leak. They claimed I was stubbornly refusing to admit the obvious out of some sort of strange loyalty to the CCP (despite the fact that I said lab leak is a danger worth worrying about, and despite the fact that I severely criticized the CCP at the beginning of the pandemic, and despite the fact that zoonosis is far worse for China’s reputation than lab leak, which explains why the CCP has vigorously tried to cover-up the zoonosis evidence.)
No doubt these commenters will show up here and apologize for their slander against my character.
Just kidding!!
PS. I enjoyed this comment, in reference to a picture of WIV scientists:
This is the Wuhan Institute of Virology’s coronavirus research group, out for a team dinner at a local restaurant on January 15th 2020 (ie a month after the pandemic started). This isn’t the most rational probabilistic evidence in the world. But we’ve already seen people take the rational probabilistic evidence twenty different directions. So let’s ask the same question Peter did – do these look like people who secretly know they just started the worst pandemic in modern history?
If they secretly knew they’d just started the worst pandemic in modern history, wouldn’t they at least be wearing masks?
Like Alexander, I don’t see the picture as being definitive evidence, rather it’s a reminder that these are real people. When I read lots of recent western commentary on China, I get the impression that the Chinese people are viewed as some sort of totalitarian cyborgs. But when I visit China and talk to people, they seem kind of like Americans. I’m slightly more persuaded by a western scientist that met the WIV people at a conference at the end of 2019, and reported that their mood seemed completely normal. Scientists aren’t like lawyers—they’re pretty transparent.
I also believe that people underweight the fact that the Chinese scientists clearly thought it was zoonosis in the early stages of the pandemic, before the CCP told them to keep quiet and began pushing alternative theories.
People who’ve never been to China often have this weird idea that the CCP is omniscient, and that nothing happens in China without the CCP knowing about it. China is different in some respects, but nowhere near as different as many of you assume. Like the US, it’s a vast and endlessly complicated place.
Tags:
10. April 2024 at 14:25
Nothing under the sun will convince a committed lab leaker.
10. April 2024 at 16:41
Hopefully Scott Alexander’s analysis and follow-up will be a reference point moving forward. I used to link to Kristian Andersen’s Unrolled Twitter Thread as a well linked and user friendly “source of truth” but I’m estimated that the number of minds I’ve changed over the last few years was closer to zero than five 🙂
10. April 2024 at 17:06
Anderson is a proven lier. No reason to quote him.
Alexander is intellectually honest. If he says it’s 90-10 I will trust he’s honest. But end of the day there is no smoking gun. No matter how much research I do on the issue I have no faith I can come to a high probability conclusion.
Not sure why Sumner keeps posting on the issue since the best argument he has is an appeal to authority fallacy of quoting the smartest guy he knows opinion. Which I agree Alexander is smart.
10. April 2024 at 17:07
As I stated once before, the Zoonosis hypothesis seems to be the null hypothesis to me, given that most such pandemics begin this way.
I haven’t bothered to pay much attention to this issue, because China can be said to be blamed either way, and China will not allow a proper investigation, so determining the truth may be impossible.
Also, I don’t think the truth in this case matters much, as China has strong incentives to try to prevent this sort of thing from happening again, whatever the origin.
10. April 2024 at 17:37
People who’ve never been to China often have this weird idea that the CCP is omniscient, and that nothing happens in China without the CCP knowing about it. –SS
I wonder about this.The modern state as a panopticon.
People in China will not express themselves freely, on the internet for example. Of course not in the mass media either.
Every large publicly held company in China has board seats allocated to the CCP, as well as official internal CCP cadres.
Of course, a large fraction of the economy is still in SOEs.
The CCP sure seems to keep a gimlet eye on everything that happens in Hong Kong, a city I am in with frequent contact during business days.
10. April 2024 at 20:06
what does lab leak mean here?
1. an infected animal kept in the lab escaped and so lab leak?
2. an infected person in the lab spread it to others and so lab leak?
3. a novel virus that was engineered escaped via #1 or #2?
#1 and #2 also are problematic even if the disease itself zoonotic. need more tight and foolproof mechanisms to prevent future occurrences.
#3 is a whole different beast of lab leak that calls the science of GOF and dual edge sword practices into question
10. April 2024 at 23:46
Scott,
…If they secretly knew they’d just started the worst pandemic in modern history, wouldn’t they at least be wearing masks…!”
The lab leak theory doesn’t necessarily say they did it intentionally.
If it leaked due to accident they might not be aware of it by January 15th as it takes time to detect an error or careless handling. Once detected, i.e. when people started getting ill, they might even be embarassed and start making up the zoonosis theory to cover your asses (happens all the time, scientists are humans)
Thus it might still come from a lab, despite them taking group pictures in January 2020 and promoting the zoonosis theory later on..
Not saying this is what happened, but I hope Alexander has better arguments…
10. April 2024 at 23:50
Well, first of all it’s not slander. Slander is related to oral statements.
In this case, you’re claiming libel.
But to prove a statement is libelous, one has to prove actual malice.
The malice qualification is precisely why Trump doesn’t sue into oblivion. You’re just ignorant, and ill-informed, and say dumb things about his character that are false. Like your recent statement that he wants to increase the debt. Or your statement about his collusion with Russia, which never happened. Or your statement about his desire to allow Xi to murder all those Uighers, which never happened. Or your statement about him being equivalent to the mass murder Adolf Hitler and Napolean.
The only reason nobody sues you for those wrong statements is because you’re just wrong. It has nothing to do with Malice, although I’m sure your hatred for Trump is very real. It oozes from your pores.
I also remember the criticism during the pandemic, because I was one of those criticizing you, and it was predominately directed at your comments around ivermectin and at republicans who supported the Great Barrington Declaration, particular the statement that they were killing people. You actually accused another party of killing people. That’s bizarre.
The CDC lost their lawsuit against ivermectin and had to remove all defamatory information, so we were correct on that front. And I think in hindsight the scientists who signed the great barrington declaration were also correct.
11. April 2024 at 05:00
@Sean – none of Scott’s analysis relies on appeals to authority.
11. April 2024 at 05:07
@Ricardo
1. The CDC didn’t lose its court case, they settled. And the case had absolutely nothing to do with defamation or ivermectin’s effectiveness.
2. Slander (and libel) both have colloquial meanings that have nothing to do with legal burdens of proof.
11. April 2024 at 07:27
I thought Scott Alexander put it well when he said that either way you look at it, this is a remarkable coincidence.
– Either the virus crossed over from animals and out of all the wet markets in China, the first super spreader event was a 20 minute city drive from a lab doing Covid research, or
– The virus came from the lab and out of all the places in Wuhan, the first super spreader event happened to be in a wet market.
Whichever theory you like, there isn’t the supporting information you’d like to see, but it’s easy to blame that on the CCP cover-up.
The debate was great if you want to know more about esoteric (but important) details like the Furin cleave site, and I would definitely bet on natural origin* if I had to bet, but I’m also not anywhere near 90-10 confident.
* For clarity, I would consider “Lab researchers brought natural strain to lab, where it escaped, possibly after light GOF work” to be “lab origin,” but even with that definition, I still favor natural origin.
11. April 2024 at 07:28
Sean, You said:
“best argument he has”
No, the argument is contained within Alexander’s two posts. Read them.
Viennacapitalist, That seems unlikely. They knew what viruses they were working on and it had been sequenced by then. Lab leak proponents all assume (correctly) that they would have known.
Again, I don’t want to make too much of this; there are far more important data points.
11. April 2024 at 07:31
J Mann, The same sort of Bayesian updating that that Saar uses to get to lab leak pushed me 90% to zoonosis.
11. April 2024 at 07:40
I was irritated at the people that were irrationally in favor of the zoonosis theory and calling anyone who considered the lab leak hypothesis plausible conspiracy theorists. Propagandist morons, at that early stage there was no need to shut people down and it was irrelevant where the virus came from, why were they so concerned about shutting the theory down? I guess we did find out the answer, some were afraid it would shut down their research.
11. April 2024 at 07:49
I guess you cam claim Alex Berenson is obscure, but he isn’t remotely, and the Biden Administration considered him important enough to actively suppress. Free speech advocate Scott Sumner doesn’t see this assault on the First Amendment as a matter of concern. Scott Alexander’s failure to even bring up Berenson’s recent posts on the lab leak theory, or acknowledge Berenson’s existence at all, suggests to me he is not an honest guy, even if he is a “msart guy.”
11. April 2024 at 07:50
XVO, Both side were irritating, both sides were biased, and both sides used improper debating techniques.
11. April 2024 at 07:53
Stan, Berenson’s claims have been discredited.
“Free speech advocate Scott Sumner doesn’t see this assault on the First Amendment as a matter of concern”
LOL, how many times have I railed against left wing cancel culture?
11. April 2024 at 09:49
Are we really sure there aren’t laboratories in the First World (or worse, the rest of the world) with potential to cause another COVID-19? Even if the lab leak theory is false, this deserves investigation.
11. April 2024 at 09:49
Scott,
The debate was great, and I admit that if God were prepared to settle the bet, I wouldn’t bet against you or Scott Alexander on this question.
It also convinced me that I don’t have the math skills to call myself a Bayesian, together with this comic.
https://www.smbc-comics.com/comic/bayesian
11. April 2024 at 11:14
Scott,
Read both of Scott A.s articles but didn’t finish either. I’ve never found it to be that interesting of a question, but I’ve always leaned zoonosis and those articles just made me lean harder.
I still think the strongest argument is that all the people pushing the lab leak theory tend to be the worst people.
I’m still not 100% which is fine, because I am 100% convinced that it isn’t very important.
Good post.
11. April 2024 at 12:44
Having first hand interest in covid have followed this moderately closely though with less interest as time goes by. The debate was very well done and packed in lots of info in a short amount of time. Alexander is correct that there will always be someone with some obscure evidence popping up, but to be fair there is no way to be 100% certain. 90/10, 85/15 seems to be pretty much in line with available evidence. Also, you need to keep the timeline correct. The initial lab leak claims in the US also claimed that the virus was deliberately created as a weapon to hurt the rest of th world, or US.
You can be pretty sure that the lab people would have known about covid well before the picture and would have gone over their own records to see if they had any part in it. They would not have attended a public gathering like that and risked exposure if they thought they had anything to do with it.
Steve
11. April 2024 at 14:58
SS seconds this opinion: “For now, I’m still at 90-10 zoonosis” – funny, but you never hear Scott talk about the 10% chance. It’s always the 90%. As for Alexander’s analysis, I will only read it if it goes into Bayesian statistics and discusses among other things the 1977 CHN-RU lab leak, the leaked US State Dept memo on WIV’s lax standards, the redacted memo from the NIH’s Dr. Fauci et al, the fact that WIV came up with the DNA sequencing so quickly, the weird coincidence of the spike protein matching previous chimeric virus research (Shi Zhengli et. al’s Duke research) and the like. I doubt Alexander gets into these probabilities since they are not easily available. But I will say this: in a court of law WIV would have been found guilty as a matter of law due to the “cover up”. Scott and Sinophiles say it’s because that’s how the Chinese roll, but that rationale would not save them in most Western courts of law.
11. April 2024 at 15:50
Harding, I’ve made the same point–many times.
Ray, I see you haven’t kept up with this issue.
11. April 2024 at 20:38
Alexander convinced me that zoonosis is likelier than a lab leak. But I still think a lab leak is a more damaging story to Chinese authorities. Zoonosis happens; didn’t swine flu in 2009 originate in the US? And what about MERs? And then there is Zika and West Nile Virus, etc., etc. China not cracking down on wet markets is seems a lot less sinister, because a lot of countries keep livestock in unhygienic conditions. Nothing about that would be exceptional. However, performing gain of function research on the types of pathogens that are most likely to spark a pandemic in a lab with ties to the military and a lab that has a history of failing to adequately follow safety protocols is something that most countries don’t do.
11. April 2024 at 23:33
If C19 resulted from a wet market..why are there still more than 40,000 wet markets in China?
Also…
Granted, Wuhan is likely one of the largest wet markets in China, but there are much larger cities in China than Wuhan, and larger wet markets.
Of all the wet markets in China, including perhaps a dozen that are even larger than Wuhan’s…the one wet market in China that the virus escaped from…was next to two Wuhan labs in which gain-of-function coronavirus testing was being done on bats?
The very type of virus in question?
Seems like a looooonnnnngggg shot.
12. April 2024 at 06:00
Thank you, Sumner.
12. April 2024 at 06:45
Lizard, Wrong. Chinese wet markets are very exceptional—especially likely to produce pandemics.
Solon. Wow, you packed a lot of false claims and bad logic into one short comment!
12. April 2024 at 09:08
@steve said:
In the early days of the vaccination campaign we regularly discussed rare events that the 30-40k RCT trials were not sufficiently powered enough to detect. I have long argued that the Lab-Leak Hypothesis was plausible but unlikely, however, that does not imply that the probability has to be a whole integer greater than or equal to one percentage point. I started off saying “near zero probability” since 1/200 seemed charitably closer to reality than 1/100. For me, the Bayesian/Fermi estimate upstream of the “spillover” location was the most improbable bottleneck, not only the existence of a novel coronavirus isolate (i.e. not just a swab sample but cultured virus) with pandemic potential but that this isolate was kept secret for future nefarious reasons that are incompatible with the publishing history of Shi Zhengli’s lab. There was a possibility that Shi Zhengli’s lab was holding back data until they published but as true evidence started to build, this probability became orders of magnitude less likely. Instead of 1/200, the evidence, IMO, hit 1/1000 range then 1/10,000. New evidence could have drastically shifted the probabilities into the whole integer percentage range but the new evidence always seemed to go the opposite direction, against the involvement of Shi Zhengli’s lab. The Lab-Leak Hypothesis could still not be ruled out since other virology labs were located closer to the Huanan Market though it was much less likely that these less prominent labs were doing work on sarbecovirus isolates from the Lancang/Mekong River basin in Yunnan province and/or northern Laos. Sloppy handling of an obscure Hubei province sourced swab sample at a less prominent, less competent, and more generalist lab was still in play. I could not imagine how such a scenario could fit the existing evidence but my imagination is imperfect at best.
Saar Wilf’s rough Bayesian estimates exposed the flawed logic of the Lab-Leak promoters. He quickly went from a near-zero probability based on priors to 90%++ probability based on little more than Jon Stewart’s “Hershey Factory” joke and an “evil mad scientist” intuition that equates even the hint of a lab experiment involving genetic engineering to prohibited Gain-of-Function research. We seem to forget that Emmanuelle Charpentier and Jennifer A. Doudna 2020 Nobel Prize in Chemistry for their CRISPR/Cas9 genetic scissor work was well deserved because this technology is largely positive sum rather than an existential risk to humanity. Any virology lab that does not leverage genetic tools is not engaged in modern research, period. Similarly, Ray Lopez’s examples of previous lab leaks is not relevant since these involved well-known viruses. Secret and previously unknown to science virus isolate is a very different proposition.
90/10, therefore, seems more like an exercise in being respectful to opposing views than a careful analysis of the evidence. The Lab-Leak Hypothesis has not yet been falsified but it is very very very close. The social psychology behind the general perception that Lab-Leak is still likely is fascinating, and somewhat horrifying, but that is not evidence supporting the Lab-Leak Hypothesis, that is evidence that falsifies the Wisdom of Crowds hypothesis. YMMV.
12. April 2024 at 09:20
RAD, Excellent comment.
13. April 2024 at 15:34
@RAD – surely you’re joking? You and Sumner think that WIV publishing history and a group photo of researchers after the C-19 virus was in the wild is evidence of a non-leak? After the CCP deliberately silenced Zhi by replacing her with a CCP general at WIV? After they could care less if researchers at WIV died from exposure? Seriously? As for Bayesian analysis, nobody, not you, not me, and Saar Wilf and certainly not Scott Alexander, whose judgement I don’t trust (was he the one that was outed with his real name? Shows an alarming lack of judgement talking about his work on the internet), knows the true Bayesian numbers. They are unknowable, so you’re “85%/15%” against a lab leak and my “99% 1%” hunch for a lab leak are just guesses based on your priors. But, being familiar with western legal system, unlike you, as I said before, I’m 99% confident WIV would be held liable for the C-19 pandemic. Simply because they covered it up. As a matter of law.
@SS: I see you’ve still into personal attacks rather than reasoned debate, no surprise.
13. April 2024 at 15:47
@myself- also keep in mind re the “group photo” the WIV researchers, who were working on “spike protein” chimeric viruses very similar to SARS-CoV-2 (the C-19 virus) simply may not have know their lab accidentally or otherwise released the virus. Not everybody knows everything going on at their place of work.
But it’s telling that the first people to sequence the SARS-CoV-2 virus, in record time that stunned Western researchers who thought it would take many more months than it did, was the WIV. Why was that? Because they created it.
13. April 2024 at 20:39
The most important section from Scott Alexander’s follow-up post:
Peter’s Jan 16, 2023 blog post, The case against the lab leak theory — Every reason why lab leak theories fail and a natural origin of covid is more likely, is excellent.
I’ve wasted too many brain cycles debating with misinformation merchants like Ray Lopez who, IMO, is not a good-faith actor. For Dr. Steve and other smart and truly good-faith actors, please find an hour to work through Peter’s post.
13. April 2024 at 21:36
@RAD – your “faith based” science is predictable. Like those Baptists who won’t believe in evolution. But I have worked at the intersection of public policy, law and science (patents mainly) before I retired, including biotech. Unlike you, I don’t believe in either wild coincidences (recall Peter Fleming’s advice about coincidence, happenstance and enemy action) and I know why scientists lie (grant money and to prevent public backlash against CRISPR research). Unlike you, I know the story behind the story. No doubt you are unaware of a lot of things, not just the 1977 H1N1 leak but a plethora of things, since you’re not in the field. So I’ll leave you with an easy one, simple for people like you and Sumner: why do porous masks work to prevent SARS-CoV-2? When in reality for really good protection against dangerous viruses found in nature you need a moon suit? Because chimeric viruses made in the lab are fragile, in the same way a domestic cow is fragile compared to an African Cape buffalo. The simple heat and humidity gradient from a paper mask is enough to kill a lab made virus. And there’s a long list of things I could get into (at one point I had an email listing all the improbable things you have to believe to believe that C-19 was from nature, I even copied our host), but it would be a waste of time to believers like you.
13. April 2024 at 23:02
@Ray said:
Ah, from Peter Miller’s article:
p.s. I’m not joking, and don’t call me Shirley 😉
14. April 2024 at 09:01
“I don’t believe in either wild coincidences”
Like the fact that almost all the first cases just happened to occur near the same sort of animal market that triggered SARS-1? That sort of coincidence?
17. April 2024 at 10:56
SS: “Like the fact that almost all the first cases just happened to occur near the same sort of animal market that triggered SARS-1? That sort of coincidence?” – that was no coincidence, that was enemy action. The WIV had an adjunct office next to the wet market, which was torn down when the pandemic started. The theory is either infected WIV personnel visited the wet market or, since Asians don’t like to waste anything, they were selling infected rabbits from the WIV to the wet market to make a little extra money. The last suggestions sounds like nuts to most Westerners until you meet Asians (from those foraging in public parks for food to JIT JP inventory management, where every scrap is accounted for during manufacturing and there’s almost no inventory, to avoid inventory related issues). If you bothered to read my emails that last suggestion was in one of my early emails on C-19 origins.
9. July 2024 at 11:16
]
Sorry, that’s been bugging me.