The Trailing End of the Ivermectin War

Opining on the Nature of Experts

A long time ago in a galaxy far far away, Scott Alexander of Astral Codex Ten made a very convincing post about the nature of ivermectin efficacy against Covid-19, in which he stated generally that ivermectin seemed to work a little bit against Covid-19, and that was probably due to its anti-worm properties. This post hit at the exact right time of the Covid-19 culture war, got a tremendous amount of traffic, was referenced by dozens of large media outlets, and provided an out for the anti-ivermectin culture warriors to admit that it worked some but stand by their position that we still shouldn’t use it. If we stick to our HWFO Media Dynamics analysis toolkit, this article got traffic mostly because it was useful to an echo chamber, not because it was necessarily true or untrue, since that’s how viral traffic works. I initially found it convincing, and found it less so after talking with Alexandros Marinos.

Alexandros Marinos, a personal friend and excellent analyst who is very pro-ivermectin, disagreed heavily with the analysis. Alexandros and Scott got into a not-quite-fight. Scott is the elephant in the room and Alexandros is a mouse who deeply admires the elephant, and wanted the elephant to change his mind, and do so publicly. Alexandros started an entire Substack to lay out his case, which is here, and worth a read:

His work began with this, which is relatively exhaustive:

A Conflict of Blurred Visions

All his work, including some extensive behind the scenes dialogue with Scott, columnated in a recent post about potential mathematical errors in the Worm Article, on Scott’s behavior after being made aware of these errors, and about overall faith in the rationalist community itself.

Scott Alexander's correction on ivermectin, and its meaning for the rationalist project.

Scott responded yesterday, by posting what amounts to a very soft correction, and defending himself from certain characterizations in Alexandros’s article on The Motte, which is a subreddit discussion space affiliated enough with Scott that everyone there knows who he is, but loosely enough that the discussion won’t get nuked by the moderators. In his Motte post, he objects to Alexandros’s characterization of him in the piece, and then rationalists all argue on Reddit because that’s what rationalists do.

If you’d like to dive deep into the back-and-forth, feel free. If you’d like to do a year’s worth of education on statistical methodologies to keep up with the mathematical claims, feel free. If you’d like to decide for yourself whether ivermectin works or not, feel free. That’s not what we’re going to do here. We’re going to talk about experts.

Experts Challenged

I have not seen anything on either side I consider to be "bad faith." I have, however, seen behavior which I consider to be fear driven, and typical of that of an expert challenged. Discussing how fear affects experts stranded in the center of group populations seems more interesting to me than talking about ivermectin efficacy.

I no longer think it matters whether ivermectin works. I think the dynamics around the ivermectin discussion were classic Game B sensemaking crisis culture war stuff. Most people are irrational emotional mob followers, and most people chose their opinion whether ivermectin worked based on in-group signaling. The few who were trying to do it rationally did so with different givens fed to them by their social feeds through an echo chamber filter. While this was probably not the case for either Scott or Alexandros, this was obvious for 99% of the people in the world with an opinion on it. So the most interesting thing to talk about with this deal is the psychology of "expert" behavior while standing in the middle of a mob.

I'm an expert, but not in medicine. I'm an engineering expert in stormwater hydrology, and when you're a civil engineer PE you have a very deep ethical code to which you must adhere to keep your license. Similar to the Hippocratic oath, except related to public safety with regard to your engineering designs and analyses. And civil engineers break this oath all the time, because of the psychology of being an expert, and that sucks, and here's an example.

If you adjust a floodplain map, and you make major changes to that map that adjust the floodplain boundary, which affect tens of thousands of acres of land, and you are paid for this work, and you have a professional reputation staked against this change, and some other engineer peer reviews your work and discovers an error, you are ethically and morally bound to review that other engineer's qualms and if the error is correctly identified you are ethically and morally bound to admit your error and retract your work because failure to do so could lead to dozens of dead people and hundreds of millions of dollars of property damage. I have watched engineers refuse to retract this very thing. Also, if you live in South Georgia don't buy land on the Ogeechee River, because I know with 100% certainty that the floodplain maps are wrong, I stated so in state court, the Georgia EPD and a very large engineering firm doubled down on bad work, and people are now building houses in the floodplain because of it.

As that court case resolved, which is the only court case I've ever lost, I as a professional had to choose whether to run to the media and wave my hands and freak out and try to personally attack one of the five biggest engineering companies in the country, staking my professional reputation on a fight against a titan, or throw my hands in the air and pop popcorn waiting for the next hurricane to hit Savannah GA. I chose to pop popcorn. There was literally nothing else I could reasonably do. Even as a mouse who very much disliked the elephant, I popped popcorn.

The thing that frustrated me most about the experience was the unwillingness of the other engineer to admit the engineering failure, which was profound and egregious. They basically modeled the Ogeechee River floodplain, which is choked with cypress trees, as if it were desert sage brush. Very bad error. Nobody corrected it, because of (A) south Georgia “Boss Hog” politics, and (B) the psychological nature of experts.

this stuff ^^ impedes the flow of water way more than desert sagebrush does

When you're an expert, being an expert becomes part of your personal identity. If someone challenges your work, you take that as a personal affront, as if they're challenging you. And if you're an expert that's towards the front of a giant mob, and the giant mob is listening to you, then that other guy's not just challenging you, he's challenging your position within the mob. All those hardwired primate signal paths light up in your brain and influence your behavior to avoid losing your position within that mob.

Rationalists like to pretend that this primate wiring doesn't exist, or at least doesn't exist within them. Part of the rationalist identity is that they're half Vulcan and not subject to all this awful primate thinking in which they're mired. But when Alexandros came to me on the side in private channels with a lot of this stuff, I basically predicted Scott's response, and the response of the overall group, and cautioned him against trying to push it too hard. I advised him to state his case publicly and then pop popcorn. Walk away. I hope he does do that once this last argument resolves, even though my take is that Scott is either outright wrong or hedging his bets by saying "well I wasn't totally wrong because I admitted it might work some."

I too think ivermectin might work some. It probably reduces bad outcomes by around 15% or 20%. The combination of it with all the rest of the FLCCC early treatment protocol may reduce bad outcomes by as much as 40% or 50%. I think the mode of action of the FLCCC protocol is partly disease specific but partly just related to overall immune system improvement during an infection of any kind. I think the CDC and official channels made a choice to spike ivermectin as well as spiking all other early interventions because they were afraid that public knowledge of a treatment that was somewhat effective would reduce vaccination rates. They have a history of doing the same shit all across the public health space, with things like how they attack nicotine vapes for instance.

Scott now upgrades his estimate of ivermectin efficacy from XYZ% to (XYZ+10)% or something, thinks he's done his job, and attempts to walk away without losing credibility. While this may not be the path a pure Vulcan brained rationalist would take, it's the path most "experts" would take, because of expert psychology and mob dynamics. Alexandros becomes irate because he thinks that people died from the ivermectin “horse dewormer” backlash, which they probably did, and Scott was a contributor to the backlash. But the CDC and global health calculus went something like "X people will die from lack of vaccination due to thinking there's a treatment, Y people will die from not giving that treatment, X>Y so let’s spike ivermectin." I talked about that a year ago on HWFO. In their minds, lying is rational and good, like lying about nicotine vapes.

And in the end, the actual rationalist argument should be about whether X>Y or Y>X, and nobody's got the data to make that case in either direction. Did Scott kill people by aiding in spiking ivermectin? Or did he save people by convincing them to get vaccinated? If the latter, how many more people did he save by convincing them to get vaccinated than those who died from adverse vax events? Was the delta enough to overcome the ones that didn't get the FLCCC protocol because the CDC went hard in the paint against it? Someone do the math on that for me, because it's above my pay grade. What I do know is this.

Mobs gonna mob.

Experts gonna expert.

Popcorn gonna pop.

And the biggest lesson from the whole thing is that classic catch phrase from the 40 Year Old Virgin movie. Don't put the pussy (metaphorical semi expert whom you deeply admire) on a pedestal.