That is to say that I don't think SSC is a eugenicist or racist on purpose, necessarily. I think he's a pretty-smart guy who thinks his intellect and knowledge in one area (psychiatry and medicine) gives him the ability to reason about other areas correctly without restriction, and without context that is necessary. For example:
"HBD" as a term is a stand-in for racial science or eugenics. Those are explicitly racist fields of study, and are distinct from things like the study of human genetic variation.[1]
A common misconception is that correctness is a defense to allegations of racism the same way truthfulness is a defense to defamation. It's not the presence of data, but rather the leaps in interpretation of that data which underpin eugenics.
The problem with saying "HBD is mostly correct" is that that statement doesn't mean supporting "data shows Black people have lower IQs as a group in America". (That statement is extremely fraught itself, but it's at least got some data around it[2].) It means supporting "Black people are genetically predisposed to have lower IQs", adding "people with lower IQs are of less societal value than people with high IQs", and ending with "Black people should not be allowed to reproduce so that the population will have a higher IQ overall."
When Siskind/Alexander is saying "HBD" is probably right, he may not realize he's doing this, but that whole chain of logic is the kind of thing he's throwing his hat in, not just the first statement. If he wanted to discuss just that first part, he needs to consider environmental context he simply is not considering, and which eugenics as a "science" purposefully avoids addressing.
He doesn't seem to know what he doesn't know, which to me is more dangerous than knowing what you don't know, and not caring.
[1] https://en.wikipedia.org/wiki/Human_genetic_variation
[2] https://www.brookings.edu/articles/the-black-white-test-scor...
Edit: This tweet sums up my concern here: https://twitter.com/ElSandifer/status/1362135175912693766
The arrogance of being sure you are smart enough to never make a mistake in the ideas you pick up is pernicious and deadly.
Ed: I would also add that while thinking you have an infallible BS detector is dangerous, believing that truth can only come from certain people and your enemies never have unique insight is another form of hubris, and I think the latter has at least as good as argument for being more dangerous in the long term.
I actually really like that Slate Star Codex had a page where Scott listed his mistakes, and that he makes public predictions at the start of each year and rates whether they were correct. These are great ways to keep track of one's thinking, and more people should do it: https://web.archive.org/web/20191224063106/https://slatestar...