Something I've noticed about Schlomo here is that he writes and talks like he learned everything off of Wikipedia, which I think he admits to. Either way, I read him calling himself as an autodidact that way. My work brings me in contact with statisticians and data scientists all the time, and none of them talk about Bayes Theorem the way these guys do, let alone suck it off the way LW does.
This type of thing seems pretty common with "rationalists". If they were good at explaining and communicating, they'd be teachers, authors, or professors. Like, mathematical or philosophical logic are formal systems of reasoning meant as a way to discuss certain issues/problems, not a way to live your entire life. As it stands, the rationalists are mostly just "ideas guys" who happen to have created definitions of intelligence and value that put their useless asses at the top of the pecking order.
Another example of this type of guy, though he's not a lolcow, is
Jaron Lanier. Just another "self educated" guy who hasn't done any real work in computing or the internet (and hell, I'll count management there, because managers still have to talk to the code monkeys) but makes all these sweeping statements on how he knows how the future computers and the internet will play out, that are meant to amaze people without any technical background themselves.