Maybe you could elaborate on this so that I understand it better. How could all those people who happen to contribute money to the mitigation of AI risk not "care all that much whether other lesswrong participants care about AI risk"? For that stance to make sense, the reason for their donations couldn't possible be that they wish to support SI. Because then they would care a lot if other people took the cause seriously as well, since each person who takes AI risk seriously does increase the chance of SI receiving additional support.

Donating a trivial amount to one charity is a big leap from ostracising all those that don't.

You might further argue that it isn't irrational not to worry about AI risk.

It is irrational. It's just not something there is any point being personally offended at or exclude from the local environment. On the other hand people not believing that correct reasoning about the likelyhood of events is that which most effectively approximates Bayesian updating have far more cause to be excluded from the site - because this is a site where that is a core premise.

Could you please tell me where I can find arguments that support that stance?

I'm almost certain you are more likely to have collected such links than I. Because I care rather a lot less about controlling people's beliefs on the subject.