I'm legitimately worried about the money and the incentives it creates. What would a self-interested agent (LW seems to use "agent" in exactly the opposite sense to what I'd expect it to mean, but I hope I'm clear) in the position of the LW leadership do? My cynical view is: write some papers about how the problems they need to solve are really hard; write enough papers each year to appear to be making progress, and live lives of luxury. So what's stopping them? People in charities that provide far more fuzzies than LW have become disenchanted. People far dumber than Yudkowsky have found rationalizations to live well for themselves on the dime of the charity they run. Corrupt priests of every generation have professed as much faith that performing their actual mission would result in very high future utility, while in fact neglecting those duties for earthly pleasures.

Even if none of the leadership are blowing funds on crack and hookers, if they're all just living ascetically and writing papers, that's actually the same failure mode if they're not being effective at preventing UFAI. When founding the first police force, one of Peel's key principles was that the only way they could be evaluated was the prevalence of crime - not how much work the police were seen to be doing, not how good the public felt about their efforts. It's very hard to find a similar standard with which to hold LW to account.

It occurs to me as I write that I have no idea what the LW funding structure is - whether the site is funded by the CFAR, MIRI, SIAI or something else. Even having all these distinct bodies with mostly the same members smells fishy, seems more likely to be politics than practicalities.

The kookiness... if LW were really more rational than others, I'd expect them to do some weird-but-not-harmful-to-others things. So I suspect this is more a perception than reality thing (Though if there are good answers to "what's the empirical distinction between real and fake cryonics" and "why do you expect polyamory to turn out better for you lot than it did for the '60s hippie communes" it'd be nice to see them). IMO the prime counter would be visible effectiveness. A rich person with some weird habits is an eccentric genius; a poor person with weird habits is just a crank.

It would be really nice to have more verifiable results that say LW-style rationality is good for people (or to know that it isn't and respond accordingly). The failure mode here is that we do a bunch of things that feel good and pat each other on the back and actually it's all placebos. We actually see a fair few articles here claiming that reading LW is bad for you, or that rationality doesn't make people happier. On thinking it through this would be the kind of cult that's basically harmless, so I'm not too concerned. On the perception side, IMO discussing health is not worth the damage it does to the way the community is seen (the first weight-loss thread I saw caused a palpable drop in my confidence in the site). I've no idea how to practically move away from doing so though.

Secrets and bans rub me very strongly the wrong way, and seem likely to damage our efforts in nonobvious ways (to put it another way, secretive organizations tend to become ineffective at their original aims, and I'm worried about this failure mode). I certainly don't think the ban on the basilisk is effective at its purported aim, given that it's still talked about on the internet. And just having this kind of deception around immediately sets off a whole chain of other doubts - what if it's banned for other reasons? What else is banned?

If there really is a need for these bans, there should be a clear set of rules and some kind of review. That would certainly address the perception, and hopefully the actuality too.

I think the use of fictional evidence is actually dangerous. Given the apparently high value of LW-memetic fiction in recruiting, I don't know where the balance is. I think overuse of jargon is just a perceptual problem (though probably worth addressing).

I have... unusual views on diversity, so I don't think setting people against their less-rational friends is an actual problem (in the sense of being damaging to the organization's aims); I file this as a perceptual problem. The most obvious counter I can think of is more politeness about common popular misbeliefs, and less condescension when correcting each other. But I suspect these are problems inherent to internet fora (which doesn't mean they're not real; I would suggest that e.g. reddit has a (minor) cultish aspect to it, one that's offputting to participation. But there may not be any counter).

The hierarchy: in the short term it's merely annoying, but long-term I worry about committee politics. If some of the higher-ups fell out in private (and given that several of them appear to be dating each other that seems likely) and began sniping at each other in the course of their duties, and catching innocent users in the crossfire... I've seen that happen in similar organizations and be very damaging. Actual concern.

So in summary: actual concern: where the money goes, any secrets the organization keeps, clarity of the leadership hierarchy, overuse of fiction. superficial issues: overuse of jargon. The rest of my list is on reflection probably not worth worrying about.