A perfect example of the problem, I guess.

Many pro-LW-mainstream arguments are weak if you have significantly different priors. People with minority view quickly learn the difference in priors and learn to express their views less often and defend them less.

I also consider FOOM-as-described-on-LW quite improbable and the writings of Eliezer on the topic simply raise a few red flags; I see that it is a popular position here, but most people don't find it worth the effort to fight mainstream.

There are still many topics on LW where no relevant values or priors are parts of LW majority's collective identity and I get some entertainment and information form reading these discussions and participating in them. There are also topics close to things that are accessible to science with all its rigidity (but also stability) compared to Bayesian inference. These are very informative too.