Far Truth Is For Extremes
To answer the question posed in my last post, here are some situations where it makes sense to forgo the large benefits of things like religion, to care about far truth:
You are stuck in your ways, like a smoking addict. You admit it would have been better for you had you become more religious early on, but alas you fell in with the wrong crowd, and now the costs of change for you outweigh religion’s gains. If you are nice, you’ll warn young folks to avoid your downfall.
Contrarian far claims with big personal consequences are true. If choosing cryonics would gain you five or more expected years of life (over its costs), and you are one of the rare people who would actually do something so contrarian after being intellectually convinced of its advantages, and if you can reliably discern when a majority is wrong, then you’ll need to think accurately about far topics to find such opportunities. For non-contrarian far claims with personal consequences, you could just follow the crowd without thinking.
You have a good chance of being respected as a far topic expert, by a community that evaluates claims in truth-correlated ways. If you could be a famous cosmologist, you might try to create cosmology claims that will look good when evaluated by the tests cosmologists will apply. The gains from becoming a famous cosmologist could outweigh the risk that by becoming more truth oriented you will forgo religion’s gains. Beware, however, that truth-correlated is not the same as true – most communities say their far claim tests are more truth-correlated than they actually are.
So assuming you actually have a viable choice, the situations where it makes sense to reject religion in favor of far truth are extreme – either there are big personally-useful far contrarian claims to learn, or you have a good shot at being a rare far expert, respected by a community with truth-correlated standards. So if such extremes seem unlikely to you, far truth probably isn’t worth its costs to you. Go away, and sin no more.