Yeah of course, but... all results are equally likely, but some are "interesting" or "meaningful", like all tails or all heads. If it's 21 tails, it's pretty mu

Yeah of course, but... all results are equally likely, but some are "interesting" or "meaningful", like all tails or all heads. If it's 21 tails, it's pretty much certain that something is not right, that it's not a fair coin. In a Bayesian point of view, let's pick some prior probability distribution over lengths of human survival and consider the current age of humanity as a random variable. This gives the posterior probability over lengths of human survival.