Good ideas and conversation. No ads, no tracking. Login or Take a Tour!
I keep seeing SBF quotes regarding Bayesian statistics, and I gather that's a broad theme within the effectobro community? Bayes' theorem is simple and elegant, but suffers from a critical flaw for many applications, which is that it relies on a priori knowledge of probabilities of certain events happening. Inevitably, there is no knowledge of the true probability of many future events happening or not, so the "known" probability is replaced by an assumed probability. In the case of whether AGI will come to pass in a few years, I gather that the effectobros just go ahead and insert 100%. Same for whether the "biologic substrate" matters.