AMA | October 2024

9979.855 - 10004.982 Sean Carroll

So in cases like this, I don't think that my answers are very firm. But I think that as a methodological matter, appealing to unlikely branches of the wave function should literally be your last resort. You know, the Bayesian prior on that one is very, very small in my mind. There's almost always going to be a larger prior for there's some reason for this, we just haven't figured it out yet.