Hmm. I’m not sure I understand what you mean by “possible” or “should” enough to know how to apply it here. If it’s not the case that entropy is literally infinite in possible quantities, is it still possible just because we don’t know it’s not the case? We know from Godel that math is either incomplete or inconsistent, so what form of “should” is contained in “can’t”? I’m not sure even what “model” means here—math doesn’t model anything, it stands on it’s own axioms. Models tie math to reality (or at least predictions of observations), imperfectly.
Hmm. I’m not sure I understand what you mean by “possible” or “should” enough to know how to apply it here. If it’s not the case that entropy is literally infinite in possible quantities, is it still possible just because we don’t know it’s not the case? We know from Godel that math is either incomplete or inconsistent, so what form of “should” is contained in “can’t”? I’m not sure even what “model” means here—math doesn’t model anything, it stands on it’s own axioms. Models tie math to reality (or at least predictions of observations), imperfectly.
I suspect “no math handles all worlds” is fully addressed by a size parameter, letting the remainder of the math stay constant.
What we set it to comes down to Bayes. Let’s not assign probability 0 to a live hypothesis.