2

I found the following argument here (although the paper is about a different topic):

A General Argument Against Immortality:

The method of Theory Confirmation can be applied to the question of immortality. In general, if we are immortal, there would be two classes of observations: Those made by normal people within a normal lifespan, and those made in the ‘afterlife’. For ‘quantum immortality’ the ‘afterlife’ will be taken to mean those who find themselves to be much older than a normal human lifespan. If the ‘afterlife’ is infinite, then it will have infinitely more integral measure than the normal life. Thus, the effective probability of finding oneself in a normal lifetime would be zero. If there is no ‘afterlife’ then the effective probability of that would be unity. By applying Bayesian reasoning, this implies that if one does find oneself in a normal lifetime, as we do, there must be no infinite afterlife.

Is this argument valid? To me, it seems that it is obviously false, but I may be missing the nuances of the argument. I would object that there is no random sampling here, which the author seems to assume. We are not randomly sampling the time in which we exist. If this argument is valid, we can make a general argument against the possibility of indefinite time. If time is indefinite, there is a zero probability that time is where it is currently, yet it is.

So, is this argument valid? How does it avoid my objections, especially the seeming conclusion that time cannot be indefinite?

  • It seems to you that the conclusion is false or that the argument is invalid? If it is the latter, why? The "nuances" are listed on the same page, one has to accept MWI (which is still a minority view), even probability zero does not make normal lifespans impossible, and finite "afterlife" does not even produce zero probability. – Conifold Dec 2 '19 at 23:07
  • @Conifold I think that the argument is invalid, which makes the conclusion invalid (that immortality is extremely unlikely because of the fact that it is infinite in time), because of the reasons I lay out in my question. My question is not about anything related to MWI, I'm just wondering if the general argument that immortality is extremely unlikely stands. – user43277 Dec 2 '19 at 23:11
  • The author explicitly says, also on the same page, "This proof still leaves the possibility that the MWI is false and that we would have been immortal if only it had been true. However, advocates of “quantum immortality” believe the MWI and QI are true". The probability setup explicitly relies on MWI, hence random sampling isn't needed, and your analogy to the time point does not work. Its probability is 0 regardless of indefinite time, and for intervals you have to assume equidistribution over all points to get 0, which the afterlife argument does not. – Conifold Dec 2 '19 at 23:21
  • If the argument was valid, you could imho prove, that death or any other future event is impossible as well. Technically it's claiming "the propability of making an observation before any future point T in time is very small compared to making that observation after T". Since we are all currently making that observation before T, it's unlikely that T exists." Which is obviously wrong. Perhaps because it would assume random sampling in time, as you already mentioned...? – Omni Dec 2 '19 at 23:37
  • @Conifold Maybe I don't understand his argument then. How does this passage: "If the ‘afterlife’ is infinite, then it will have infinitely more integral measure than the normal life. Thus, the effective probability of finding oneself in a normal lifetime would be zero. If there is no ‘afterlife’ then the effective probability of that would be unity. By applying Bayesian reasoning, this implies that if one does find oneself in a normal lifetime, as we do, there must be no infinite afterlife." rely on MWI? Isn't he constructing an argument based purely on probability and time spans? – user43277 Dec 2 '19 at 23:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.