1) All I have known, my entire existence, is existence.
2) It is impossible for me to even imagine non-existence.
3) I understand that in the world I perceive, people cease to exist to me after they die.
4) As a materialist I know my mind will cease to be supported by my physical brain after I die, and hence, my mind will cease to exist as well.
5) My understanding of 3) and 4) are predicated, however, on my own existence.
6) Thus 1) and 2) seem to be of much greater weight than 3) and 4).
7) Thus, it is rational for me to not believe my existence will end upon my death.
You forgot the part about all the evidence showing you didn't exist for the first 14.7 billion years of the universe, and for god knows how long before that.
True, it passed pretty quickly for you, but there you go. I imagine the time after I die will pass pretty quickly, too, before I am resurrected somehow.
Actually, there's several rational reasons an athiest might think they'll survive past death:
1) The techno-rapture. At some point in the future, humanity will achieve sufficient technological ability they will be able to resurrect dead humans -- perhaps even those whose bodies have rotted away. If you have a body with brain reasonably intact, it's certainly just an engineering issue.
This is tangentially related to the idea of transhumanism and the Omega point, which is one such way this might be accomplished, but not the only one.
2) In the tradition of WhatBadgersEat.com, you may be
living in a computer simulation already. This is an inversion of the old argument from the 1970's which suggested we were about to destroy ourselves.
In that argument, since you exist, you most likely would have been born in the largest bulge of humanity rather than a long time ago, which would be far more unlikely. Hence if humanity extended into the indefinite future, where the bulge would grow ever-larger, then you would probably have been born then, in the future, rather than now.
You weren't. Hence there must be no indefinite future, and Doomsday must therefore be right around the corner.
This simulation argument inverts that by suggesting that, yes, humanity will extend into the future, at least far enough to be able to create massive simulations, which will have many more virtual citizens than "real world" ones. Hence the vast majority of people who will ever live will live in those simulations.
Hence, statistically, you are probably in one of those simulations.
The flaw with both arguments, if you ask me, is that some people, of course, are born in the real world, even back in the distant past.
Every one of them could come to these two conclusions. And every one would be wrong. Hence the theories give us no new information, and is thus useless in a partitioning/decision theory sense.
Having said that, decision theory would, however, validly conclude that the proper thing to do is presume everyone was in a simulation, because (in theory) 99% of the time you'd be correct. And unless you can come up with a better judgement algorithm, that isn't too shabby.
The only thing remaining is to judge how likely it is we will successfully not kill ourselves off and develop simulations
and want to populate them with non-trivial numbers of "people"
It's important to point out that these "simulated" people would have to be real, conscious people (though automata may also exist) and hence are not really "simulations" but real people with a (perhaps) radically different physical substrate instantiating them. Searle pointed out how consciousness, being a real thing, must therefore be created by some real-world physics (whether known or not) and therefore cannot be purely derivative of information processing. And therefore the old philosophical notion of replacing neurons with electronics, or buckets and pulleys, probably would fail.
