What evidence might convince you? Simply curious, I can’t think of any.
Well, the most obvious one is dying and finding myself in an apparent afterlife. But that’s a bit flippant. Any of the following would make assign the idea of an afterlife a high probability:
1) A collection of mediums who are able to talk to the dead and can demonstrate that they are actually talking to the dead. Evidence that they can actually do so would a) correctly passing on pass-phrases to living people that were agreed upon before people died b) access to large amounts of information that the individual should not otherwise have (e.g. give them a simple number theory problem and they should be able to get a solution from say Erdos. And physics problems from Feynman, etc. c) the systematic production of information known to no living individuals which is subsequently confirmed by archaeology (the exact location of certain lost cities and ships would be the most obvious result). d) the production of new results in multiple areas of learning (again, people like Erdos and Feynman presumably have been working since their deaths). Any single one of these would be strong evidence and at least 2 of them would be convincing. (There are more likely hypotheses to explain any single one of these four by itself. For example c could be by itself explained by access to something like an Akashic record which seems unlikely but about as likely as an afterlife. )
2) This one has two parts a) There exist prophets who demonstrate repeatedly the ability to engage in miracles (and not silly miracles like making a congealed goo become liquidy when you shake it once a year. Miracles means repeated, testable examples of serious violations of the laws of physics in a variety of different ways) and the ability to prophesy. b) those prophets assert there is an afterlife and give a coherent description of it.
3) Any evidence that we are in a simulation drastically increases my estimate that there’s an afterlife because keeping back up copies is a natural thing to do. That by itself isn’t enough to assign a high probability because the entities running the simulation may not care enough about intelligent life (Tangent: We could have a Singularity, start modifying stars and building Dyson spheres and all that fun stuff. Then some higher-level equivalent of a grad student gets annoyed because its simulation of a what a universe with only 3 spatial dimensions would look like has developed some sort of self-organizing gunk that’s messing with the macroscopic behavior. That’s going to push its PhD back a while. Don’t worry, it will keep the copy running long enough to figure out which parameters to tweak to prevent the self-organizing junk from showing up again.)
4) Any end-times scenario which fits in with a major religion that has an afterlife also elevates the chance of an afterlife. Thus, if most of the Evangelical Christians disappear along with every little child I will assign a high probability to the Rapture having just taken place and thus there’s an afterlife.
Note that these are just off the top of my head. They are by no means exhaustive.
I really don’t like saying that God can’t exist, there can’t be an afterlife, etc., because we don’t have evidence for which isn’t the same thing as evidence against. My reaction is make no beliefs, not update my beliefs against.
Well, by “beliefs against” do you mean certainty that they don’t exist? A good Bayesian won’t do that because assigning probability 0 or 1 to something is a bad idea. But, one can assign the existence of God or the existence of an afterlife a low probability. Both are extremely complicated hypotheses. That means they shouldn’t get assigned a high prior probability, regardless of whether one is assigning priors using Solomonoff induction or some other approach. The fact that people have been searching for evidence of an afterlife and have come up woefully sort adjusts that probability downwards if it goes in any direction.
Well, the most obvious one is dying and finding myself in an apparent afterlife. But that’s a bit flippant. Any of the following would make assign the idea of an afterlife a high probability:
1) A collection of mediums who are able to talk to the dead and can demonstrate that they are actually talking to the dead. Evidence that they can actually do so would a) correctly passing on pass-phrases to living people that were agreed upon before people died b) access to large amounts of information that the individual should not otherwise have (e.g. give them a simple number theory problem and they should be able to get a solution from say Erdos. And physics problems from Feynman, etc. c) the systematic production of information known to no living individuals which is subsequently confirmed by archaeology (the exact location of certain lost cities and ships would be the most obvious result). d) the production of new results in multiple areas of learning (again, people like Erdos and Feynman presumably have been working since their deaths). Any single one of these would be strong evidence and at least 2 of them would be convincing. (There are more likely hypotheses to explain any single one of these four by itself. For example c could be by itself explained by access to something like an Akashic record which seems unlikely but about as likely as an afterlife. )
2) This one has two parts a) There exist prophets who demonstrate repeatedly the ability to engage in miracles (and not silly miracles like making a congealed goo become liquidy when you shake it once a year. Miracles means repeated, testable examples of serious violations of the laws of physics in a variety of different ways) and the ability to prophesy. b) those prophets assert there is an afterlife and give a coherent description of it.
3) Any evidence that we are in a simulation drastically increases my estimate that there’s an afterlife because keeping back up copies is a natural thing to do. That by itself isn’t enough to assign a high probability because the entities running the simulation may not care enough about intelligent life (Tangent: We could have a Singularity, start modifying stars and building Dyson spheres and all that fun stuff. Then some higher-level equivalent of a grad student gets annoyed because its simulation of a what a universe with only 3 spatial dimensions would look like has developed some sort of self-organizing gunk that’s messing with the macroscopic behavior. That’s going to push its PhD back a while. Don’t worry, it will keep the copy running long enough to figure out which parameters to tweak to prevent the self-organizing junk from showing up again.)
4) Any end-times scenario which fits in with a major religion that has an afterlife also elevates the chance of an afterlife. Thus, if most of the Evangelical Christians disappear along with every little child I will assign a high probability to the Rapture having just taken place and thus there’s an afterlife.
Note that these are just off the top of my head. They are by no means exhaustive.
Well, by “beliefs against” do you mean certainty that they don’t exist? A good Bayesian won’t do that because assigning probability 0 or 1 to something is a bad idea. But, one can assign the existence of God or the existence of an afterlife a low probability. Both are extremely complicated hypotheses. That means they shouldn’t get assigned a high prior probability, regardless of whether one is assigning priors using Solomonoff induction or some other approach. The fact that people have been searching for evidence of an afterlife and have come up woefully sort adjusts that probability downwards if it goes in any direction.