This question has little to do with LW-style rationality. The contrivance of postulating extra entities like souls becomes obvious once you accept that mind is a process in a living brain, affected by drugs, injuries, etc. You can, of course, postulate anything you want, but the odds are against you, and definitely not 50⁄50. See this illuminating debate, discussed here. Moreabout the 50⁄50 fallacy you’ve fallen prey to.
Edit: fixed the debate video link to a better-quality one.
I don’t think I’m committing the Fallacy of Gray. The Fallacy of Gray is when you treat probabilities between 0 and 100 as the same. This instance is a question is of what to do with a completely unknown probability.
Regarding the 50⁄50 fallacy, there’s two questions:
1) What is the likelihood of something happening for which you have no data that would allow you to predict it?
2) What is the likelihood of an afterlife?
The second question is the overarching question of this post, for which I may be wrong about the probability. But what I meant in the past comment is that I don’t think the Fallacy of Gray applies to the first question. Sorry that I wasn’t clear about that. Now that I’ve clarified, do you think it would be committing the Fallacy of Gray to say that something for which you have no data on has a 50⁄50 shot of occurring?
You don’t seem to understand what it’s really like to “have no data.” A question with no data is something like: Will the Emperor of Alfa Centauri eat fried lummywaps or boiled sanquemels today?
On human consciousness we have lots and lots of data, more than enough to predict confidently that it’s lost forever when we die.
What is the likelihood of something happening for which you have no data that would allow you to predict it
It depends on where your priors come from. If you mean Knightian uncertainty, then this is a whole area of research. Afterlife is not like that.
There is plenty of experimental data to falsify the afterlife model. If you take the souls model seriously, there are concrete testable predictions it makes, these have been investigated and found to be false. The video, among many other sources, discusses a bunch of them. Based on the results of these experiments we can evaluate the probability of afterlife, and the number is tiny, though the exact number would depend on who does the calculations.
If you take the souls model seriously, there are concrete testable predictions it makes, these have been investigated and found to be false. The video, among many other sources, discusses a bunch of them.
I don’t have time to watch the video, can you give an example? The only experiments I can think of leave the experimenter apparently unable to report the results.
Edit: Ok, there have been reports of some experimenters who have successfully reported positive results, but not in a reliable or reproducible manner.
First, feel free to state your favorite model, or at least some starting point, e.g. “all information in the brain is duplicated in an extra-brane substrate not interacting with the currently known physical forces, such as electromagnetism”, or “the mind is not located in the brain, the latter is only a conduit it uses to communicate with other minds”, and we can start refining it, eventually finding what it predicts. Then we can start looking whether relevant experiments have been done.
For example, the observation that damaging a certain part of one’s brain leads to cognitive changes may or may not be relevant, depending on the model used.
Similarly, noting that ape brains are very similar to human, yet monkeys have no souls is only an argument against some specific models.
Another example: an experiment that tests whether a person reporting an out-of-body experience is able to read a message secretly hidden on the upward-facing part of a ceiling fan would not falsify an epiphenomenal model where a soul simply accumulates the memories and personality, only to separate some time after the brain death becomes irreversible.
Sure, if you want to put it this way. I see it more as making a commitment to analyzing a single afterlife model vs the no-afterlife version “consciousness is a process in a living brain”. Committing to analyzing a single clearly defined model helps against inadvertently moving the goalposts if a contrary evidence is found. An explicit goalpost moving is fine, as long as it is of the form “we have found this model to be invalid, but we can construct an alternative model which does not suffer from the same weakness, so let’s reject the original model and consider this one instead”. I see it done in Physics all the time, by the way. So, please do pick a model. (Also, I wasn’t the one who downvoted your reply. Downvoting a living conversation is one sure way to end it.)
1) What is the likelihood of something happening for which you have no data that would allow you to predict it?
Do you mean anything that fits that criteria, or a given thing that fits it? I find it quite likely that there is a large number of things that are true that we have no realistic way of finding out, but it’s dwarfed by the number of things that are false that we have no realistic way of finding out. If there are five things that are true that you can’t verify, so you decide to believe five unverifiable things, it just means you’ll be wrong ten times instead of five.
This question has little to do with LW-style rationality. The contrivance of postulating extra entities like souls becomes obvious once you accept that mind is a process in a living brain, affected by drugs, injuries, etc. You can, of course, postulate anything you want, but the odds are against you, and definitely not 50⁄50. See this illuminating debate, discussed here. More about the 50⁄50 fallacy you’ve fallen prey to.
Edit: fixed the debate video link to a better-quality one.
I don’t think I’m committing the Fallacy of Gray. The Fallacy of Gray is when you treat probabilities between 0 and 100 as the same. This instance is a question is of what to do with a completely unknown probability.
Watch the debate I linked, please. The probability is not unknown, it’s vanishingly small.
I will watch it as I go to sleep tonight.
Regarding the 50⁄50 fallacy, there’s two questions: 1) What is the likelihood of something happening for which you have no data that would allow you to predict it? 2) What is the likelihood of an afterlife?
The second question is the overarching question of this post, for which I may be wrong about the probability. But what I meant in the past comment is that I don’t think the Fallacy of Gray applies to the first question. Sorry that I wasn’t clear about that. Now that I’ve clarified, do you think it would be committing the Fallacy of Gray to say that something for which you have no data on has a 50⁄50 shot of occurring?
You don’t seem to understand what it’s really like to “have no data.” A question with no data is something like: Will the Emperor of Alfa Centauri eat fried lummywaps or boiled sanquemels today?
On human consciousness we have lots and lots of data, more than enough to predict confidently that it’s lost forever when we die.
I get your point, but your example is poor—I think we have more than enough data to answer this question: No, with 99%+ probability.
It depends on where your priors come from. If you mean Knightian uncertainty, then this is a whole area of research. Afterlife is not like that.
There is plenty of experimental data to falsify the afterlife model. If you take the souls model seriously, there are concrete testable predictions it makes, these have been investigated and found to be false. The video, among many other sources, discusses a bunch of them. Based on the results of these experiments we can evaluate the probability of afterlife, and the number is tiny, though the exact number would depend on who does the calculations.
I don’t have time to watch the video, can you give an example? The only experiments I can think of leave the experimenter apparently unable to report the results.
Edit: Ok, there have been reports of some experimenters who have successfully reported positive results, but not in a reliable or reproducible manner.
First, feel free to state your favorite model, or at least some starting point, e.g. “all information in the brain is duplicated in an extra-brane substrate not interacting with the currently known physical forces, such as electromagnetism”, or “the mind is not located in the brain, the latter is only a conduit it uses to communicate with other minds”, and we can start refining it, eventually finding what it predicts. Then we can start looking whether relevant experiments have been done.
For example, the observation that damaging a certain part of one’s brain leads to cognitive changes may or may not be relevant, depending on the model used.
Similarly, noting that ape brains are very similar to human, yet monkeys have no souls is only an argument against some specific models.
Another example: an experiment that tests whether a person reporting an out-of-body experience is able to read a message secretly hidden on the upward-facing part of a ceiling fan would not falsify an epiphenomenal model where a soul simply accumulates the memories and personality, only to separate some time after the brain death becomes irreversible.
So, pick your model.
So you’re asking me to pick some hypothesis to privilege.
Sure, if you want to put it this way. I see it more as making a commitment to analyzing a single afterlife model vs the no-afterlife version “consciousness is a process in a living brain”. Committing to analyzing a single clearly defined model helps against inadvertently moving the goalposts if a contrary evidence is found. An explicit goalpost moving is fine, as long as it is of the form “we have found this model to be invalid, but we can construct an alternative model which does not suffer from the same weakness, so let’s reject the original model and consider this one instead”. I see it done in Physics all the time, by the way. So, please do pick a model. (Also, I wasn’t the one who downvoted your reply. Downvoting a living conversation is one sure way to end it.)
Do you mean anything that fits that criteria, or a given thing that fits it? I find it quite likely that there is a large number of things that are true that we have no realistic way of finding out, but it’s dwarfed by the number of things that are false that we have no realistic way of finding out. If there are five things that are true that you can’t verify, so you decide to believe five unverifiable things, it just means you’ll be wrong ten times instead of five.