Welcome Andrei! Iāve actually mused on this question a lot myself, and I agree itās under-discussed!
What immediately jumps to mind is that this postās argument requires SSA, a view of anthropics (the study of how one should reason about their own existence). Under SSA, you are randomly sampled from a āreference classā of beings. You rightly conclude that under SSA, being born a human is extremely unlikely, so your existence seems to be strong evidence against nonhuman sentience. (This would also imply the sum of artificial and/āor future sentience wonāt be much more than the sum of past sentience, also known as the doomsday argument.)
However, SSA is not the only view of anthropics. SIA is another view (heavilypromoted by EA blogger Benthamās Bulldog) which says that worlds with more beings who could be you are more likely, in such a way which (by design) cancels this postās argument, as well as the doomsday argument and several others. Iāve personally always been more convinced by SIA.
But I donāt think itās even necessary to settle anthropics for this. As you wrote, thereās a post-birth filter where only a tiny % of people will have come up with this argument. Itās like if thereās a lottery where you get sent a letter only if you win, and then you get one and think āwow, the lottery must have been rigged in my favorā. But winning was guaranteed given youāve received that letter. In your last paragraph, it seems like you exempt your own birth from this: āI rolled a oneā. But thatās just what any lottery winner would say! Thatās why Iām personally still convinced by this filter.
Thank you for the comment Ariel! Iām finding it pretty surreal learning about how much thinking has actually been done on the topic after some more digging. You are right that this entire post requires me to have been born for the probabilities to be the same as the number of births of different species, which is not necessarily true, since I might have also not been born at all.
I agree that any argument made by someone receiving the letter is not good evidence of the lottery being rigged, since they were always going to say that. Only winners will know that the lottery existed in the first place, and only they will tell you how it is likely to be rigged. So I agree that you hearing this argument from me should not be evidence of anything. By design, this seems like an argument that cannot convince another person. Is that true for a person arguing about their own life as well? It is my own sentience that makes me think I played the lottery (and could have lost, e.g. been a chicken), but that is simply an intuition I have. You might or might not have the same intuition.
I guess you could make the argument that there are many other lotteries I never knew I was part of to begin with. If there were a billion different lotteries played at birth, on average I would be expected to win a lottery, and I guess I simply ended up winning the sentience one. This would invalidate my sentience argumentbecause of the filter.
Welcome Andrei! Iāve actually mused on this question a lot myself, and I agree itās under-discussed!
What immediately jumps to mind is that this postās argument requires SSA, a view of anthropics (the study of how one should reason about their own existence). Under SSA, you are randomly sampled from a āreference classā of beings. You rightly conclude that under SSA, being born a human is extremely unlikely, so your existence seems to be strong evidence against nonhuman sentience. (This would also imply the sum of artificial and/āor future sentience wonāt be much more than the sum of past sentience, also known as the doomsday argument.)
However, SSA is not the only view of anthropics. SIA is another view (heavily promoted by EA blogger Benthamās Bulldog) which says that worlds with more beings who could be you are more likely, in such a way which (by design) cancels this postās argument, as well as the doomsday argument and several others. Iāve personally always been more convinced by SIA.
But I donāt think itās even necessary to settle anthropics for this. As you wrote, thereās a post-birth filter where only a tiny % of people will have come up with this argument. Itās like if thereās a lottery where you get sent a letter only if you win, and then you get one and think āwow, the lottery must have been rigged in my favorā. But winning was guaranteed given youāve received that letter. In your last paragraph, it seems like you exempt your own birth from this: āI rolled a oneā. But thatās just what any lottery winner would say! Thatās why Iām personally still convinced by this filter.
Thank you for the comment Ariel! Iām finding it pretty surreal learning about how much thinking has actually been done on the topic after some more digging. You are right that this entire post requires me to have been born for the probabilities to be the same as the number of births of different species, which is not necessarily true, since I might have also not been born at all.
I agree that any argument made by someone receiving the letter is not good evidence of the lottery being rigged, since they were always going to say that. Only winners will know that the lottery existed in the first place, and only they will tell you how it is likely to be rigged. So I agree that you hearing this argument from me should not be evidence of anything. By design, this seems like an argument that cannot convince another person. Is that true for a person arguing about their own life as well? It is my own sentience that makes me think I played the lottery (and could have lost, e.g. been a chicken), but that is simply an intuition I have. You might or might not have the same intuition.
I guess you could make the argument that there are many other lotteries I never knew I was part of to begin with. If there were a billion different lotteries played at birth, on average I would be expected to win a lottery, and I guess I simply ended up winning the sentience one. This would invalidate my sentience argument because of the filter.