For example, if sentient life exists elsewhere in the universe, your odds of being a human are vanishingly small. This would suggest sentient life does not exist elsewhere in the universe.
That’s not how the anthropic principle works.
The anthropic principle lets you compute the posterior probability of some value V of the world, given an observable W. The observable can be the number of humans who have lived so far, and the value V can be the number of humans who will ever live. The probability of a V where 100W < V is smaller than the probability of a V only a few times larger than W.
It’s unclear if you get to count transhumans and AIs in V, which is the same problem Yvain is raising here about whether to include bats and ants in the distribution.
You can’t conclude that there aren’t other planets with life because you ended up here, because the probability of different values of V doesn’t depend on the observable W. There’s no obvious reason why P(there are 9999 other planets with life | I’m on this planet here with life) / P(there are 9999 other planets with life) would be different than
P(there are 0 other planets with life | I’m on this planet with life) / P(there are 0 other planets with life).
(I divided by the priors to show that the anthropic principle takes effect only in the conditional probability; having a different prior probability is not an anthropic effect.)
Disclaimer: I’m a little drunk.
I’m troubled now that this formulation doesn’t seem to work, because it relies on saying “P(fraction of all humans who have lived so far is < X)”. It doesn’t work if you replace the “<” with an “=”. But the observable has an “=”.
BTW, outside transhumanist circles, the anthropic principle is usually used to justify having a universe fine-tuned for life, not to figure out where you stand in time, or whether life will go extinct.
The anthropic principle lets you compute the posterior probability of some value V of the world, given an observable W. The observable can be the number of humans who have lived so far, and the value V can be the number of humans who will ever live. The probability of a V where 100W < V is smaller than the probability of a V only a few times larger than W.
This argument could have been made by any intelligent being, at any point in history, and up to 1500AD or so we have strong evidence that it was wrong every time.
If this is the main use of the anthropic argument, then I think we have to conclude that the anthropic argument is wrong and useless.
I would be interested in hearing examples of applications of the anthropic argument which are not vulnerable to the “depending on your reference class you get results that are either completely bogus or, in the best case, unverifiable” counterargument.
(I don’t mean to pick on you specifically; lots of commentors seem to have made the above claim, and yours was simply the most well-explained.)
This argument could have been made by any intelligent being, at any point in history, and up to 1500AD or so we have strong evidence that it was wrong every time. If this is the main use of the anthropic argument, then I think we have to conclude that the anthropic argument is wrong and useless.
First, “the anthropic argument” usually refers to the argument that the universe has physical constants and other initial conditions favorable to life, because if it didn’t, we wouldn’t be here arguing about it.
Second, what you say is true, but someone making the argument already knows this. The anthropic argument says that “people before 1500AD” is clearly not a random sample, but “you, the person now conscious” is a random sample drawn from all of history, although a sample of very small size.
You can dismiss anthropic reasoning along those lines for having too small a sample size, without dismissing the anthropic argument.
Thank you for saying this. I agree. Since at least the time I made this comment, I have tentatively concluded that anthropic reasoning is useless (i.e. necessarily uninformative), and am looking for a counterexample.
That’s not how the anthropic principle works.
The anthropic principle lets you compute the posterior probability of some value V of the world, given an observable W. The observable can be the number of humans who have lived so far, and the value V can be the number of humans who will ever live. The probability of a V where 100W < V is smaller than the probability of a V only a few times larger than W.
It’s unclear if you get to count transhumans and AIs in V, which is the same problem Yvain is raising here about whether to include bats and ants in the distribution.
You can’t conclude that there aren’t other planets with life because you ended up here, because the probability of different values of V doesn’t depend on the observable W. There’s no obvious reason why P(there are 9999 other planets with life | I’m on this planet here with life) / P(there are 9999 other planets with life) would be different than P(there are 0 other planets with life | I’m on this planet with life) / P(there are 0 other planets with life).
(I divided by the priors to show that the anthropic principle takes effect only in the conditional probability; having a different prior probability is not an anthropic effect.)
Disclaimer: I’m a little drunk.
I’m troubled now that this formulation doesn’t seem to work, because it relies on saying “P(fraction of all humans who have lived so far is < X)”. It doesn’t work if you replace the “<” with an “=”. But the observable has an “=”.
BTW, outside transhumanist circles, the anthropic principle is usually used to justify having a universe fine-tuned for life, not to figure out where you stand in time, or whether life will go extinct.
This argument could have been made by any intelligent being, at any point in history, and up to 1500AD or so we have strong evidence that it was wrong every time. If this is the main use of the anthropic argument, then I think we have to conclude that the anthropic argument is wrong and useless.
I would be interested in hearing examples of applications of the anthropic argument which are not vulnerable to the “depending on your reference class you get results that are either completely bogus or, in the best case, unverifiable” counterargument.
(I don’t mean to pick on you specifically; lots of commentors seem to have made the above claim, and yours was simply the most well-explained.)
First, “the anthropic argument” usually refers to the argument that the universe has physical constants and other initial conditions favorable to life, because if it didn’t, we wouldn’t be here arguing about it.
Second, what you say is true, but someone making the argument already knows this. The anthropic argument says that “people before 1500AD” is clearly not a random sample, but “you, the person now conscious” is a random sample drawn from all of history, although a sample of very small size.
You can dismiss anthropic reasoning along those lines for having too small a sample size, without dismissing the anthropic argument.
Thank you for saying this. I agree. Since at least the time I made this comment, I have tentatively concluded that anthropic reasoning is useless (i.e. necessarily uninformative), and am looking for a counterexample.
Best time to do anthropic reasoning. Save the sane reasoning for when you’re sober! ;)