So finding evidence of life that went extinct at any stage whatsoever should make us revise our beliefs about the Great Filter in the same direction? Doesn’t this violate conservation of expected evidence?
Is there a counter-weighing bit of evidence every time we don’t find evidence of life at all, and every time (if ever) we find evidence of non-extinct life?
According to Hanson’s article, non-extinct life that didn’t reach sentience counts as failing the Great Filter, and no life at all also counts as failing at a very early stage. I believe my point still stands.
No, the total evidence for a great filter is conserved (lack of observable galactic colonization), the evidence merely shifts where we expect this great filter to be.
Let’s have another go. By Bostrom’s logic, witnessing a failure of life at any stage (which includes life failing to develop in the first place) implies that the great filter happens later than we thought, because each failure tells us that the steps that preceded it (e.g. planet formation, liquid water, etc) probably didn’t include the great filter. But life eventually fails on all planets except ours: the “silence of the sky” is a background assumption for the whole thesis. So what kind of evidence would tell us that the filter happens earlier than we thought?
Failing to find any life whatsoever on Mars would be evidence for the great filter being development of life or earlier (and thus evidence for the GF being earlier than we thought), but it is only very weak evidence of that, since even if life were very common (say 20% of all planets in the liquid water zone) we still wouldn’t be very surprised by the absence of life on Mars in particular. Matters would be different after investigating hundreds of planets and failing to find signs of life anywhere.
Finding (independent) life at any stage would be evidence for whatever step this life failed to make and/or whatever wiped it out being the great filter, but for any reasonably well understood step or mechanism of extinction (e. g. Mars losing most of its atmosphere) the shift in probability will be much smaller than the shift in probability for having life in the first place, which is a total unknown, and such life would also be evidence against a black swan before that point without discriminating between black swans between that point and us and after us. So the slightly increased probability of that particular GF wouldn’t come anywhere close to making up the lost probability mass earlier, leaving a great filter after us much more likely.
If we discover life wiped out by a black swan that would shift a lot of probability mass to that black swan, but it would have to be something not taken into account at all before, seeming certain to happen to almost all life after discovery, and also something that would be very unlikely to happen to us in the future if it was to make up for the lost probability mass for any earlier GF. A sufficiently certain seeming former black swan could even shift probability mass away from a GF after us, but that’s not the way I’d bet.
the shift in probability will be much smaller than the shift in probability for having life in the first place, which is a total unknown
This is a crucial step in your argument. It depends on our initial prior for extraterrestrial life being very low. If that prior were slightly higher, the argument would work just as well in reverse, or maybe balance out. There’s something icky about this whole business.
So finding evidence of life that went extinct at any stage whatsoever should make us revise our beliefs about the Great Filter in the same direction? Doesn’t this violate conservation of expected evidence?
Is there a counter-weighing bit of evidence every time we don’t find evidence of life at all, and every time (if ever) we find evidence of non-extinct life?
According to Hanson’s article, non-extinct life that didn’t reach sentience counts as failing the Great Filter, and no life at all also counts as failing at a very early stage. I believe my point still stands.
No, the total evidence for a great filter is conserved (lack of observable galactic colonization), the evidence merely shifts where we expect this great filter to be.
Let’s have another go. By Bostrom’s logic, witnessing a failure of life at any stage (which includes life failing to develop in the first place) implies that the great filter happens later than we thought, because each failure tells us that the steps that preceded it (e.g. planet formation, liquid water, etc) probably didn’t include the great filter. But life eventually fails on all planets except ours: the “silence of the sky” is a background assumption for the whole thesis. So what kind of evidence would tell us that the filter happens earlier than we thought?
Failing to find any life whatsoever on Mars would be evidence for the great filter being development of life or earlier (and thus evidence for the GF being earlier than we thought), but it is only very weak evidence of that, since even if life were very common (say 20% of all planets in the liquid water zone) we still wouldn’t be very surprised by the absence of life on Mars in particular. Matters would be different after investigating hundreds of planets and failing to find signs of life anywhere.
Finding (independent) life at any stage would be evidence for whatever step this life failed to make and/or whatever wiped it out being the great filter, but for any reasonably well understood step or mechanism of extinction (e. g. Mars losing most of its atmosphere) the shift in probability will be much smaller than the shift in probability for having life in the first place, which is a total unknown, and such life would also be evidence against a black swan before that point without discriminating between black swans between that point and us and after us. So the slightly increased probability of that particular GF wouldn’t come anywhere close to making up the lost probability mass earlier, leaving a great filter after us much more likely.
If we discover life wiped out by a black swan that would shift a lot of probability mass to that black swan, but it would have to be something not taken into account at all before, seeming certain to happen to almost all life after discovery, and also something that would be very unlikely to happen to us in the future if it was to make up for the lost probability mass for any earlier GF. A sufficiently certain seeming former black swan could even shift probability mass away from a GF after us, but that’s not the way I’d bet.
This is a crucial step in your argument. It depends on our initial prior for extraterrestrial life being very low. If that prior were slightly higher, the argument would work just as well in reverse, or maybe balance out. There’s something icky about this whole business.