[csee]

> Before the great filter, there's the jump to multi-cellular life.

Well, the filter may exist prior to the jump to multi-cellular life. Finding early life on Mars would necessarily be bad because we would now be able to mostly eliminate that possibility, which means that the probability that the filter is ahead of us increases.

[s1artibartfast]

I disagree. While finding dead life on other planets decreases the probability of some filters, it also increases the probability of other filters.

For example, if we go to every planet in the solar system and see evidence of extinct multi-cellular life, but no sapient life, this would suggest that the filter is indeed behind us, and not in front of us.

[bpodgursky]

I don't think this is right. It moves the filter forward from "no life" to "no intelligent life" (If we have a TON of examples – remember, we are also a data point. so even if we find 200 planets with dead multicellular life, we still have 1/200 making it to spacefaring — that's a huge % when extrapolated out to a galactic scale! It's a large enough % that I don't even find this to be a meaningful prior adjustment).

It also adds no information as to which, if any, filters are in front of us. How would "lot of dead fish" help us understand "do civilizations destroy themselves with nuclear weapons"?

[s1artibartfast]

I agree with the first part, but not the second. Agreed it would rule out some early filters. But for the second part, the ratio doesn't go to 1 in 200. That's not how the anthropic principle works. You're already assuming that the Earth is extraordinary oh, so you can't just put it in the data set. Instead you would look at 199 out of 199 planets where sapient life capable of making an observation failed to develop.

Agreed that it provides no data on future filters ahead of us

[csee]

> Agreed that it provides no data on future filters ahead of us

This was my instinct, too, after reading both of your posts, but on reflection I think that it's unlikely to be no information.

If we instead found 199 examples of extinct complex life (say, fossilized rodent-like creatures), instead of extinct simple life, we all probably agree that this would increase the odds of the filter being in front of us.

Now, imagine gradually changing the example from extinct complex life (rodent-type creatures) and going down to extinct simple life. At what point in this thought experiment does it become "no data"?

[s1artibartfast]

I guess I agree that discovering extinct life is information about probabilistic filters in earths future. I But, it does have to be weighed against the newly observed data supporting filters humans have already past.

To take your example, as we keep finding rodent fossils, we become more confident that a filter exists between rat and rocket-men. As we go from 0 planets with rodent fossils to 1 trillion planets, the probability that this is the filter keeps going up.

That is to say, we are not only decreasing the chance of pre-rat filters, but simultaneously increasing the chance that the filter is in the rat stage.