20 Comments

But that is just a direct consequence of the pdfs being near-constant, as we've agreed.

Expand full comment

That doesn't sound right to me either.

Expand full comment

Thanks Robin. It was the repeated refs to the distributions being exponential that threw me. While technically correct, they made it harder for me to see that you were really comparing Earth’s durations with sets of values obtained by N repeated draws from a *nearly uniform* distribution over [0,W-S], after discarding those sets whose sum exceeded W-S.

Expand full comment

Yes the pdf is near constant, but no we aren't using any more than the total time constraint.

Expand full comment

"if an oasis succeeds in doing all its required hard steps within its W-S duration, then the time durations required to do each of the hard steps are all drawn from the same (roughly exponential) distribution, regardless of the value of E for those steps!"

OK but isn't the pdf for each of these time durations nevertheless essentially constant over the relevant N-dimensional region (in which the ith duration Di is at most W-S for all i)? And if so, why would you expect the observed values for Earth to be any closer than what’s implied by the constraint that the Di sum to no more than W-S?

Why I think the pdfs are essentially constant: By the definition of 'hard step', the expected value E of each step >> W. But for an exponential distribution, E = 1/L (writing L for lambda), so 1/L >> W => LW << 1 => L(W-S) << 1 => the value L*exp(-LDi) of the pdf of the (un-truncated) exponential distribution for Di is always, within the relevant truncation region, roughly L.

Expand full comment

Yes, there are surely many constraints and factors I'm not considering above. And if they correlate with the features I examine, my analysis could be misleading.

Expand full comment

That would be about future filters. This post is focused on explaining data with past filters.

Expand full comment

You must discuss the "political" filter in a future post. I'm afraid that's the one that may be key, at least for us https://www.nationalreview....

Expand full comment

You can represent a chance of something causing extinction via a try-once step.

Expand full comment

Any ideas about positive and negative filters?Positive ones: thing that should happen, like sex.Negative ones: thing which should not happen like large asteroid impacts or wrong genetics.

Expand full comment

I've now added that link to the post.

Expand full comment

There is page in wiki about it and it may have some links https://en.wikipedia.org/wi....

Expand full comment

Those sound like plausible effects to consider; has anyone worked them out systematically to get relative rates of habitable planets?

Expand full comment

That certainly seems like something plausible, but I think that just ends up equivalent to try-try steps.

Expand full comment

I have read somewhere that smaller stars seems to have intense magnetic explosions, and as potentially habitable planets are closer to smaller stars, these explosions will strip atmospheres from earth-like planets; they also tend to tidaly lock planets, so they face the star one side, and it is bad for life. Also, most terrestrial planets have nuclear energy for plate tectonics for less than 10 billion years, so longer living stars are useless, as without plate tectonics they will lose their atmospheres etc.

Expand full comment

Also, I worry a bit that in some sense one of the major things to be explained here is choosen preciscely because it's hard to explain. The reason we are talking about the fact that we are seemingly pretty late in the universe but don't see any other life is something we're thinking about *because* it seems hard to explain. I mean in a universe in which we found lots of creatures about we wouldn't be asking this sort of question and thinking about it.

Does that mean this is kinda like p-hacking and we should be adjusting our expectations because we specifically choose a notion that's hard to explain so maybe it shouldn't surprise us that is a low probability aspect of our experience? I dunno, all the analysis seems pretty plausible to me so maybe not but I thought maybe someone could explain why we don't need to correct for this.

Expand full comment