Anthropic effects imply that we are more likely to live in the universe with interstellar panspermia

post by avturchin · 2020-03-10T13:12:54.991Z · score: 11 (4 votes) · LW · GW · 18 comments

The sketch of the argument which favours panspermia.

1. There are around 10 billion potentially habitable planets in our galaxy.

2. Interstellar panspermia of small organisms will inseminate all of them with life in a few billion years.

3. Abiogenesis could be the main filter. Recent estimate of the minimum length of self-replicating RNA is around 100 bases. Totani thinks that only one of 10^100 Suns will generate correct RNA and start RNA-world. Thus only one of 10^80 of universes has life. Thus it looks like Rare Earth is true and we are alone.

4. However, there are many different universes in the multiverse and some of them may allow panspermia.

5. The density of planets with life in a pasnpermia-universe is 10 billion times more than such density in a non-panspermia universe as each life will colonise the whole galaxy.

6. We don't know the relation of non-panspermia to panspermia-universes. If it is less than 10 billion, we are more likely to be in panspermia-universe based on anthropic considerations, because each of them will have 10 billions times more attempts to create a civilization of a planet with life.

7. If we are in the panspermia universe, it means that the Great filter from the Fermi paradox is likely ahead, as major abiogenesis filter doesn't prevent many planets in our galaxy from having life. Rare Earth hypothesis is false locally.

8. It means that either there are a lot of aliens nearby or almost all civilizations self destruct. (Or there are some other earlier great filters like a higher level of asteroid impacts etc.) Not good from x-risks view point.

18 comments

Comments sorted by top scores.

comment by James_Miller · 2020-03-10T20:30:52.376Z · score: 4 (2 votes) · LW(p) · GW(p)

What if we update on the age of the universe? Imagine that the normal course of events after a high tech civilization arises is for it to grab all the free energy it can as fast as it can and universes where life forms easily do not have civilizations at our level of development at the current age of the universe.

comment by avturchin · 2020-03-11T12:15:10.109Z · score: 2 (1 votes) · LW(p) · GW(p)

Panspermia implies that civilizations will appear earlier, because the number of planets with life is growing exponentially. This also means that first civilizations will likely be of approximately the same age. This idea was used by the proponents of SETI as justification of the search of radiosignals inside our Galaxy.

comment by Decius · 2020-03-10T19:39:53.043Z · score: 4 (2 votes) · LW(p) · GW(p)

Doesn't the anthropic effects also imply that abiogenesis is easier than studies suggest it is, for reasons that we don't yet understand?

If the chance of our science looking like it does, given that abiogenesis is trivial, is as high as one in ten billion, wouldn't the anthropic effect make it an even proposition? Is the evidence against abiogenesis being trivial really 10,000,000,000:1 against? (roughly p=10^-11)?

comment by avturchin · 2020-03-10T20:00:43.103Z · score: 3 (2 votes) · LW(p) · GW(p)

The question could reformulated: "what is more easy: panspermia or abiogenesis?" If we look at Earth, we could see that "panspermia" between different parts of the Earth worked better than abiogenesis. We have only one type of life here.

The answer could come from Mars: if we find life there, and it is of different origin than one on Earth, than abiogenesis is easier. If there will be ours type of life, it will be evidence for easier panspermia.

comment by avturchin · 2020-03-11T11:47:48.537Z · score: 2 (1 votes) · LW(p) · GW(p)

Also, this article suggests irreducible difficulty of abiogenesis: it needs random generation of the correct string of around 100 bases long. This difficulty is rather general and could be applied to any universe: any life form should pass through random generation of the first self-replicating unit. Such random generation requires enormous amount of attempts.

comment by Donald Hobson (donald-hobson) · 2020-09-09T11:36:53.863Z · score: 2 (1 votes) · LW(p) · GW(p)
the correct string of around 100 bases long.

Which is "the correct string"? Many different strings could self replicate.

Such random generation requires enormous amount of attempts.

Only if the fraction of molecules that can self replicate is microscopic.

comment by avturchin · 2020-09-09T12:36:27.985Z · score: 2 (1 votes) · LW(p) · GW(p)

Here we speak about minimum length of self-replicating string. If first such string 100 bits and the next is 110, the second one will be 1000 times less probable during random generation of strings. Thus longer strings could be ignored.

comment by G Gordon Worley III (gworley) · 2020-03-10T17:58:29.905Z · score: 4 (2 votes) · LW(p) · GW(p)

I think this is line of argument is also consistent with life being easy, i.e. life frequently spontaneously appears millions of times throughout the universe basically whenever the conditions are even moderately right for it. We could replace "panspermia" with "easy life" or equivalently "late Great Filter", in which case I think all is equal between these options within this argument and so we have to choose between them by non-anthropic reasoning, for example arguing non-anthropically whether panspermia or easy life is more likely.

comment by avturchin · 2020-03-10T19:05:24.801Z · score: 2 (1 votes) · LW(p) · GW(p)

Agree. The main difference between panspermia-universe and easy-life one will be the life in other galaxies. In panspermia-universe, life will be only in our galaxy, so it may have some special features, like different spectrum in some waves because of higher abundance of organic molecules.

comment by Teerth Aloke · 2020-05-04T12:49:57.865Z · score: 3 (2 votes) · LW(p) · GW(p)

Well the probability might be even lower, according to Eugene Koonin.

https://arxiv.org/ftp/q-bio/papers/0701/0701023.pdf

comment by Jiro · 2020-03-11T18:36:30.527Z · score: 2 (1 votes) · LW(p) · GW(p)

This is similar to the simulation hypothesis, and in fact is sometimes used as a response to the simulation hypothesis.

comment by Daniel Kokotajlo (daniel-kokotajlo) · 2020-03-10T23:06:50.589Z · score: 2 (1 votes) · LW(p) · GW(p)

I disagree with 5. In about a hundred years or so, probably, a shockwave of interstellar colonization will radiate out from Earth and convert the visible universe into life (or simulated life). Even though most of that life will be very different from ours, enough of it will be like ours that I don't think it's fair to say the density of life-like-ours is higher in panspermia universes than in non-panspermia universes.

comment by avturchin · 2020-03-11T12:18:16.231Z · score: 2 (1 votes) · LW(p) · GW(p)

But our descendant 100 years from now probably will know this, so they are not exactly like us and our not members of reference class in question.

comment by Daniel Kokotajlo (daniel-kokotajlo) · 2020-03-11T14:17:50.768Z · score: 2 (1 votes) · LW(p) · GW(p)

Yes, but enough of them won't know this. I'm thinking of ancestor simulations in particular, and zoo planets, and stuff like that.

comment by avturchin · 2020-03-11T14:47:53.748Z · score: 2 (1 votes) · LW(p) · GW(p)

This argument would work only if there is a difference between simulations created by civilizations in paspermia-universes - and simulations created by civilizations in non-paspermia universes. And such difference should be inverse to screw the relation.

For example, if most of simulations are trustworthy past simulations, when panspermia-civilizations will simulate past of panspermia-civilizations including the outer space with panspermia and vice versa. Adding simulations in the equation is not changing final result.

Putting it more simply: there are more civiliations in panspermia universes and more simulations created by these civilizations.

comment by Daniel Kokotajlo (daniel-kokotajlo) · 2020-03-11T16:35:04.469Z · score: 2 (1 votes) · LW(p) · GW(p)

I disagree that there are (many) more civilizations and simulations in panspermia universes. Panspermia spreads life at a much slower rate than colonization shockwaves do, and so the colonization shockwave will catch up to and surpass the fruits of panspermia. The pre-colonization, post-panspermia civilizations will be a drop in the bucket compared to the simulations and whatnot created post-colonization. So yeah there will be more, but only a tiny drop more.

comment by avturchin · 2020-03-11T18:00:47.873Z · score: 4 (2 votes) · LW(p) · GW(p)

I got your idea, but still not convinced. I think that there will be more civilizations in panspermia-universes, because panspermia helps to evade other types of early filters:

If there is only one planet with life in a Galaxy, it may have only 1 in 1000 chance to survive other risks like asteroid impacts, and if it doesn't survive, this galaxy is done.

In the panspermia-galaxy, there are 10 billion planets with life, and at least some of them will survive all other early filters. Thus panspermia-galaxy will almost certainly create an intelligence explosion wave.

So, for 1000 singularities in the panspemia galaxies there will be only one in non-panspermia-galaxy with life. So if I am in a simulation of whatever type, it is 1000 times more probable that it is created by a singularity in a panspermia-galaxy.

However, inside a simulation could be simulated a world with different laws of panspermia than in real world of simulators. To object this, I suggest general principle: "Most facts about outside world in the most simulations are true". I will prove it based on the contrary evidence: Imagine that in all simulations created by all possible civilizations there is a lie A, while in the real world non-A is true. ( e.g. 2=2=5 or Sun's size is square). In that case, there should be a coordination process which is applicable to all possible civilizations which create simulations. However, there is no such physical process. So all simulations can't share one lie. (There is one exception: all simulations lie that they are real world).

Thus, any random fact A about the outside world in a simulation is unlikely to be false (but could be). The fact that we are in panspermia galaxy or not is a random fact and all simulators can't coordinate to lie in one way about it. Thus we should give high credence to the idea that our simulation represents the real state of affairs regarding the probability of panspermia in our galaxy. Thus, being in a simulation or not doesn't affect significantly our estimations about the type of the galaxy where we originated.

comment by Daniel Kokotajlo (daniel-kokotajlo) · 2020-03-11T21:20:20.227Z · score: 4 (2 votes) · LW(p) · GW(p)

Ahhh, OK now I think I agree with you. Thanks.

Not sure about the proof by contrary, I'll need to think about it more.