4 karmaJoined


I think this almost perfectly describes my problem with these videos/accounts/sensor readings. The same thing that makes them better evidence of aliens also makes them less likely to be real evidence. The crazier the physical constraints, the more likely "if this is real, the explanation is extra-terrestrial" becomes, but the less likely "it is real" becomes. Evidence that significantly increases the probability of  "this is real" without significantly decreasing the probability of "if this is real, the explanation is extra-terrestrial" seems necessary yet elusive.

The discussion of UAPs lately reminds me of the "How would Magnus Carlsen beat me at chess?" example that is popular in alignment these days. The still-unexplained phenomena that people will demand explanations for must be rare and hard to explain without a lot of good observations, or they wouldn't still be unexplained.

It seems similar to assuming that dark matter must be far more mysterious than just a particle, because we have so much trouble confirming any explanation of it, despite the fact that its observed behavior tells us that it should be extremely hard to confirm for any methods available to us.

I think the fact that the accelerations are close to, but not, a complete violation of physics is the most interesting, but it depends on how likely you think it is that a non-extraterrestrial explanation for a rare phenomena would also not seem to violate those laws. Or how likely a non-extraterrestrial explanation might be to appear to violate the laws of physics before further investigation. I do think this actually would make me update a bit in favor of extra-terrestrials if I thought about it more.

I wish my thoughts on this were better formulated, but I've been avoidant of UAP stuff for a while because engaging with it usually left me very frustrated and annoyed, and I don't think it's something we are likely to make meaningful progress on.

Answer by Y_Zhang3

Why haven’t we even seen evidence of power seeking AIs expanding in the universe? Any answer to that should also explain why they haven't arrived here.

The easiest potential answer is that we see what we should expect to see, even if there were many earlier civilizations, because civilizations are so spread out through space. If that’s true, you just need to explain why they would be so rare. But we really can't be sure if any of the potential explanations are true.

Are we (naively) early?

This is a question of whether more independent births of civilizations exist in the future than have already existed. This is difficult to actually address. The universe will last for trillions of trillions of years, the integral of even a very low rate of planet formation over that time would vastly outnumber the planets that have already formed. A huge amount of dust is gravitationally bound to galaxies still waiting to fuel star and planet formation. Somewhere around 92% of the total baryonic mass that can form stars is still in the galactic halos. That could potentially mean 92% of Earth-like planets have yet to form. It’s at least an upper-bound.

This is where that 92% comes from:


It does seem like we are early if the formation of Earth-like planets is a good proxy for formation of civilizations, and if the emergence of each civilization is an independent event. That’s the “naive” part.

We don’t want to know how many planets, or even Earth-like planets, will form in the future. We want to know how many planets capable of birthing civilizations will form in the future. If we don’t like the statistical implications of simply being early: 

What might make the emergence of different civilizations not independent? How could one civilization reduce the number produced in the future? What prevents those 92% of planets yet to form from satisfying the condition “capable of birthing civilizations”?

Grabby aliens are an option. If every early civilization inevitably leads to an expanding, power-seeking AI, then that would prevent future civilizations from forming within its sphere of influence. That expansion would be slower than light, so we should see evidence of it before one hits us. Now the question could be “How late among the initial civilizations are we?” If we were very late, we might expect to actually see some evidence of grabby aliens out in the universe.

Great sentence from that paper:

if the Milky Way today contained another civilization, it is likely that Earth would be at least the ten billionth planet to host a civilization in the observable universe, which would eventually contain at least a hundred billion civilizations.

The solar system formed after more than 50% of other Earth-like planets in both the Milky Way and the observable universe. So “Where are the grabby aliens/power-seeking AIs?” is a very valid question. Even a slow rate of grabby alien expansion doesn’t explain why we haven’t seen anything. Given how late we are, it seems like civilizations would have to be quite rare. Or somehow all civilizations to form very close together in time. The Milky Way is only 100,000LY across. That’s the upper limit for how long ago grabby alien expansion could have started if we aren’t alone in the galaxy.


Bottom-left shows the formation of the solar system ⊙ outside the 50% contour.


Every statement above can be followed by 10 “unless”es, some of them more likely than what I said here.

The error bars are huge on everything.

I was assuming all civilizations create ever-expanding AIs, and not really differentiating between that and grabby aliens.

I definitely don’t want to give the impression this is rigorous or exhaustive. That paper is from 2015. A lot of work has gone into thinking about these things since. It’s probably outdated. I just wanted to use a couple numbers from it. It goes way more in-depth.

I tried to avoid committing to any specific anthropic reasoning by saying "if you have a statistical problem with being early"

David Kipping is a good source for more rigorous statistical analysis and constraining.