Kipping on Grabby Aliens

If you recall, our grabby aliens analysis tries to explain 3 key datums: a) we don’t see any huge alien civs in our sky b) we have arrived at a very early date in the universe, and c) there is only ~1Gyr left for life on Earth. Each datum allows us to estimate a matching parameter of our 3 parameter model: a) alien civ expansion rate, b) time constant for appearance rate, c) power in power law of appearance rate. We can thus roughly estimate the spacetime distribution of such visible alien civs.

In a new video, David Kipping respectfully disagrees:

I have three major criticisms …

[1.] Grabby aliens is predicated among the Hard Steps Model, which we don't know to be true.

[2.] There is a much easier [solution] available to us, … that M dwarfs are not habitable to begin with.

[3.] The theory invokes what to me is a kind of Get Out of Jail Free card. It states the grabby aliens are all expanding at speeds close to the speed of light. … [But] there is no experiment we can design that can falsify this hypothesis, even in principle.

Instead Kipping prefers: 

far simpler hypothesis is twofold. First, M dwarfs don't permit creatures like us to develop on them, solving the Red Sky Paradox and our apparently early emergence in cosmic history. Second, the emergence of aggressive expanding civilizations just doesn't happen for some reason. …

Earth's environment is changing on a timescale that is linked to the sun's main sequence lifetime, and those changing conditions would surely precipitate evolutionary adaptations for Earth's inhabitants. …

If intelligence bestows evolutionary advantages, then perhaps perhaps it is inevitable given a few billion years. … the probability of transitioning from a human-like civilization to a grabby aliens one, perhaps … is just incredibly small, and that explains why we don't see them.

That is, Kipping says (like others) that in our history there were no hard try-try steps, not even at the origin of life; intelligent life at our level was largely inevitable on a planet like ours, though it had to wait for the Sun to evolve to a matching state, which coincidentally took about half the Sun’s history. (A coincidence explained well by the hard steps model.) 

If our type of planet isn’t crazy rare, that should create a crazy huge number of places with alien intelligence among the ~10^24 stars in the observable universe. Kipping says we don’t see any because of a crazy small chance that intelligence like ours gives rise to a distantly-visible civilization. This seems to me very bad news about our future, though Kipping doesn’t seem bothered by it. 

I don’t understand Kipping’s issue with “falsifying” very fast aliens. We won’t directly test any hypothesis on alien speeds until we see them, at which point we’d confirm a fast speed as easily as a slow one. And we will know much sooner how fast we can go, first based on engineering analyses, and then on our own realized travel speeds. 

I agree with Kipping that it is modestly surprising that we find ourselves on a star as big as the sun; smaller ones are more numerous, though their fraction of potential planet metabolism is smaller. I also agree that planets at small stars face more obstacles, though I’m more optimistic than Kipping about ocean worlds there. 

My biggest issue with Kipping’s scenario is his crazy small estimated chance of intelligence like ours getting grabby. Seems to me we’ve already done rough engineering calculations which suggest that interstellar colonization looks possible, and even at high speeds. And the biological/economic pressures that would induce expansion, when its possible, have long seemed obvious. 

I could much more easily believe a version of “rare Earth”, either with very rare Earth context, or some crazy huge try-once steps in the early history of life. That is, life appears widely, but it almost always falls fast into a bad structure, from which it can’t escape.

Three corrections. First, Kipping highlights me but doesn’t mention my coauthors: Martin, McCarter, & Paulson. I couldn’t have done this without them. Second, Kipping says:

We'll encounter this wave of [grabby alien] colonization in about a billion year's time, and Hanson predicts a one in 10,000 chance that we will ourselves become such a grabby civilization before that time.

Actually, while we calculate observational implications for many different chance levels, we give no estimate ourselves. Finally: 

Hanson's model not only assumes hard steps, but also that the number of hard steps is a universal value on every planet and moon out there.

As step difficulties vary over such a wide range, if there are a wide range of possible hard steps paths, the most likely number of steps is probably far more likely than other numbers.

Added 9Sep: Avi Loeb also has a recent post criticizing grabby aliens! Odd that 3 astrophysicists all posted critiques the same month, 3 years after we published. Is there a campaign behind the scenes?

Loeb says that small stars might not be able to long host life. Which we agree is possible, though far from obvious; ocean worlds are very common and seem immune from most small start problems.

Loeb also complains:

Travel at relativistic speeds requires an unrealistic energy supply … and unrealistic shielding.

But we are talking about aliens millions to billions of years more advanced than us, we already seem to have promising designs for 0.15c speeds, and our high speed estimate is an inference from data, not an assumption as Loeb calls it.

Loeb errs here:

the authors argue that the ratio of quiet to loud alien civilizations must be over 10,000 for there to be even one alien civilization ever in our galaxy. Since our civilization is regarded as quiet for now, our chance of becoming grabby is the inverse of this ratio, namely less than 1/10,000.

No, if we accept a no-Milky-Way-aliens stat, then 10,000 becomes an upper bound, not a lower bound, on the ratio of quiet to loud aliens. Our future chances might be high.

Finally, Loeb ends with this dismissive dig:

Instead of performing theoretical gymnastics with unknown parameters, we better engage in an experimental search