Welcome to a short in-between episodes episode.
The reason I’m releasing this now is there was a draft paper released recently, titled
“The Fermi Paradox and the Aurora Effect: Exo-Civilization Settlement, Expansion And
Steady States” and it started getting discussed and I had several requests to do a video explaining
it, so I thought we’d go through it and then get into some of the pros and cons for
Let’s start with the title.
If you’re watching this episode odds are you already know what the Fermi Paradox is,
which is the apparent contradiction between how empty of life the Universe appears to
be, and just how huge, ancient, and filled with potentially habitable planets it is,
or basically, where are all the alien civilizations?
But the Aurora Effect is probably unfamiliar unless you’ve read Aurora by Kim Stanley
Robinson, author of the Mars Trilogy.
We’ll use the definition from the paper, that “Good worlds are hard to find”, which
is to say, just because you might find a planet you could settle and terraform, doesn’t
mean it’s really one you’d invest the effort into, or for that matter be able to
transplant a civilization there that could successfully grow.
History is full of settlements that failed or stalled out after all.
Let me start by saying that I have issues with the theory, which I’ll expand on once
I’ve laid out what that theory is.
So, let’s first go through that paper in a more relatable way so we understand what
it’s all about.
If you assume that most systems don’t have a place anyone wants to live or will thrive
at, it really puts a dent in the classic expansion model for the galaxy.
That assumes you colonize in a first wave from Earth, or our solar system anyway, and
then those colonies colonize those near them eventually, who do the same.
This is a very common approach for modeling galactic settlement timelines and we often
refer to it as percolation because you can then model it like a fluid going through some
How many worlds are viable for settlement and how quickly those settlements grow to
being willing and able to dispatch a daughter colony of their own, rinse and repeat.
Models based on this are very sensitive to a number of variables, and can thus produce
a wide range of galactic colonization timelines.
For instance, if our colonists can and want to settle every star system, can travel at
1% of light speed, and we assume the typical colony begins with 10,000 people, doubles
in numbers every 50 years like we did in the 20th century, and will colonize a new system
when they get to 10 billion people, they need to double 20 times, to go from 10,000 to 10
billion, and it will take them just a millennium after settlement to get to colonizing their
own new systems.
Just to keep the math easy, if they only travel at 1% light speed and only settle within 10
light years, it takes them another millennium to get a colony ship to all those systems
in range, so basically your colonization wave percolates out at 10 light years per every
It means you colonize at 0.5% of light speed, you travel at 1% but spend half your time
stopped and growing in numbers.
Galactic Colonization, everything within 100,000 light years, would take 20 million years.
That sounds like a long time, but to put that in perspective, a single rotation of our galaxy
takes roughly a quarter of a billion years.
So, this 20 million-year timeline means not much in our galaxy has changed, At the local
level, a score of centuries won’t rearrange the locations of neighboring stars much either,
so we treat the galaxy as basically static.
In other words, the stellar motion can be ignored for colonization.
However, you can play with those numbers a lot, or even outright discard the model when
using something like our Gardener Ship approach.
If your ships can move at a modest fraction of light speed, say 10%, then most of the
percolation time is growing your numbers on each new planet, and you can advance from
each one a lot further for each step, since they can hurl new colonial ships much faster
& further each time.
You might make steps of 100 light years every two millennia, colonize at 5% light speed,
and settle the whole galaxy in just 2 million years.
You are also way less likely to stall out on any given front because you’ve got such
a large range encompassing so many settled worlds that if some don’t grow quick and
send out colonies, many others are plugging that gap, like broadcasting lawn seeds.
On the flip side of this, if you have to go slower, or if many worlds just aren’t desirable
for colonization or if they grow much slower or tend to lose a desire for colonization
and don’t send out ships once they can, then you percolate far slower and might even
stall out completely.
If we were assuming it took 100,000 years to grow enough to make each new jump and it
was 10 light years a jump, you’re creeping outward from Earth at just a percent of a
percent of light speed and will need a billion years to colonize the galaxy, during which
time you can basically discard the notion that humanity is even doing this, as your
civilizations are diverging massively at each step so that even your nearest neighbors or
mother world are basically different species and you might argue about who colonizes what,
since there’s not likely to be much team spirit even in your local expansion area.
This is even more hindered if you can’t expect to find a suitable candidate in 10
light years of you.
If you can only make a successful colony around maybe 1 in 100 stars similar enough to our
own, and only maybe 1 in 100 of those has a planet that’s a desirable candidate for
terraforming, you’re only going to have a couple candidates within a 100 light years
of you, not the many thousands we normally assume when discussing this topic here on
If your ships aren’t relativistic, and you’re focused on terraforming planets, that’s
a real long voyage to ask folks to make for the honor of living on desolate wastelands
for the many millennia it will take to terraform them properly.
And it is the basic reasoning behind the classic scifi space opera of galactic empires with
only a few million inhabited worlds.
That colonization process is very slow and very hard and very disorganized and can peter
out as colonies fail or just barely prosper and aren’t interested in repeating the Herculean
task of colonizing again themselves.
It also implies you might have dead colonies, ones that just failed and that went extinct,
and timelines long enough that even if an older alien one did so on a world we came
across, we might not even know they had.
That’s also a popular notion in scifi, that great civilizations arise and eventually end,
leaving little or nothing behind in their wake when a new civilization evolves and sends
out its sons and daughters to do as they did before us.
On such timelines you really can’t treat the galaxy as a static place either, not only
might a civilization that once looked like a vaguely spherical blob have smeared itself
across the galaxy as all the stars moved around, but they might actually be taking advantage
of local stellar motion for colonization.
If you are going the slowboat route of colonization, taking many centuries to prepare for missions
that might move not much faster than our modern space probes, you might bypass nearer stars
in favor of ones moving toward you.
So the paper lists a number of parameters and assigns various values to them, those
values we need not consider, they weren’t chosen out of a hat, but still are fairly
arbitrary, even if reasonable guesses.
And like Drake’s equation when you’ve got a lot of parameters, many of which you
can’t nail down to even an order of magnitude, you can’t draw many concrete conclusions
We’re not going to walk through their math, though it’s an interesting model.
Some of those parameters are the fraction of systems that are settleable, the density
of systems, which is not uniform, some regions of space will be near wastelands while others
will be much closer and richer in worlds than our own region.
We’ve also got probe or ship range, and velocity, as those both are major factors
in not just how far and fast you can go, but how many candidate worlds you have access
to and if you’re even willing to do it.
Others include probe or ship build and launch times, lifetime of the settlement, average
stellar motion if you’re going slow enough you need to factor that in, and many more.
This includes the notion that if settlements are dying off, folks would have to resettle
them to continue expanding.
Now this gets back to the Fermi Paradox more directly by reminding us of Michael Hart’s
Fact A in his original conjecture about the Fermi Paradox, which is sometimes called the
Fermi-Hart Paradox because Hart is basically the guy who began the discussion of the Fermi
“Fact A” is that there are no aliens currently on Earth, obviously not everyone agrees with
that but it’s the basis of discussion for the Paradox, since there’s obviously no
Paradox about where all the aliens are if the answer is out in a farm field crushing
corn stalks to make modern art.
I also usually don’t like calling that a fact, as a lack of evidence to me is more
of an assumption than a fact.
I’ve never seen leprechauns, don’t believe they exist, but I’m hesitant to say “Fact:
Leprechauns are not on Earth”.
Whichever, it’s the basic assumption of the Fermi Paradox along with the idea that
while Earth and it’s life might be fairly special, it’s probably not so special we
wouldn’t expect lots of other places where life can pop up and get a foothold and potentially
be spawning other civilizations that might be interested in colonizing new worlds too.
But there’s a second half of that, because while it says there are no aliens currently
on Earth, it also implies none in the past too.
By fossils and genetics, we can rule out us being an alien colony that got abandoned,
unless it either happened a few billion years ago, since we can see a clear relation to
other life on Earth and a common origin, or that they took more of the bioforming route
and basically adapted themselves into our existing world very subtly, which is possible
but tends to involve a lot of weird choices and handwaves to make much sense without ramming
into the fossil record.
That brings us to an outright failed colony, one that just didn’t take off and was being
done to minimize disruption to the local and current life, and could maybe escape notice
in the fossil record.
Just to throw one out there for daydreaming, if some aliens came here 65 million years
ago and settled near Yucatan, they might have domed things over at their base to avoid too
much back and forth ecological leakage while they were setting up and since they and their
own life would probably not match well for existing conditions.
That’s something we’d do, even if we meant to totally replace the ecosystem already on
a world, we’d want to study it in detail first and keep ourselves isolated.
If after a few centuries things just weren’t working out, we weren’t growing much, and
many of those born since were fond of the existing ecosystem and didn’t want it destroyed,
I could see them abandoning the colony or setting the reactor to blow up and vaporize
all the extraterrestrial life, leaving behind a big crater.
This would effectively erase most of the evidence they were here, and time would do a good job
removing the rest.
This is an example of how you can decouple Hart’s Fact A, no aliens here now, from
no aliens here ever.
So the paper argues there’s an effective temporal horizon that would obscure our ability
to see previous visitors or settlements.
See the Cyclic Apocalypse episode though for why this normally won’t tend to work for
expecting time to erase signs of prior civilizations, or short form, bones aren’t the only things
that fossilize or otherwise leave a long-term remnant, and civilizations would tend to leave
a ton of mysterious right angles in their geological records.
The caveat here is you must have been a fairly large civilization to leave enough of those
kinds of footprints that we’d be likely to have some survive and numerous enough we’d
So this is the basic notion: civilizations arise probably not too often, don’t find
too much to colonize, take a very long time to do it, and often have local pockets grind
to a halt or die off.
When new civilizations arise, the time between these civilizations and their smaller scope,
means you could easily not even see them if they’d had a colony next door at some point
or maybe even on your own planet.
We’ll mostly bypass the paper’s discussion of stellar motions because I think that overdoes
things, if you are already assuming colonization so slow that each colonization front is incorporating
the motions of stars and the galaxy not being static into their plans, I’d think you could
just assume they will throw their hands up in the air and not colonize just because of
the time needed.
Whether you’re doing it by slowboat generation ships or probes that arrive and unpack and
replicate and build a civilization, time remains a hurdle that many are not going to want to
try jumping over, especially if that’s such a long time that whatever is doing the colonization
isn’t going to even vaguely resemble what their mother system has mutated into or the
original colonists who left.
Hard to get the dinosaurs to colonize the galaxy if they know it will be humans who
actually arrive, essentially.
All right, that brings us to the question: what’s the flaw in the paper’s reasoning?
Channel regulars are probably raising an eyebrow with me at that “Good worlds are hard to
find” aspect, asking why you even care about finding good worlds since we normally wouldn’t
even expect spacefaring civilizations to much care about planets, to quote my friend Fraser
Cain, “Gravity wells are for suckers” you’ve just gone to all that effort to escape
your own planet’s gravity well, why would you go set up shop on another one?
So, when we talk about colonizing other star systems, we usually emphasize that you start
not by landing on a planet there, but by linking up to various resource-rich low gravity objects
like Asteroids and Moons, build rotating habitats there if you want or need gravity, and settle
planets if you want to, but only after you’ve setup your basic space based industry and
Planets may or may not have their value and might end up very prized by civilizations
as potential centers, but in their own right are no bottleneck to expansion.
If we assume that approach, it pretty much ends the theory under discussion today right
Interstellar colonization is mostly done by civilizations that back home already mostly
don’t live on planets, as they build and breed their way up to K2 civilization status,
and while Earth-like planets might be highly valued, or not, most folks live in artificial
They’ve lived on one for the long trip to the new system, and ought to be happy enough
in one at the destination or even prefer those over spending centuries to terraform a planet.
Such being the case, virtually every star and planet, habitable or not, becomes prime
real estate, because to you they are really just an electric outlet and building supply
But even if those systems are more valued, as they may well be, you will tend to backfill
as you go, same as settlers always claim the best spots first and fill in over time.
So, let’s get into the meat of my objection.
This concept is very dependent on the notion that colonists are likely to be either limited
or picky about what systems they settle and might tend to fail even then.
That’s a perfectly reasonable assumption inside the usual space opera context of just
settling Earth-like planets, but even ignoring the notion that we’re likely to build most
of the places we live in, not terraform them, we aren’t going to make our first colonized
world around another star.
We’d try for Mars or Venus first, and either we find out that’s not practical, and either
stay at home or go the artificial habitats route, or for a more cybernetic or digital
existence, or we’d have gotten practice making places like Venus and Mars habitable,
in which case we already know how to do it and if we can do them, then “Earth-like”
gets very unimportant.
Worlds of similar mass and temperature to Earth might not be too common, but if we widen
the range to include wretched miserable rocks like Mars and Venus, then odds are you won’t
ever have to look far for a star that’s got a candidate, and you also have a couple
planets whose inhabitants won’t find them unappealing and can supply colonists.
Even if you expand the habitable world definition to be safe, amenable star systems, without
exposure to dangerous stellar events like supernovae, pulsars and quasars, the time
horizons and backfilling of systems from continued growth will just leave gaps in your galaxy’s
colonization, not roadblocks.
The second flaw is that it probably won’t be a process of planet-hopping anyway.
Even if planets are preferable, especially to folks who live on them, a future population
of this solar system is likely mostly to be people who weren’t born on a planet, and
live inside a great big cylinder, or on a computer chip, and neither of those cares
much about either travel times or the destination’s planets, just so long as the destination has
free matter and energy to exploit.
The former already live on a spaceship, since a rotating habitat is a spaceship, and the
latter can literally just send copies of themselves.
Remember, we all suffer from confirmation bias, the tendency to search for, interpret,
favor, and recall information in a way that confirms one's preexisting beliefs or hypotheses.
We all currently live on a planet, so our bias is to naturally think that planets are
good things to live on and in the future, our descendants will continue with that trend.
However, that’s dangerous thinking because if we asked a farmer a couple of centuries
back whether the majority of the world’s population would live in cities in the future,
they would have laughed at us.
They would have rightly argued that the majority of folks live as farmers and if everyone moved
to towns or cities, there wouldn’t be enough farmers to feed all the city dwellers.
Now, we know from our history that the farmer was ultimately wrong.
Mechanization and technology revolutionized farming and today, the majority of folks now
live in urban environments.
We have to be careful not to apply the same confirmation bias to the notion that humanity
will always want to live on planets too.
By the time we’re ready to aim for galactic colonization, we’ll be a K2 civilization
or moving toward that.
Our solar system will be peppered with thousands or millions of space habitats, possibly far
more, and the majority of humanity will not have set foot on a planet let alone lived
They won’t be wedded to the notion that planets need to be colonized any more than
we are wedded to the notion that most of us need to be farmers.
Another objection is just this general notion that civilizations can go extinct — they
They can collapse and we’ve tons of examples, but they don’t just die off without external
pressures, usually a rival civilization eroding them, even more so when they are high-tech
and thus resilient to things like natural disasters.
Collapse is not synonymous with extinction then, it’s just an alteration, usually to
something tougher and stronger, except for when discussed by a historian who had a crush
on the civilization under discussion and dislikes the barbarians who ended them.
I don’t usually like to broaden the term evolution to include social aspects, but it’s
valid on longer timelines and particularly in this case.
Applying the survival of the fittest litmus tests to a large sample size, you should generally
expect civilizations over time to get better, not worse, at surviving.
You can argue that evolution, which generally breeds to make things tougher and more adapted
to their area, does not apply to high-tech civilizations, but it basically doesn’t
because through forethought, planning, and strategy we can do better than random odds.
Which again kind of negates the notion that such civilizations will dwindle on new worlds,
when they’re drawing on centuries of planning and prior experience and know what challenges
And even if we accept the idea that a combination of slow colonization, a low fraction of settled
worlds of stars available, and modest long-term success rate for colonies would create “silent
bubbles” in our galaxy, that won’t necessarily solve the Fermi Paradox.
It will strongly depend on the visibility of civilizations; even if Earth is in the
middle of a thousand-light-year bubble of space that’s not settled yet, or not anymore,
the closest Dyson swarm or space beacon may still be visible.
Especially as the folks living there know they are not alone and not hidden, since their
launch would have been a big and recorded event, thus have every reason to be heard
and none to hide, and we see none.
Or short form, it’s a good theory for the more classic notions of planet colonization
but like a lot of Fermi Paradox approaches for intelligent life being reasonably common,
it’s leaning too heavily on classic notions of settling new places.
What we’d tend to expect is our own system boasting many trillions of people living in
space habitats that aren’t too interested in living on or colonizing planets.
If you’re curious about some of the solutions we’ve discussed for that, check out our
Generations Ships Series and Outward Bound Series, and if you want to know more about
other solutions for the Fermi Paradox or the problems with them, try our Fermi Paradox
And of course all of those, along with this paper, focus on the idea that while our technology
will improve, it probably won’t get a lot of the neater advantages often seen in science
fiction that tend to violate the speed of light or the laws of thermodynamics.
This upcoming Thursday we’ll relax about that a bit and contemplate some of those more
hypothetical technologies so advanced they are indistinguishable from magic, and what
their implications would be for the civilizations that wield them, in Clarketech.
For alerts when that and other episodes come out, make sure to the subscribe to the channel.
And if you enjoyed this episode, hit the like button and share it with others.
You can also support the channel on Patreon.
Until next time, thanks for watching, and we’ll see you Thursday!