Beyond “Fermi’s Paradox” XVI: What Is the “Dark Forest” Hypothesis?

We’re in the light cones of all the stars we can see, and as far as we know, we can’t get out of them. If a civilization ended even 100 years earlier than our equivalent “now” in their light cone, we wouldn’t have seen them. And if they became visible even 1 year later than our equivalent “now,” we would not have seen them yet.

We’re proceeding through time at 1 sec per sec and basically if we’re going to see an alien civilization at this point, I think the only way would be if one happens to achieve the necessary technology to be detected while we’re looking at it. If there were existing civilizations that were easy to see, we would have seen them already.

I think it’s far more likely we will confirm alien life first by indirect means, for example spectroscopically detecting free atmospheric oxygen on an exoplanet, or finding tiny fossils on Mars.

If such a technology is possible at all, only the very first advanced civilization had the ability to hide from all others. And only if they were able to develop it before any other civilization had been able to track them.

To some extent it might nevertheless be a reasonable strategy to keep as quiet as possible. But this strategy is less and less useful for the latecomers. If they were already tracked by a multitude of other advanced civilization they would hardly benefit from keeping quiet. Unless they (wrongly) think that they are an early advanced civilization.

However, if a couple of these civilizations start to openly seek contact to others, what can the hidden ones do? If a hidden civilization starts to fight one of these latecomers, it would need to leave its cover and make itself known to all other civilizations in its vicinity. If it follows the dark forst hypothesis, it could only do so, when it is sure that it is the only dangerous civilization in its forest.

This leaves me with the following alternatives:

- The forest is dark, because there is only one civilization out there that is very capable in hiding and has the ability to exterminate any latecomer efficiently and without traces.

- The forest appears dark, because it is thinly populated and we just have not looked enough for the others.

- We are the only ones in our cosmic vicinity.

> Overall, the Dark Forest Hypothesis has an internal logic and consistency that makes it an appealing (if somewhat somber) potential resolution to Fermi’s age-old question. Unfortunately, it also suffers from an inherent flaw that is capable of unraveling the whole thing. Like many other Fermi-related hypotheses, it only takes one exception to this rule to prove it wrong.

Given this, I thought it was going to follow on to point out that we are the exception to the rule, but instead it goes on to talk about malevolent exceptions. But, as the article mentioned earlier, we have made many active attempts to communicate our existence to other hypothetical civilisations, and we make no effort in obscuring our radio signals.

Is it possible that other civilisations follow the “Dark Forest” principle? Of course it is! But why would we happen to be the only civilisation that doesn't really worry about getting seen? The possibility that we are not alone and we are the only ones trying to communicate sounds even more fanciful and anthropocentric than any of the alternatives.

It's orders of magnitude of orders of magnitude less energy to communicate with (including obscuring your origin, if you want) than destroy a civilization. And communication has potential benefits (cooperation) that destruction does not.

Civilizations that communicated would very quickly and easily out-compete/out-advance civilizations that did not. "Malevolent" civilizations would inevitably run into more advanced cooperating civs sooner or later and be checked, either learning to cooperate themselves or lose.

Put another way, the assumptions used to build out the hypothesis, "Chains of Suspicion" and "Technological Explosion", don't make sense either. "Chains of Suspicion" assumes civilizations cannot communicate, yet the dark forest hypothesis assume civilizations can destroy each other. That's a contradiction. If you can physically interact, you can communicate (and as I pointed out before, with much, much less energy/resources than it takes to destroy). And "Technological Explosion", as written in the book, assume exponential technological growth could occur at any point, which seems to be nonsense to me. Any kind of exponential growth necessarily depends on a medium primed with the resources for that growth and stops when the resources are expended (which is never all that long, given the nature of exponential growth).

I cringed every time this idea was forced into the forefront of the books. They are such great and imaginative books, I don't mind at all that not all the ideas really pan out, but it was a little hard to stomach every time the plot turned on this weak idea.

There could be trillions of habitats out there and we'd never know it. From small low-orbit stations orbiting home planets to larger complexes orbiting their stars instead up to generation ships moving to the next star to mine for more resources or to build more ships due to population limits being hit. Once you've perfected the building of habitats that are large enough and are tailored to your species specific needs, why mess around with planets that can kill you in so many different ways?

Maybe the reason we don't see anyone is because we can't resolve even very large habitats at light-year distances and the reason we don't run into them is because we don't have anything they can't get somewhere that isn't already inhabited. Perhaps they avoid inhabited systems because of the dangers involved.

If, instead of doing a product of the probability of each event, you actually take the various distributions into account (nowadays you can easily test a wide variety of distributions with monte carlo methods) then, the probability of getting into contact with another life in the universe becomes vanishingly small.

I'm not sure this constraint is so strict to force a strong competition for survival, given how vast the universe is. It seems to me this projecting our earth-bound mentality of limited resources to a whole different scale, where it may not really apply.

1. You don't have enough easily accessible hydrocarbons in ground to build a civilization advanced enough to even get you into space.

2. You do have enough hydrocarbons which leads to inevitable overshoot, and accompanying collapse before you ever get close to figuring out interstellar travel. Exponential growth patterns lead you to either exhaust the carrying capacity of your planet, or you end up warming your planet too much and die off (or a little bit of both).

We're a pretty good case study between 1 and 2. Right now we're in a race to see if we can exhaust our hydrocarbons or overheat our planet first, there seems to be no realistic alternative (plenty of nice fantasy ones though).

If we had reached peak oil in the early 80s (or sooner) we would have likely avoided catastrophic climate change, but would have started a major population shift downwards towards 1 billion, where we would have likely stabilized but with no major technical progress (technological progress is largely a function of energy).

We didn't though, so now it seems like we are going to continue to increase the rate we combust hydrocarbons until we create an unlivable planet (at least for us). A few billionaires are making some cool toys, but we don't seem to be able to survive until we're anywhere near interstellar travel.

The dynamic that guarantees it is that great stuff is almost always cheaper/easier to experience via simulation than in reality. We already know that TV plots and video game situations are endlessly more entertaining than our real-world lives. Arranging computing power that can fool our eyes and ears is way easier than mustering the materials to get ourselves into space.

https://arxiv.org/abs/1806.02404

Most people when doing the Fermi estimate for the number of civilizations in the Galaxy use numbers that tend to result in a lot of them. But different people use very different numbers in different steps. If you look at the ranges of values that different people assign to the different terms then suddenly it doesn't seem very unlikely that intelligent civilizations could be rare.

https://www.youtube.com/watch?v=o9kbcGfX35M&ab_channel=Peter...

As of myself, I also have a theory why it's difficult for civilizations to see each other:

1 - You need very advanced technology to break through the barrier of speed of light (worm hole, someone?)

2 - Such advanced technology will need a long, continuous period of peace and prosperity for researchers to figure it out. Wars are "useful" in certain way but not always.

3 - Intelligent beings are in general greedy (regardless of their biological construction) and they want to take as much resources as possible.

4 - The resources in the universe is unlimited, but the resources that one civilization can obtain is very limited.

5 - Combine 3 & 4, you can probably figure out that the "long, continuous period of peace and prosperity" is very much unlikely and very likely to be broken by wars within the civilizations.

6 - The more advanced the civilizations they are, more likely those wars will wipe them out completely.

7 - Even if they managed to avoid those destructive wars (unlikely already). Another factor we need to consider is the number of smart scientists and technicians that can figure out and implement those fancy technologies.

8 - Yeah for sure, one man, such as Newton or Einstein can leap forward our understanding of the universe.

9 - But you need a very large of population to "grow" one Newton or Einstein.

10 - Go back to 3 & 4, now tell me, even if they manage to avoid wars, can they also avoid creating a caste and shut everyone else out of a prosperous life? I mean look at ourselves.

11 - If such thing happens (eventually), there is not going to be enough population to create a Newton or Einstein. And science/tech eventually just don't go forward.

12 - Combine 6 & 11, it is really very, very unlikely for any civilization to develop advanced technology to travel beyond the speed of light, which also means we don't get to see them (very often).

I personally find it impossibly unlikely that we are the only intelligent life. It's just, there's nothing in the universe that doesn't repeat. And there is a lot of space for us to repeat. Our solar system likely isn't all that special, so I just don't get that we could be the only thing that appeared in all the conceivable universe.

It also might be the case that we're not alone, but that life is just so hard to get going that maybe the nearest civilization is galaxies apart from us. In that sense I suppose we really are "alone" for any practical purpose.

It's pretty similar to one of my favourite quotes, from Peter Watts:

> Equidistant to the other two tribes sat the Historians. They didn't have too many thoughts on the probable prevalence of intelligent, spacefaring extraterrestrials— but if there are any, they said, they're not just going to be smart. They're going to be mean.

(I'm admittedly finding a lot of the other plot points of the series, especially around how people respond to adversity, less convincing though.)

If you think about how humans behave, I think if we ever get to the point where we're a multi-planetary species, those two planets aren't going to trust each other very much. If we expand to other star systems with years of communication lag, it'll be even worse -- both societies are going to be constantly expecting a hail of nukes to fall out of the sky one of these days, unless there's either a plausible defense against such an attack, or a reasonable expectation of mutually-assured destruction.

The same idea was in Gregory Benford's novel In the Ocean of Night in the 1970s.

Bracewell-von Neumann self-replicating interstellar probes are not impossible. I think we're not too far away from being able to create and launch them ourselves, if we survive.

Has any technical civilization in galactic history launched a BvN probe? It only takes one launch of a BvN probe to saturate the galaxy for all time afterwards, esp. if they can mutate and compete. But we don't see any such probe activity, and I think that we would, if they were busily mining asteroids and frantically manufacturing copies of themselves. So I think we can tentatively assume that nobody has launched one yet in our galaxy; and explaining that possibility leads to some unsettling consequences.

It only takes one xenophobic von-Nuemann capable civilization to ruin the party for everyone.

Honestly, it puts things into perspective when considering the terror of the 'dark forest' hypothesis. If intelligent life is anything like us, they are just as stuck on the prosaic (arguing amongst themselves about invasive ads) as they are on the profound (ending other civs).

I don't agree that Civilizations need to expand without limit. Even humans, having reached a certain level of comfort, have difficulty reproducing enough to maintain population levels. Conceivably then, civilizations will reach a comfortable size and then stop. Certainly they might be willing to stop when they reach a "boundary".

The problems with enemies is that, unless we are of sufficient size, we might bump into an enemy that is, say, twice our size, and therefore can thump us in a war. However, novels by Iain M. Banks [2] and Neal Asher [3] provide insight here (Culture vs Iridans, Polity vs Prador). The issue is not how many weapons you have (W), or even how many weapons you can make (dW/dt), but how fast you can make weapons factories (d2w/dt2) (etc). And ultimately, organics simply have absolutely no chance against AI: in both Culture and Polity, the AI runs the war for humans.

On the matter of technological breakout, the entire Lensman series has an underlying theme of the back-and-forth of new weapon vs it's mitigation [5]. Liu makes a good case that such back-and-forth is unrealistic: in the time it takes light to travel from the enemy's star to yours, they might go from poking each other with pointy sticks to fusion bombs, or from fusion bombs to star-killers. But there is a simple mitigation there too: don't hang around stars. Or rather, think of stars as Navies think of islands: nice place to refuel and R&R, but don't put all your ships in one harbor.

Ultimately, if survival is the goal, then Salvation or We are Bob [4] are the way to go: expand as much as possible. Just be willing to defer to existing civilizations.

[1] https://www.goodreads.com/book/show/34068552-salvation

[2] https://www.goodreads.com/author/show/5807106.Iain_M_Banks

[3] https://www.goodreads.com/author/show/56353.Neal_Asher

[4] https://www.amazon.com/Are-Legion-Bob-Bobiverse-Book-ebook/d...

[5] https://www.goodreads.com/series/49225-lensman

Examples: Blow a star up and make it project a unique territorial mark. Make a diamond lighttower, beaming threatening messages to probable green zones, to lure unsuspecting enemies in.

I absolutely can't see how that follows exactly. The Universe, hell, just our own galaxy, is pretty damn big. The time and fuel it could cost you to get to the adjacent civilization and steal its resources might be way too much compared to, you know, invest in some cosmic mining operations where you can get a Bonanza of resources just by poking into your local asteroid cluster.

To address the rest of this argument, a civilization might be as malicious as they come and they could still be very powerless and can't just go terrorize somebody.

If you assume that we live in an FTL universe subject to the laws of thermodynamics (as I do), mass and energy (being basically the same thing as we know) are the ultimate limiting factor. This basically means bigger is better. Remember such a civilization is likely thinking on time lines where a trillion years is a blink of an eye.

So how will a spacefaring civilization evolve? Likely into a Dyson swarm for three primary reasons:

1. It requires no new physics. It's essentially an engineering problem that only requires materials as complex as stainless steel and energy tech no more complex than solar power. If you get better tech (eg fusion, graphene for construction) this outcome becomes more likely not less;

2. Orbitals are a highly efficient way of creating living area. A full Dyson swarm around our Sun would likely consume <1% of the mass of Mercury, for example; and

3. The energy output of the Sun is essentially "free".

A full Dyson swarm makes you a Kardashev-2 (K2) civilization. On the Carl Sagan scale, that means you're consuming ~10^26 Watts of power. And that's just for one Sun. Now consider:

1. Such a civilization is capable of sterilizing the galaxy in 100,000 years if they choose to;

2. There's really no hiding from such a civilization;

3. There's really no hiding a Dyson swarm from a Dyson swarm from an advanced civilization due to the IR signature it would produce. This has nothing to do with say radio communications.

Remember with the Fermi Paradox you don't have to establish what the average civilization does. Say there are 5 spacefaring civilizations in the Milky Way. What are the odds that all of them hide? Now imagine there are 100. The odds of them all hiding are much lower. And it only really takes one who decides to build Dyson Swarms and expand beyond their star to be detectable.

So I can imagine a splinter of a civilization decides to "hide". A good example of this would be if fusion becomes viable. One could sink habitats into Neptune, which is not particularly dense, and probably live there undisturbed and unfound for eons is quite high. That's probably more hidden than, say, trying to hide between stars.

It's just not likely that as the number of civilizations increases that all of them go that route.

My view is that spacefaring life is just quite rare. We may well be the only one within out light cone within the Milky Way. Why that is is an interesting topic.

[1]: https://www.youtube.com/watch?v=tEBn8bc0k-I

TLDR: The forest analogy fails because civilizations are not singular entities (like animals are) but rather complex structures which can merge and evolve. The supposed technological and societal barriers can very well not materialize in which case the theory falls apart.

Details below...

> Suppose a vast number of civilizations distributed throughout the universe, on the order of the number of observable stars. Lots and lots of them.

"On the order of observable stars" is extremely wide. Perhaps there is one life-bearing planet per galaxy. Perhaps there is one per 100 stars. Those are vastly different outcomes.

> Survival is the primary need of civilizations.

On individual level humans have no hope to "survive" (i.e. live immortal lives). Yet despite that fact we don't commit mass suicide. The "primary need" of human civilization to survive is merely emergent behaviour due to other factors. Perhaps advanced civilizations would be happy to achieve some grand undertaking and then perish.

Now for the second "axiom":

> Civilizations continuously grow and expand ...

Expansionist civilizations with unlimited space certainly can do that. But does it really bring any benefits? The more complex the societal structure, the harder it is to change. Perhaps there is a critical mass above which it becomes detrimental to multiply beyond and all "sane" civilizations will stay below it.

> ... , but the total matter in the universe remains constant.

It may be finite, but perhaps most of it will be "stranded" anyway - so far away that access to it becomes non-feasible anyway.

> ... In other words, the finite nature of resources will ultimately pit one civilization against another as they all struggle to sustain their growth.

This part I believe is the biggest logical fallacy of all. What is civilization if not a set of societal structures? And we can easily imagine merging of two such structures - they are already composed of other structures; what is just another level of complexity? If we merge two "competing" civilizations suddenly we have a single larger civilization... supposedly without conflict? The amount of resources is the same as before but not we only have a single contender: the merged civilization.

This paradox stems from the failure to account for the internal race for resources inherent to all civilizations. The civilizations are not unitary entities. Treating them as such and assuming blind dedication to it from its members is short sighted and lacking imagination.

> If open communication exists, then conflict could be averted and tensions diffused. But given the time-lag involved with any attempt at interstellar communications, civilizations that are light-years apart are unlikely to resolve their fears quickly.

Sufficiently advanced civs can simply send AI envoys or do mind upload of their representatives with the speed of light. This makes the learning of language, communication etc. very quick. Perhaps FTL communication can exist too. Etc. etc. Endless ways to break another assumption.

> By the time they arrived, the attacker’s technology would not have matured one bit, while the defender would have decades or centuries to progress.

It may be possible that advanced technology does not progress as fast as it does now. See our experiments with fusion and particle accelerators. What used to be easy now takes years and billions in funding. Perhaps going at some point getting any kind of advancement will be nearly impossible and all technology will become "old". In this scenario the centuries of technological advancement are moot. OTOH potential gains from trade from foreign civs is very real possibility.

He can't make such a statement about probability when the circumstances aren't even known.

We don't even know how life forms so how can we even know what the probability of it forming is?

How do we know that probability is high enough that it is very likely to occur? The chances of it occurring could be incredibly low.