In 1974 we sent the Arecibo radio beamed a message  towards the few hundred thousand stars of Messier   13, a globular cluster near the edge of the Milky  Way,.

It’ll take 10s of thousands of years to   arrive, so no surprise we haven’t heard back yet.

But if there are other civilizations out there,   surely we expect many of them to have a head  start on us—and so could have been shouting   into the void louder and longer than we have.

So why haven’t we spotted their own efforts to   make contact?

What if the silence from the stars  is a hint that we shouldn’t be so outgoing?

What   if aliens are deliberately keeping quiet  for fear that they might be destroyed?

For over 60 years we have been listening for  messages from aliens.

Yet we haven’t heard,   or seen, anything or anyone.

This is dubbed  the Great Silence by author David Brin.

With   over five thousand exoplanet discoveries confirmed  to date, it seems increasingly likely that there   has to be life out there.

The mismatch between  the number of possible origins of life and the   apparent absence of any other technological  life is known as the Fermi Paradox, which of   course we’ve discussed once or twice in the past.

But today we’re going to explore one of the more   terrifying solutions to the Fermi Paradox - the  idea that there are other technological species in   our galaxy, but they’re all silent - because those  that broke their silence were quickly destroyed.

This is the dark forest hypothesis, and  it’s the core idea of Liu Cixin’s book of   the same name, the sequel to the Three  Body Problem.

Liu Cixin coined the term,   although the idea has been proposed by others.

But the imagery of the universe as a dark  forest is so chilling, let’s lean into it.

Imagine you are a lone hunter stalking through  dense forest.

Visibility range is almost zero   so you have no idea what’s hiding over the  rise or behind the next tree.

You suspect   there may be other hunters but you’re not  sure.

You carry a deadly weapon and know you   could destroy other hunters in an instant … but  only if you spot them first.

You are forced to   assume that they have the same capability,  which means they could also destroy you.

So what do you do?

Do you call out and  hope to find allies?

Friends?

Or do you   keep very still and quiet and hope no  one finds you?

Or do you keep stalking,   ready to destroy any other hunter you find out of  fear that they’ve also chosen this last option?

The dark forest solution to the Fermi Paradox  proposes that almost all civilizations   will choose silence—whether a  passive silence or a watchful,   trigger-happy silence.

And those that  do not so choose are no longer with us.

It’s certainly a great creepy premise for a scifi  novel, but the idea is really interesting because   it’s based on some reasonably concrete game  theory.

And also some pretty solid physics.

Let’s look at the scenario again without the  forest metaphor.

We have two planets that have   developed advanced civilisations.

Let’s call them  the Alicians and Bobarians, or A and B for short.

Both A and B are capable of sending messages  across the universe and could, if required,   destroy another planet relatively easily and  without personal risk.

We’ll come back to how.

We can model their interactions in game theory  as a sequential game, just like chess—in which   ”players” choose their actions in sequence.

Each of the civilisations act in response to   the other’s last action, leading to a tree of  possibilities and some payoff associated with   each branch.

Each civilization should choose its  responses to maximise the payoff likelihood.

For   example, if the civilizations make friends that  would have a positive payoff for both—they gain   strength in numbers and maybe share technology.

If  one civilization destroys the other that would be   a relatively neutral payoff for the destroyer, but  an extremely negative payoff for the destroyee.

How negative?

Well nothing could be worse, so  maybe infinitely negative payoff or infinite cost.

So let’s look at one possible scenario.

Civilization B intercepts a signal from A,   but A doesn't know about B.

Maybe  the Alicians are actively sending   out signals hoping for a response.

Or  maybe their own radio communications   are just really loud—the equivalent  of walking noisily through the forest.

So when B notices A, what options do they have?

They could ignore A, reply to A, or destroy A.

Instead of Text see the following  part of the Matt decision tree Let’s say they ignore.

A will remain  blissfully oblivious of the existence of B,   and B will continue its antisocial ways.

This scenario has a zero payoff for both   A and B.

Slightly boring, but  at least nobody gets hurt.

Add an insert of this part of the  decision tree in the above section What if B takes the second option and  destroys A?

There’ll be a finite cost   to destroying a planet in resources—but  for an advanced civilization that’s   pretty small.

And there might also be a  resource gain if that solar system had any   good loot.

Maybe there’s a cost in guilt at  annihilating billions of sentient creatures,   but the point is that if there’s a cost  it’s finite.

For the Bobarians anyway.

Add an insert of this part of the  decision tree in the above section At worst, B experiences a finite negative  payoff by taking the destroy option,   while A experiences infinite negative payoff.

Finally, if B chooses to reply to A’s message,  A now knows about the existence of B.

This   gives additional knowledge to A that they  previously did not have.

Then, A has the   same three choices that B had—including  the potential to attack and destroy B.

SEE THIS PART OF DECISION TREE FOR ABOVE pragrah So of all the options—ignore, destroy,  reply—only reply leads to a potential   infinite negative payoff for B.

If you just sum  the potential payoffs of the three branches,   it’s clear that B should never reply—assuming  the Bobarians care about their own survival.

SEE THIS PART OF DECISION TREE FOR  ABOVE paragraph and highlight “reply”,   “turn A” “detroy” and the “cost  to “B (infinity symbol)” boxes In fact, if we extend the tree a little  more we uncover new outcomes—for example,   in the “ignore” branch there’s  a possible future branching in   which A discovers the existence of B  and gets its turn to play the game.

That means both ignore and reply have the  potential for infinite cost … and by that   logic a civilization should always choose to  destroy any other civilization they detect.

Highlight the chains or reasoning.

Red paths all  lead to B death and green path ensures B survival So, if we assume that civilizations that  can detect other civilizations also know   game theory, they know the game-theoretic  dominant strategy—destroy the other.

Or,   if they place very heavy weight on the guilt  cost, at the very least they will remain   very very quiet.

Hence the Great Silence  and the solution to the Fermi Paradox.

This is a pretty dark conclusion, so let’s  see if we can find flaws in this reasoning   in the hope that the universe isn’t as  terrifying as the dark forest proposes.

First up, let’s look at some of the assumptions.

We said that these civilizations can destroy each   other relatively easily.

That’s actually a  reasonable assumption if both civilizations   have concentrated population centers—say,  on one or a handful of planets.

AT the risk   of telling you how to destroy planets—Perhaps  the simplest approach is the relativistic kill   vehicle.

An advanced civilization worth  its salt should be able to harness, say,   a percent of its home star’s energy towards  accelerating one or more masses to a good   fraction of the speed of light.

Send those bodies  to the offendin world and you at the very least   ionize their atmosphere and vaporize their oceans.

Moreover, the target planet really has no way to   protect itself.

The vehicle is traveling at a good  fraction of the speed of light, so the target will   have little warning before the weapon arrives,  and not much they could do about it anyway.

It doesn’t even matter if the Bobarians don’t  think the Alicians have the technology to do this.

Technological progress can be exponential,  and so enormous advances can happen on the   timescale of the light travel time between two  star systems.

Say those systems are 100 light   years apart.

In 100 years, Moores law yields a  quadrillion-fold increase in computing power.

Other technologies have also been shown  to obey exponential improvement rates,   albeit with different doubling times.

Even if  that rate is more staggered, one thing is clear:   whatever technological state you perceive  your neighbours to be in is probably not the   technological state they’ll be in when they  receive your reply, or when they find you.

The vast distances between the stars  is really the driver of the dark forest   hypothesis.

It ensures that both sides have  potential access to the instant kill option,   and in general it’s what allows us to approximate  this as a sequential game.

Civilizations can’t   interact with each other in real time—they  can’t “feel out” the intentions of the other,   or react to perceived hostility.

They can only choose their move,   knowing that if the other side chooses hostility  in response it likely means complete annihilation.

Those vast distances also mean you’re not safe  even if you send a friendly message—how does the   receiver of the message know that you’re truthful,  or that you’ll stay friendly over the centuries?

So the physics behind the hypothesis seems to  hold together.

But there are also assumptions—for   example there’s some psychology.

We assume that  all aliens will ascribe an overwhelmingly large   cost or negative payoff to the extinction  outcome.

Seems fair—the “better them than   us” philosophy seems mostly universal among  humans.

And if all species crawled out of the   mud via Darwinian processes then they should all  have competitive and self-preserving tendencies.

But it’s also possible that we’re projecting our  own primitive psychological tendencies unfairly.

We don’t know what value systems advanced aliens  might have.

Perhaps there’s a psychological   transition that essentially all civilizations go  through in which they, like, realise the value of   all sentience of something.

Perhaps the ones that  don’t transition exterminate themselves.

In that   case they may not value their own existence as  infinitely higher than that of their neighbours.

There’s also another aspect to the cost  analysis besides fear and empathy—and   that’s curiosity.

Humans did not spread  across the globe or invent fire or build   great civilizations or discover the laws of  physics by each of us placing infinite value   in our personal survival.

Yes we sought the  resources and strength needed for survival,   but we were also curious.

And that curiosity  proved a huge survival advantage in the end.

It’s that interplay—that balance between  the desire to be safe and the curiosity   to peer over the next horizon, or round the  next tree—that kept us walking through the   metaphorical forest to reach the metaphorical …  I dunno, sun-dappled meadow of the modern world.

Anyone we meet out there will  also have curiosity and wariness   in different measures.

Some will want  to know what other life and cultures   and minds are like.

And maybe that  curiosity will outweigh their fear.

This is wild speculation.

The point is,  we can’t really assume that the payoff   calculation will be the same for everyone.

Others may be playing different games.

So, how well are we playing the game?

So  far, we’ve sent out a few messages here   and there.

The Arecibo message will reach  the M13 globular cluster in something like   27,000 years.

We’ve sent a motley array of  signals to 30 or so stars that’ll arrive over   the next few decades to few centuries (some  have already arrived).

It’ll take a pretty   advanced civilization actively looking for  signals to see any of them.

Some signals   have reached their destinations and there’s been  time for a response—for example the invasion fleet   from Altair was due in 2015.

Most likely there  aren’t advanced civilizations quite that close,   and we haven’t been shouting too loudly.

We’re  walking pretty quietly through the forest, maybe   breaking a twig or two.

We haven’t really started  playing the game yet—at least as far as we know.

So should we try to contact aliens?

Well maybe  it’s wise to be a bit cautious.

On the other hand,   there’s one last assumption in the Dark  Forest hypothesis—and that's the idea   that any civilization gets to make  unilateral decisions that follow the   perfect logic of game theory.

“We” don’t  decide to things.

Individuals decide,   and collective action emerges in very  complex ways.

Even now, individual humans,   or at least smallish groups, could start projects  that make quite a lot of noise on a fairly short   timescale.

And maybe we will.

And maybe when we  meet our first non-human civilization we’ll wish   we’d respected the Great Silence.

Or maybe the  curiosity and empathy of the individuals from   both civilizations will win the day.

Perhaps  we’ll continue through the dark forest side   by side and stronger for it, in hope of finding  brighter metaphorical landscapes of space time.