What does Robin Hanson think is the most likely “Great Filter”?

Here is a long and excellent post, whereby Robin outs himself as a strange kind of environmentalist.  Do need the whole thing, but here is one summary excerpt:

So, bottom line, the future great filter scenario that most concerns me is one where our solar-system-bound descendants have killed most of nature, can’t yet colonize other stars, are general predators and prey of each other, and have fallen into a short-term-predatory-focus equilibrium where predators can easily see and travel to most all prey. Yes there are about a hundred billion comets way out there circling the sun, but even that seems a small enough number for predators to careful map and track all of them.

“At first they came for the rabbits…and then they came for me.”  I find that intriguing, but I have a more marginalist approach, and perhaps one which encompasses Robin’s hypothesis as a special case.  The death of human (and other) civilizations may be a bit like the death of the human body through old age, namely a whole bunch of things go wrong at once.  If there were a single key problem, it would be easier to find a patch and prolong things for just a bit more.  But if we have reason to believe that, eventually, many things will go wrong at once…such a concatenation of problems is more likely to defeat us.  So my nomination for The Great Filter, in a nutshell, is “everything going wrong at once.”  The simplest underlying model here is that a) problems accumulate, b) resources can be directed to help solve problems, and c) sometimes problems accumulate more rapidly than they can be solved.

This is also why, in many cases, there is no simple “fact of the matter” answer as to why various mighty empires fell in the past.  Here is my earlier review of Apocalypto, a remarkable and still underrated movie.


"So my nomination for The Great Filter, in a nutshell, is “everything going wrong at once.” The simplest underlying model here is that a) problems accumulate, b) resources can be directed to help solve problems, and c) sometimes problems accumulate more rapidly than they can be solved."

Also known as the "Fire Upon the Deep" theory.

I think I prefer the Zones of Thought explanatio from those books...

It is about as realistic as treating "TheGreat Filter" and the Fermi Paradox as facts.

This idea was in A Deepness in the Sky (the prequel to Fire); a civilization was in crisis because their optimizations had reached their bounds.

There is also the possibility that the filter is in the past.

I think there are two fairly good reasons why this possibility is of less interest.

1) It's less relevant to what we do now. For example, if the griefer hypothesis is true, we should be shutting down our external transmissions with haste, and indeed it may already be too late for us.

2) There might be great filters both ahead and behind us. So a great filter behind us doesn't disprove a great filter ahead of us, although it does reduce its likelihood and rule out some hypotheses for what it could be. I personally don't think it's at all inevitable that we will ever progress to be even a Type 2 civilization, let alone Type 3 etc.

'we have met the enemy' etc. if you're spitballing long-term problems and solutions, perhaps we should focus on changing humanity's baser instincts rather than just planning on hiding out while everyone tears the solar system apart. i'd put money on some kind of end-model iphone implant hive mind, but that just spirals into authoritarian dystopia. these discussions are always depressing.

Agreed Apocalypto is a vastly underrated movie.

Critics didn't seem to grasp the theology of it or the fact that its commentary on the War on Terror.

The Mayans, like Americans, were believers in their own exceptionalism. However, they became gripped by fear during the corn blight. The superior Mayan troops pursue the protagonist Jaguar Paw into the jungle where they become bogged down in a quagmire/asymmetric warfare. Jaguar Paw leaps from the waterfall and, upon surviving, exults and loses his fear/overcomes terror and begins to score victories over the enemy.

However, I thought the theology was more subtle than a frame-by-frame remake of Passion of the Christ. At the end, Jaguar Paw's pursuers are brought up short by the Spanish galleons: they stop pursuing him when they see this incredible sight. In a very real way, the arrival of Christianity saves the pagan Jauguar Paw. Perhaps surprisingly, Jaguar Paw turns his back on the Spanish galleons dropping anchor and decides to seek a "new beginning" in the jungle.

Thanks for engaging the topic! One clarification: I said this was my best guess for our *future* filter. Most likely our future filter is smaller than our past filter, and perhaps vastly smaller.

Suppose you believe Farhi and Guth's hypothesis that we may eventually be able to create new universes in a laboratory via quantum tunneling (though we will immediately lose contact with the new universes). Once you create one, you might as well create a billion... bigger sample size.

Which is more probable in the new universes? a) Life will develop; or b) Life will develop but inevitably will destroy itself? If you think the future filter is smaller than our past filter, then you'd have to say its more unlikely life will develop... right?

Obviously, we all hope you're right Robin!

I agree that you were discussing the future filter - but that presupposes the argument that there is a great filter. If you agree that the future filter is smaller, this discussion is just about ranking most-likely X-risks - worthwhile, but a very different question.

Yeah the scenario I have in mind is most likely after a bunch of other things have gone badly wrong. So this is my candidate for the last thing that goes wrong.

You are too soft on Tyler. Sure lots of things will go wrong in the future, but societies are not like human bodies. Bodies have only one gene line, societies are more like ecosystems where many lineages indepently struggle to survive.

An animal is a temporary vehicle for genes, each female is designed to have roughly one daughter, more if she get them. Sure old age happens because of a pile-up of ailments, but evolution can tolerate that and still keep the genes immortal.

Ecosystems are different. There you have many lineages doing what it takes to survive. Many fail, others turn into something else. Regular disasters happen, but there is always someone taking advantage of the new situation.

The popularity of the Great Filter / Fermi Paradox / Drake Equation (and the amount of theorizing devoted to them) among the lesswrong crowd (and internet science enthusiasts more generally) and others like them always struck me as a bit odd given the vast uncertainty involved in the posited probabilities.

I think it is a place where they can be much less constrained with their imaginations, as they usually try to be more constrained in nearer subjects.

The great filter is one of our few concrete data points on existential risk. Do you see the focus on existential risk as also odd?

How is a hypothesis (or argument?) a data point? It seems to me the only data point we have is that we have not yet encountered a non-human, extraterrestrial civilization, but given how little we know about the number of planets, their formation, their local geologies, etc. (not to mention factors about cultural development, psychology, evolution, neuroscience, and on and on), it certainly seems premature to make any probabilistic judgments concerning how many civilizations should exist in the galaxy.

I'm confused by the phrase "Premature to make any probabilistic judgments" - you are making a judgement that the risk is implicitly small by ignoring it, in a similar way to trying to discuss it.

And despite all of this, we do have a huge data set for inference: we know a lot about planetary formation, have order of magnitude estimates for their formation, etc. In addition, we have information about the lack of obvious large scale / intelligent life on other Solar planets, the lack of clear signals from other systems, etc. This should serve to eliminate many obvious possibilities, at the least.

No, withholding judgement is not making a judgment. You can, of course, discuss potential risks to the existence of human civilization, I'm not against that. The background assumptions that spur that conversation are what I find dubious: namely that we should have encountered signals from another intelligent civilization, but we don't, therefore there is something that prevents advanced civilizations from persisting over long periods of time. The numbers used in the Drake equation are rough at best, and that's being extraordinarily generous, which shouldn't inspire a lot of epistemic confidence in conclusions drawn from it (or other similar calculations of how many advanced civilizations should be in the universe).

"namely that we should have encountered signals from another intelligent civilization, but we don’t, therefore there is something that prevents advanced civilizations from persisting over long periods of time."

If you doubt that we should have received the signals, you are still making a judgement. What is your judgement and what is its basis?

I'm not doubting either way. I'm saying there's not nearly enough information to make an informed judgement on the question.

"How is a hypothesis (or argument?) a data point?"

The fact that we exist is a data point. The fact that we have looked for ET and not yet found him is a data point (albeit a much less determinative piece of information).

'The great filter is one of our few concrete data points on existential risk.'

Seriously? The assumption of a lack of adequate information is more than reasonable, considering that our knowledge of how to recognize and generate Hertzian waves is less than 150 years old.

And our ability to handle coherent light waves is about half that old.

Are the aliens unaware of the nature of scientific development? Why would they send waves that civilizations like ours couldn't recognize?

Actually, I'll jump in here because this is precisely my interest: Why is existential risk interesting?

Existential risk is, from the perspective of any individual of our species, essentially irrelevant. Most of us will be dead of something else, along with our friends, neighbors, lovers, and children's children's children, before any existential event seriously threatens.

Even if it did, it remains a vanishingly small subcategory of threats to the organism.

The idea that we should ignore shared global public goods is a strange one. In fact, it's exactly the sorty of public good that is almost guaranteed to be under-provisioned if not explicitly collectivized.

And despite all of that, I do care personally. Any long term decision is "essentially irrelevant" to individuals; the odds it will effect me personally is relatively small compared to the overall effect... but the odds that it will wipe out the future I envision for my children and grandchildren is still significant, and since I care about them, I care about the large scale risk.

But what are the odds that you have better information and ability to influence this hypothesized filter than your great (great, great, etc) grandchildren will? Approaching zero.

existential risk is not only relevant to everyone alive today, but it's also relevant to countless future generations as well. If you were a problem solver seeking the highest levels of status, you would focus on a problem that affects the most number of people.

This is your brain on "philosophy," kids.

I burst out laughing the first time I read about these guys staying awake at night quaking in terror of "Roker's Basilisk." Rationalism indeed.

You didn't even get the name right, chances are you didn't understand the argument either ( good for you, actually).

A theory is needed to explain why we have no good indication at all of life on other planets.

Lack of spacecraft? After all, the very first probes with even the barest capability of detecting life on another world landed in 1976.

Lack of knowledge? At this point, the number of places on this planet alone where life is found has expanded considerably - including the discovery of chemotrophs, roughly around the same time Viking 1 and 2 had landed.

Lack of imagination? The above two examples merely describe a search for life, not intelligent life. But it took us a long time to realize that forms of life that live for a couple of centuries were likely communicating over thousands of miles in an ocean basin, without our awareness of that fact for essentially all of human history.

Lack of scale? To quote another distinguished authority in this area, '"Space," it says, "is big. Really big. You just won't believe how vastly, hugely, mindbogglingly big it is.'

The real answer is, they've all watched a lot of Star Trek.

Saying this as a guy who *loved* watching Star Trek

"By golly, Jim, I'm beginning to think I can cure a rainy day!"

The Great Filter is at least true: something is causing us to not see any evidence of life out there.

The belief that life is common is strikes me as anti-religious article of faith among the lw crowd, as much as they think brain-uploading or cryogenic restoration is trivial in N years.

Yes, something is causing that, but it may be something as trivial of not looking in the right place or something having to do with the geology of planetary formation or a host of any number of things. What strikes me as odd, as you point out, is the unsupported assertion that life should be more common and thus we should have had evidence for it by now.

I mean, there are like 8 million bacteria on my fingernail right now. But I don't see any evidence of them.

I don't understand your point. We have lots of evidence of bacteria, and if you went to a biology lab they no doubt could scrape your skin and start culturing bacteria right away. There is no such comparable evidence that we should (or should not) have received signals (or other kinds of evidence) of intelligent life in the universe.

Right, but it took humanity tens of thousands of years to come to this realization, despite the fact that the bacteria were literally *on their fingernails* for the entire time. (And even today, I'm basically taking someone else's word for it). The aliens, if they exist, are significantly farther away.

It's just that "one" is the most unbelievable number in the Universe. Things either do not happen, or they are ubiquitous. We've known for a long time that stars are ubiquitous. We've known for a couple of decades that planets are ubiquitous. It's extremely odd to think that life would be a one-in-a-octillion thing. Not impossible! And _someone_ has to be the first intelligent species in the universe...but does it really seem like the way to bet?

"Things either do not happen, or they are ubiquitous." That's a pretty huge assumption there, especially to ground these conclusions. Neither way seems like the way to bet. My whole point is we don't have enough information to make any sort of rational judgment.

Except "things do not happen or they are ubiquitous" applies to post-abiogenesis Filter evasion as well as the evolution of life. And to the formation of universes.

Universe formation not happening is inconsistent with any possible observation. Cogito, ergo universes are ubiquitous.

If (independently-evolved intelligent) life is vanishingly rare, "do not happen" or "one in an octillion", then almost all universes are lifeless and almost all of the infinitesimal fraction of life-bearing universes have exactly one independently-evolved intelligent race. This is consistent with observation.

If independently-evolved intelligent life is common, and Filter evasion is common, then the universe is filled with starfaring civilizations. This is not consistent with observation.

If independently-evolved intelligent life is common, and filter evasion is rare, but the Filter comes no later than the dawn of spaceflight and atomics, then the human race is extinct. This is not consistent with observation.

Independently-evolved intelligent life being common is only consistent with observation if the Filter is A: late and B: nigh-infallible.

Assuming all of these common/rare and early/late branches are, a priori, 50/50 probabilities, I get p=0.8 that the natural emergence of intelligent life is vanishingly rare, less than once per universe. I'll maybe knock that down to 0.75 to allow for a Filter that only blocks intergalactic travel and communication.

"...but the Filter comes no later than the dawn of spaceflight and atomics, then the human race is extinct."
Don't get cocky. If we aren't still at the "dawn" of spaceflight and atomics, we certainly haven't had time for a second cup of coffee yet.

I'm not sure how you stack any number of 50/50 chances to get 80 percent, but it was generous of you to knock five off as a favor.

It could be per-galaxy: say, 80% of galaxies have 0 civs, 19% have 1 civ, and 1% have 2 civs.

I have a hard time imagining two galactic-wide civilizations even communicating their existence to each other, much less having a conversation or physical contact, except on the rare occasions when galaxies overlap.

ckb: "Universes are common, life is rare", daisy-chains two coin tosses, for an a priori probability of 25%. "Universes are common, life is common, filter evasion is rare, and the filter comes late", chains together four coin tosses for a 6.25% probability. All other possibilities are inconsistent with observation. Thus, 31.25% of possible cosmologies result in someone existing and looking up at an empty sky, and in 80% of those (0.25/0.3125) the sky is empty because there never was or will be anyone else in their universe.

Dan: I don't think a Kardashev III civilization would have any difficulty making itself noticed; the problem would be staying hidden if they for some reason wanted to do so. But the math doesn't change much if we substitute "galaxy" for "universe", though there's a complexity I don't want to deal with if we have a weak universe-scale filter and another weak galaxy-level filter.

I too find it odd the self described most rational people out there are spending all their time gazing at their own navels.

(1) Advanced civilizations could exist without colonizing other stars.

(2) An advanced inter-stellar civilization could exist but recognizing that life exists here, has put our system under quarantine for fear of contagion.

(3) We could only discover an advanced civilization by its radio emissions but high energy radio broadcasts, capable of interstellar detection, are wasteful and advanced civilizations might give up on such technologies in favor of low energy and/or non-broadcast alternatives.

Non-colonizing advanced civilizations could exist, but just one colonizing civ is enough to leave traces all over this galaxy (and possibly others, based on galaxy intersections).

Especially if their preferred mode of colonization involved comet like generation ships that travelled between gas giants, that being such a civiliation's more than sufficient environment for all the resources it required, with the proper environment to prosper, using slingshot orbits around suns.

Or we wouldn't notice them at all - after all, we didn't notice the constant rain (20 a minute - http://www.spaceref.com/news/viewpr.html?pid=3997 ) of water comets hitting this planet until the 1980s.

We haven't exactly explored the galaxy. Heck, we've only just begun checking out of Solar System. For all we know there's alien monitoring probes keeping tabs on us out on the moons of Saturn or on some unremarkable asteroid-- and communicating home via Elsewhere (AKA the Bulk) in ways we don't have the physics to even start to grasp. Assuming everyone is using radio waves, or some other form of EM, to communicate is like assuming all civilizations use clay tablets inscribed with cuneiform.

I find the "Killing Star" scenario to be quite compelling. Once you have a civilization has the ability to send a ship at close to the speed of light, they effectively have the ability to destroy any planet without any significant warning (relativistic weapons).

This sets up an incredibly potent prisoners dilemma with almost no communication and monitoring possible. Imagine a Cold War dynamic where the first strike was the only rational action.

Even if there are no other space civilizations, our own continued expansion would increase the chances of one or more rogue colonies wiping out the others.

Suppose it was 1492 or so and we were applying the mathematical tools used by SETI analysists to question life on earth. Suppose we knew that the earth was very old, that human life had at about 100,000 years or so on earth. Given that smart people of the time knew the earth's true size and could calculate some basic models of population growth, it might seem very sensible to assert humans should have the earth filled up.

Yet humans had not filled up the earth. Large areas of North America, for example, were sparsely populated. What went wrong?

Well it might be colonizing the stars is possible in theory but has a much more modest return for advanced civilizations that have been around our galaxy for a while. As a result colonization slows to a craw after some initial spurts. While in theory the galaxy will eventually see every star populated if a civilization does one star per 100 years, the actual rate could be even slower than that.

To bring the analogy home, it was possible in theory for humans to have already built up a Europe sized civilization in N. America to greeted the Europeans when they 'discovered' it. But they didn't. Perhaps if Europe had taken an extra century or two to make it over the Atlantic they would have or perhaps it would have taken a thousand years. The game of colonizing the larger galaxy may just be possible, and even likely and yet we are at the front of the race. Or the race may have been steaming ahead for quite a while now and we just haven't noticed it yet.

Limited, initial European contact with coastal native American populations spread diseases to the rest of North America that wiped out many of the natives there before Europeans could meet them. In 1492, there were large civilizations in North America (mound builders), who were wiped out before more than just a few white men were able to see them.

The migratory tribes that European explorers encountered in the 1500's and 1600's were post-apocalyptic survivors, a rump population.

High estimates of the population of the America's right before Columbus top out at 100M. That includes South America. Not seeing it at the moment, but I suspect North America had a very low population density relative to Europe and Asia in the 15th century, before Columbus.

Running a simplistic calculation of starting with a small human population of a few housand sometime around 98,000 BC in Africa and figuring the population will expand so many square miles every year, I suspect you'd get a result that would say the entire world should have had the density Europe did by 1492...if not more. Just as some calculations today say that if advanced life could have developed maybe a few billion years ago at the earliest, we should already see nearly the entire galaxy population...even if they took their time moving from star to star.

The population of Europe in 1400-1500 was on the order of 50 million. Land Area of Americas is 3-4x that of Europe. So 200m sounds like a reasonable order of magnitude guess for having the same population density... and behold, it's only off by about an order of magnitude.

An order of magnitude is 10x

"An order of magnitude is 10x"

How decimal centric of you. Real men use base two.

Agriculture was only practiced in about a third of North America (and was a recent development everywhere north of Mexico). Hunter-gatherer lifestyles predominated elsewhere. Except in areas of unusually high natural fecundity you get very low population numbers with hunter-gatherer cultures.

There were parts of North America that were thinly populated-- in fact there still are. A huge fraction of the continent lies north of the US-Canadian border and the population density thins pretty fast as you travel north of it.

"Yet humans had not filled up the earth. Large areas of North America, for example, were sparsely populated."

The population probably reached the carrying capacity everywhere given the level of technology and the climate. Russia remained (and still is) sparsely populated given climate. Even in North America today, there are large regions that are sparsely populated. My quick estimate is that a one hundred mile radius from where I am sitting has an average population density of 5 people per square mile. Pre-contact, density would have been much lower.

There could be a "forbidden" technology waiting to be discovered that wipes out civilizations a little more advanced than ours. Man made black holes? Self replicating nano bots?

Well there might likely be a forbidden technology, but I don't think either of those two are it. Any human-scale black hole would dissipate nearly instantly due to Hawking radiation. As for self-replicating nanobots, Earth's been full of them for billions of years, and no existential disaster yet. Human-made nanobots face the same limitations as bacteria strains. Among many other factors, their spread is limited by the rate that they can extract energy and the very high viscosity that microscopic objects face in fluids like air.

There's a misconception that predators are inefficient. Romans and Germans may have been brutal predators, but they were very efficient, efficient in ways other than killing their prey, such as saving and investing, investing in things other than armaments (like highways and aqueducts for transportation people and water (Romans) and factories and equipment for building more armaments (Germans)). Indeed, as predators Romans and Germans share many characteristics, including tactics for scaring the prey into submission. The Romans crucified rabble-rousers, a horrific death for sure but they would leave the body on the cross for days to rot and be devoured by scavengers, a sight other rabble-rousers couldn't forget. Everyone knows about the German brutality during WWII, but few are aware that the Hessian mercenaries who fought the colonists in America would cut off the heads of the revolutionaries and put them up on stakes for the rest of the revolutionaries to see in an effort to frighten them into submission. I think Hanson's image of predators is based on the Mad Max series of movies with Mel Gibson. It's the civilized predator that is the problem.

Romans or Germans aren't great filter threats though. Yes they may be brutal and evil, but they're not progress-halting. Precisely for the reason you identified: they still invest in non-military related things. At any non-trivial level of investment, technological progress is inevitable. A galaxy filled with Nazi planets is still a galaxy that will eventually be colonized. It certainly is worth worrying about "civilized predators" like this, but it's not worth worrying about in the context of the great filter.

I'm not sure you could accurately describe either the Romans or Germans (by which I think you mean Nazi Germany) as 'predators'. The Roman army might have been a predator in the sense that it fed itself by the spoils of conquered land but Rome as a whole fed itself through trade and industry.

In terms of long history, Nazi Germany was a historical blip, gone from the scene almost as fast as it arrived. But even there its agenda was to empty out the land it took and fill it with its own industry. Not take consumption products from the lands it defeated.

A better example of the predator type society I believe Hanson is talking about is the Walking Dead series. There it seems while human groups are far apart, they are not so far apart that it doesn't make sense for at least a few groups to hunt down others and take what they have to survive. Less predatory groups are essentially trying to scavage leftovers from civilization before the zombies. Brief attempts to revive industry (think the farming attempted at the prison) end up failing because predatory groups will always seek to take over such enterprises. But predatory groups never get large enough to be able to defend enough turf to start investing in industry.

Despite brief shining stars, the long run prognosis for such a society is a gradual decline in the total human population until it gets so low that essentially small random factors will ensure human extinction.

Actually the Roman army was predatory. The problam for rayward is problem is that this was highly inefficient. My guess is that the Nazis would have gone down the same road had they survived.

It's true that the late republic and early empire were predatory but (arguably) constructive. But then then legions found that by selling crowns, they could extract wealth from the rest of society. But the process required them to plough much of the plundered revenue into civil wars. The crisis of the 3rd century was entirely down to predation.

Eventually Diocletian et. el. came up with some new institutional arrangments that could cope with this predation better and things settled down a bit. But even those normalised a situation where the army was mostly useless for actually defending against barbarians.

Oh. And barbarians are also predators.

The Roman Empire expanded about as far as it could given existing communication and transportation technology. Hadrian abandoned Trajan's conquests in the Middle East because they were just too far away to be controlled and defended.

I think Lord Humungus wanted to rebuild civilization no less than Pappagallo did, with the condition, as he said, that "we do it my way."

Bodies generally were not left on crosses like that: there was only so much room available for crucifixions so you had to take the dead ones down to hang up new victims. We've even dug up the properly buried body of a man who was crucified so we know that burial was possible (presumably if a family asked for the body and could afford to provide for burial). The ancients were also superstitious and leaving the bodies of the dead unburied was apt to provoke the attention of their unhappy ghosts. Consider "Antigone"-- yes, I know, Greek, not Roman-- where leaving the bodies of the rebels unburied is assumed to be a monstrous atrocity right from the get-go. (Yes, sometimes that did happen-- but usually in slave revolts or wars.)

I don't necessarily disagree with the hypothesis. But predation is highly inefficient. Ecologically speaking, 90% or more of produced calories are lost at each level of the food chain. It would be far more efficient for gazelles to simply produce detachable meat and peacefully remit it to the lions. Like the wheel, such solutions simply seem to be outside the scope of biological evolution. But this type of behavior is pervasive across human civilization: rents and taxation are far more common than outright predation. I see no inherent reason why this would change in the far future. The technologies that enable this (modern forms of law, states, and militaries) are unlikely to be lost.

Efficient stationary bandits, even when controlling much smaller domains, can marshall many more resources than inefficient roving bandits. In human history we only see short bursts where the latter overwhelm the former: the Huns, the Mongols, the Conquistadors, the Wehrmacht. But then equilibrium quickly sets in and the predators are either defeated or convert to law-respecting governments. Pervasive predation would portend the near stopping of progress, the vast majority of resources would be devoured or wasted in cat-and-mouse pursuits. But even if all governments are run by corrupt and brutal tyrants, say on the order of North Korea, progress would likely slow no more than an order of magnitude. Probably not enough to qualify as a great filter candidate.

"But even if all governments are run by corrupt and brutal tyrants, say on the order of North Korea, progress would likely slow no more than an order of magnitude."

When Rome collapsed and was replaced by small warring kingdoms - progress essentially stopped for a thousand years.

China turned inward in the 15th Century and progress seems to have stopped until Westerners forced progress on China. Japan seems to have turned inward around 1600 and that lasted until the Americans sailed into Tokyo Bay.. Islam basically deliberately abandoned progress in the 11th Century.

We have historical examples that a complete stop to progress is possible.

But many technologies kept advancing after the fall of Rome, from metallurgy to ship building to agricultural implements. All of these were better than they were in the Roman Empire within 500 years of the Fall.

A thousand years after Rome fell in the Westeven architecture and military engineering had advanced beyond Roman capabilities. Only in fields like mass production of basic goods and engineering projects such as aqueducts and roads that required large peacefully integrated states did Rome surpass Western Europe by the 15th century.

AlsoIhave towonder how much of the perceived technical decline happened not in the dark ages but during the last two to three centuries of imperial rule in the West.

> When Rome collapsed and was replaced by small warring kingdoms – progress essentially stopped for a thousand years.

RoyL's comment covered how progress wasn't as slow in the middle ages as commonly believed. The question for the great filter isn't whether progress slows by a measurable amount, but whether it slows enough to stop eventual interstellar colonization. For that to happen progress has to slow tremendously. At anywhere current rates of growth and innovation, interstellar colonization will be possible well within 10,000 years. The Earth has about 1 billion years left to support life. For stagnation to be a plausible future filter, progress would have to slow by at least four orders of magnitude. The growth and innovation rate difference between Rome and feudal Europe or South and North Korea is nowhere near this drastic.

In addition to stagnation there could be periodic bouts of massive destruction. But as long as some amount of knowledge is preserved across the bouts of destruction, the effect probably isn't enough to stop long-term growth. After the re-emergence of strong nation states, Europe quickly surpassed the peak of the Roman Empire. It could simply adapt preserved innovations to catch-up much quicker than it took Rome to pathfind that level of development.

'progress essentially stopped for a thousand years'

Movable type printing was invented in China, and metal type in Korea. It just took the Europeans another couple of centuries to imitate it.

Re: When Rome collapsed and was replaced by small warring kingdoms – progress essentially stopped for a thousand years.

This is simply not true. Technological progress continued at a slow pace during the Middle Ages. As did social progress. Granted, the pace was glacial compared to the last couple of centuries, but it did happen.
The only part of the Dark Ages that really deserve that name was the period from the mid 6th century through the mid eighth century. Demographic collapse produced by natural disaster (this happened globally, though some areas recovered faster) did cause a severe downturn. But by Charlemagne's day Europe was on the move again.

Colonizing Space would require a whole lot of money, money which can only be provided, it looks like, by large governments. It would be unthinkable for our government to make that commitment today. The left would rather spend that money on the underlcass and government workers, while the "right" wants to spend it on tax cuts for the wealthy.

America looks to me like it's bound to become at best a second rate country in the future. What if the other countries that will replace it have the same dynamic? Imagine the world is 2400 where China, Japan, and Russia are the economically dominant countries and technology has not notably advanced in the last 100 years. Their elites, like ours, will see no reason to want to colonize space. Who would benefit? What would be the reason? They'll have all the resources they need on earth. Overpopulation won't be a problem. Nationalism might be dead, they've proven vulnerable to terrible ideas from the West before, or it might be unnecessary because there would be no one to dispute China's dominance. Maybe the best and brightest in Chinese society will no longer see a reason to study science. Why do it, when going into business is so much easier and pays more? Everything has already been discovered.

Well even if that's true in 2400, what about 20,000 AD? At anything even close to current trends in energy production, miniaturization, artificial intelligence and nanotechnology, self-replicating interstellar probes should be well within the economic grasp of many small to medium entities. Similar to how today many smaller governments and even private organizations can put satellites into orbit. A cultural shift away from science and exploration would certainly delay future interstellar colonization from our perspective, but it seems highly unlikely to make a good future filter candidate. A galaxy where intelligence species tended to colonize 10,000 years post-industrialization, rather than 500 years, would still almost assuredly be colonized in short cosmological timeframes.

How fast would colonization happen? The 'seed' idea is small ships sent off in different directions, once they hit a different star they spout up new civilizations after a centuries or so.

But I'm not sure colonalization followed this pattern in our history. Europeans who came to America did not shoot off in random direction but, for the most part, headed towards already established colonies and then started moving off in different directions from there.

In this model a civiliation might take a few thousand years to establish an outpost on a nearby star but then take a 100,000 years slowly sending people to that star system. Only after an extended period of colonalization are efforts made to establish something on yet another star. If colonialization is extremely slow, then it is possible that we have spacefaring civilization that are millions or hundreds of millions of years old and yet have not filled up such a large portion of the galaxy that we can't not notice them.

Many or even most civilizations could follow the model of patient colonization. But it only takes one to colonize moderately aggressively for us to notice. Even assume colonies take 100,000 years to get up and running, and they only colonize two more star systems after doing so. With 400 billion stars in the Milky Way, that only takes 39 cycles to colonize the entire galaxy. That's only a 4 million year delay relative to an aggressive colonizer that replicates new seeds in very short time. Relative to geological time that's a rounding error.

For slow colonization to be a plausible filter, colony replication time would have to be on the order of 10 million years. I can see scenarios where civilizations choose to stop colonizing or never do so in the first place. But it seems unlikely they'd want to colonize, but only at an implausibly slow glacial pace.

Perhaps aggressive colonlization doesn't make economic sense. Perhaps the energy required is massive and one gets more 'bang for the buck' by building up local colonies rather than filtering out to new star systems. Barring anything like wormholes or warp drives, the technology required to do colonization would have to depend a lot on energy conservation and recycling. it would also probably require strict population control as well.

Given that technology and culture, what is going to happen when the colony is established? Exploitation of the resources in that new solar system would happen very slowly IMO. It would be a long time before there would be any economic need to entertain a new colony.

Another factor to consider is perhaps the ability to sustain long term space travel is limited. You are essentially required to build a machine that has to be able to function for several hundred to several thousand years. We have never done anything like this. This might work if the next solar system is a handful of light years away but I'm sure there are places where 'short hops' run out and an expanding star civilization would need to make a leap of a few dozen light years. At that point your choices are either to accelerate closer to light speed (which requires a huge amount of energy) or manufacture ships that can maintain themselves for much longer periods of time.

A final factor is 'stellar gentrification'. Imagine a civilization that is a million years old and has colonized a dozen star systems with a few hundred billion people in each. Yes those looking for a move might make out for a new, 13th star. Or they might go to one of the 11 other systems instead. Perhaps on some stars civilization has collapsed and the next colony will be a 'return' to that system.

As the number of populated stars goes up, the number of choices for interstellar travel that consists of going to an already colonized system rather than a brand new one increases.

It seems somewhat unlikely to me that the human race will go extinct in the near future. Drastic population decline? Sure. Return to pre-industrial existence? Sure. But not extinction.

There may not be an extinction but there will be dramatic changes, especially in the advanced societies. The survival of positive genes is even now being compromised by modern medicine. Genetically inferior humans that wouldn't have survived to reproduce and pass along their bad DNA just a short time ago are now doing so. Type 1 diabetics can pass along their genes to their children. Eventually, populations will be composed of increasingly blind, deaf, semi-mobile, unintelligent subhumans with really crooked teeth.

As Razib Khan notes, this is unlikely, as parents want the best for their children and pre-birth genetic analysis is getting easier and easier, making the future more likely to be one of decentralized eugenics.

If there were a million civilizations like ours through the galactic history, and we expect that its even odds that all of them have been "filtered," then the probability that any one of them would have been filtered is roughly 0.99999931. If we are to assume that the great filter isn't any one thing, but rather a number of things working together at the same time (as opposed to the belief that the great filter is a number of different things working independently of one another, which is a better argument, albeit one I think is still wrong), then we can actually place some probability on the outcome of any given condition taking place.

(assumptions for this is that a bad things last for 20 years, it takes 4,000 years from when a civilization becomes "like ours" to the onset of interstellar colonization, it takes 4 bad things happening concurrently to filter a civilization, and, as previously mentioned, 1 million civs like ours). Doing the math (that any given civ must have a 0.99999931 chance of being filtered), we can come to the conclusion that, at any given time, there is a 50% chance that a given bad thing will happen at any given moment. (200 instances of a condition of a probability of 0.5115 gives us a probability that, at least once, all 4 end up favorable of .999999307, which is the probability we'd need to assume a 50% chance of a million civilizations being destroyed during their history).

So basically, we should assume that at any given time, there's a 50% chance we're halfway to being filtered (two bad conditions). Furthermore if we say we became a civilization like ourselves in 1950, (which I think is as good a year as any), then we'd have had a 60% chance that three of the four things happened at once. (of course, for all these examples, I'm using only best guesses, but they are all reasonable. The ones that seem the most "made up" are the number of civs, which the model is least responsive two. Changing the number of civs by a degree of magnitude seems to change the chance that a given "bad thing" occurs at a given instance by about 2 percentage points.) Given that, as far as I know, we've never seen anything which can be close to a "bad thing," (nothing which puts us a quarter of the way to extinction, by any means) I'd go ahead and reject the model out of hand.

But I'm getting ahead of myself here. Looking at the original analogy of again, dying of old age isn't really "a bunch of bad things happening at once," at least not in the way my model explained it. If that were the case, we'd be measuring life expectancy in half lives, which isn't what we do.

What old age really is the accumulation of the ravages of time. So, according to Prof Cowen, we can expect that bad don't so much happen as they accumulate; so increase in environmental degradation and wars will eventually do us in. While this may seem more reasonable, it is still most certainly wrong.

The probably of A and B and C has to be less than the probability of A or B or C (except at the 1 and 0 bound, but those aren't interesting). So if we are to agree that the probably of the great filter is A and B and C, and we fix the minimum value of the probability of the great value at 0.99999931, then the probability of any one of them has to be 1-(1-0.99999931,)^3, Which again, seems wrong: basically its the belief that there are three problems which every civilization will run into, each with a probability 0.999999999999999999667, and furthermore that the civ can withstand any two conditions, but the third will doom them? Or, to move away from math terms, that any civilization which could stop environmental damage or achieve a world government or a disarm its nukes would colonize the galaxy, but no civilization out of a million has managed to do any of those three? Its much much more explanatory to say that any given civilization would have to do all three to survive. (Its even more explanatory to say that a single one of those is the problem, and that, after a long enough time, enough people have nukes that not a single civ manages not to blow itself up).

We should always spend some time thinking about these puzzles, but I think the most important conclusion to be drawn is that we know very little about the nature of Reality and our place in it. Our species has learned a lot of clever tricks in the last few generations, and that puts us at risk of getting a bit puffed up. It's good to reflect that Kant thought we had things just about worked out two centuries ago...and Aristotle, two millennia ago. It's good, too, to be alive in a moment when we are realizing that we don't even know what the physical universe is _made of_--much less do we know to what extent the physical universe is the fundamental story of Reality. (Kant taught us, after all, that space and time themselves are merely the forms of our perception. Whether they are properties of things as they are in themselves...I think we can't know.)

Which is to say, the Great Filter might not be such a bugbear after all. It might be that interstellar travel is something that just doesn't get done because it's not something that turns out to be worth doing. Species that could, see no reason to. Or are moved into other channels of activity by some greater instrumentality. And if that sounds a bit mystical, I can only say that we are talking about a scope where Clarke's Laws come into play.

It seems to me once truly immersive virtualization comes about (and we're pretty close here on Earth) most civilizations would be happy to stay put and only virtually explore/zone out/whatever. It's almost infinitely easier and somewhat scratches the same itch as expensive, dangerous, extremely time-consuming interstellar exploration.

For a great filter to be 'great', it has to apply to almost everyone - not an act of randomness, or just 'lots of bad stuff happening at the same time', but some sort of logical progression that civilizations take that ends in their destruction before they develop the ability to move between stars.

My candidate for that is simple: The inevitable course of technology is that it unleashes the capacity for destructive power faster than it can be counteracted or defended against. In the past, malevolent forces needed armies to carry out wide-scale destruction. Today, a madman who gets his hands on a backback nuke can start WWIII. In the future, it may be biological weapons, or nanotech weapons. What happens to us when we have molecular assembly available to the masses? What if terrorists or apocalyptic cults have access to technologies that can create widespread pandemics or self-replicating assemblers that can turn a planet into gray goo? Can we avoid that kind of scenario?

Another possibility is that our science will get to the point where we carry out some kind of experiment that has devastating consequences that can't be foreseen. A particle experiment that releases massive amounts of energy from an unknown source or something like that.

However, we don't need to even posit a 'great filter' so long as the Drake equation can be fed reasonable terms that result in an answer of approximately one. No Fermi paradox or great filter required if the answer to the Drake equation is that the average galaxy should have no star-hopping civilizations at all.

This is not difficult to conjecture. We have not found life anywhere else in our solar system so far, or even fossils of old extra-terrestrial life, despite finding many environments suitable for life in the solar system (Mars, Europa, Enceladus, Ganymede, the clouds tops of Venus).

Nor have we any evidence of a second, independently evolving chain of life on Earth.

The more we learn about how robust life on Earth can be and how surprisingly hospitable to life other worlds in our solar system are, the more perplexing it will be if we don't find it anywhere else in the solar system, and so far we`ve come up with nothing.

And even if we can quantify how often life will form around other stars and prove that it does form elsewhere, it`s a long way from basic multi-celled creatures to space-faring civilizations. It may be that life is ubiquitous in the galaxy, but that civilizations are extremely rare - rare enough that you would expect a galaxy that contains one to be extremely rare.

I say `galaxy` because realistically we can`t hope to detect civilizations in other galaxies unless they are so powerful as to be able to absorb a significant amount of the galaxy`s energy. Certainly standard communications using transmitters and receivers of electromagnetic energy would not be detectable by us if they were happening in other galaxies.

So, we should restrict the debate to our own galaxy. And within that debate, we can rule out the center of the galaxy where the radiation is high and collisions common. We can eliminate all the places where gamma-ray bursts may have cut swaths through in the past. We can eliminate any stars that that were sterilizingly close to supernovae within the last billion years or so.

Then take out the young stars were life could not have formed, the ones that are too hot, too short-lived, or too variable . Remove the stars in regions low in heavy elements, etc.

How many potential civilization-carrying stars are left after you remove all that stuff? Perhaps millions, or maybe even a few billion. That sounds huge, but it's a number small enough that other factors may result in civilizations developing only in extremely rare cases.

The laws of probability may be the only 'great filter' we need. There may not be one big event that stops a civilization from developing or surviving, but so many low-probability destructive events that given billions of years one of them getting you is a virtual certainty. We may just be incredibly lucky.

Of course, that collection of low-probability events may BE the 'great filter'. Maybe it should be called "The great strainer'.

This is a very good argument.

The big problem with tthis Fermi paradox stuff is it assumed we know far more about the universe than we do.

In any event, you and I may well live to see the spectrograms of a few hundred alien planets. And won't that be a heck of a thing?

Which considering that the very first exoplanet was discovered two decades or so is an interesting data point when looking at the fact that the Drake equation was first posited about 5 decades ago. We might actually be able to survey an exceedingly tiny number of stars' planets within a few millenia at this rate. Assuming we stay civilized and interested, that is.

My guess is that there is at least one MR reader trying to figure out to monetise the apocalypse.

No commenter left behind.

I wish I had thought of this first:

"the future great filter scenario that most concerns me is one where our solar-system-bound descendants have killed most of nature - "

Sounds like a call to action to me. I've been reading Jeff Sach's book, "The Age of Sustainable Development." He talks about related issues and what we can do about it. The threats are not what a few freaks can go. The threats are what mainstream, business as usual, processes can do.

What good does it do humanity if it escapes to the stars only to engage in continuing pillage and plunder strategies? Maybe humanity has to learn to live within limits because there is no new wilderness left to plunder.

The great filter is humanity's built in propensities to destroy the very life that sustains it.

Comments for this post are closed