At Least Two Filters

Where lies the great filter, i.e., the obstacles that make it extremely unlikely that any one chunk of pre-organic matter originates a visibly expanding interstellar civilization? While it seems unlikely our ancestors passed through much of a filter in the last half billion years, our descendants may face a big filter in the next few thousand years, and there may have been big filters associated with the origin of life, the spread of life, the invention of complex cells, sexual reproduction, or multicellular life.

In many folks eyes, an elegantly simple resolution, which is likely because of its simplicity, is to assume there is just one huge filter: the origin of life. Assuming that first step is enormously hard allows one to think all the other steps are pretty easy. They wouldn’t be sure things of course, but conditional on a big enough origin-of-life filter, one wouldn’t have a strong reason to fear that common analyses underestimate future filters.

Unfortunately, the elegantly simple hypothesis that the great filter is mainly a big origin-of-life filter seems at odds with our best evidence. Why? Because if the spread-of-life step had the weakest possible associated filter, then life spreading must be easy. Over billions of years life could have spread to many star systems from its place of origin:

Life could spread across a galaxy via giant molecular clouds reliably collecting life from the stars they drift near, and then passing that life on to a few of the thousands of new stars they create.

If over billions of years life spread to many hundreds, or even billions, of star systems, and no substantial filters stood between arrival of life near a star, and its eventual development of advanced technical civilizations like ours, then why would we now see no any evidence of other civilizations? Yes it is possible that we are the very first, but that hypothesis is of course unlikely by default.

It seems to me that if the great filter is to consist of just one big step, the only plausible possibility is the development of multi-cellular life. All the steps before that one seem able to spread to other star systems via single-celled life hidden in dust, and it seems we haven’t had a big filter step since the multi-cellular innovation.

So if the idea of just one big filter appeals to your sense of elegance, you’ll have to presume that life, including complex life with sexual reproduction etc., is very common in our vast universe, but that Earth is one of the handful of places in all that vastness with multi-cellular life.

If you don’t find that plausible, well then you’ll have to grant there are at least two filters. And if two, why not three? So you must find the possibility of a third filter in our future plausible; beware future filters.

GD Star Rating
Tagged as: ,
Trackback URL:
  • Roko

    I still don’t believe your argument that intelligence is not a filter.

    To paraphrase, you say that since brain size increased continuously, getting to a large brain can’t be the hard step.

    But this doesn’t follow.

    If you run N=10^10000 runs of evolution, you might find that only k=100 of them result in large brains, so big brains are a filter. Then you look at the 100 where you did get big brains and see that in all of them, brain size increased continually.


    Because conditional on evolution achieving something, it’ll probably achieve it incrementally.

    • Roko

      To be precise, treat log brain size as a discrete time random walk where the change in brain size at each time interval is drawn from a normal distribution with, with time measured in units of 1 billion years. Run 10^1000 simulations, and then restrict to cases where log brain size is very high. You will probably find that all of these anomalously high achievers look back at a history of continuous increase, rather than one sudden jump.

    • Selection would only predict continually increasing brain sizes along the lineage that lead to us, unless there’s a lot more DNA sharing than we think. But we see increasing brain sizes in other lineages as well.

      • Anonymous from UK

        Or unless there are other cross-lineage corellations that you haven’t thought of. For example, predator-prey relationships would be a source of cross-lineage corellations, as would “lucky” choices of early biochemistry, lucky choices for reproductive mechanisms like sex.

        And by the way, it seems to my naive mind that brain size corellates pretty well with phylogenetic distance from humans. For example, of the five kingdoms, none other than ours developed anything like brains.

        This is exactly what we would expect to see if intelligence was a filter.

      • Chris T

        For example, of the five kingdoms, none other than ours developed anything like brains.

        This is exactly what we would expect to see if intelligence was a filter.

        Exactly, as I’ve pointed out before on this blog, out of 70 phyla, only one class in one of those phyla has ever led to an organism capable of radio. To make the statement that Hanson has made here requires ignoring the vast majority of evolutionary lines that didn’t result in greater encephalization.

  • Eric

    Sadly agreed with all of this.

    It would be nice to hope that the “great filter(s)” are behind us, but evidence doesn’t give me much comfort. Looking at the history of life on Earth, I don’t really see any evidence for past “great filters” unless life itself is really hard to get started (timing wise, that seems unlikely, since life looks as old as it can be on the Earth). Multi-cellular life appeared early and repeatedly, so this doesn’t look like a filter either (

    I also don’t see hostile aliens acting as a filter. Looking at the Galaxy, it seems any other life out there is very inconspicuous. A couple of points:

    (1) If there were hostile aliens, they don’t seem to worry about life, even complex life. The Earth’s atmosphere can be easily observed by outsiders for indications of a rich ecosystem. For the past 3+ billion years, it looks like nobody interfered. Life destroying Berserkers seem unlikely.

    (2) We’re currently making ourselves conspicuous in new ways, with radio/radar emissions and industrial pollutants in our atmosphere. It’s too soon to say if hostile aliens will take notice, but I doubt it, since if they wanted to rid the galaxy of smart competition, they could have just as easily wiped out complex life on Earth 100 million years ago or more. I don’t see radar or atmospheric CFC’s as that meaningful of a threshold, but of course, I could be wrong and their invasion fleet may arrive tomorrow. 🙂

    I suspect we’re now reaching the important filters, probably involving some interplay between intelligence and self-destruction. Since no aliens seem evident, the “coming filter(s)” seem pretty severe. Perhaps refuges are futile, since such measures should have protected at least a few of the more paranoid in alien civilizations.

    Let’s pray that future missions to get spectra from the atmospheres of extra-solar planets find only dead inert worlds in chemical equilibrium. That’ll make the past filter argument stronger.

    • “I suspect we’re now reaching the important filters, probably involving some interplay between intelligence and self-destruction.”

      You may not think Bezerkers likely, but I think a decent case can be made to put most of our investment into hiding from Bezerkers (at least from Bezerkers with whatever our own cutting edge detection technology manages to be). Because I think the probabilities indicate all other strategies to avoid future filters will fail.

    • With panspermia, our life could have started before Earth, so the early appearance of life on Earth doesn’t mean its an easy step.

    • if multi-cellularity arose many times on Earth, then if there is a big associated filter it would have to be regarding a particular kind of multi-cellularity, e.g., animals.

  • Michael Kirkland

    If life can easily spread between solar systems, shouldn’t we expect to find it in more than one place in this one?

  • I wonder if finding yourself in an atmosphere full of O2 is like finding yourself in an ocean full of jet fuel or like roaming across the savanna and finding lots of convenient supermarkets already built – that the ridiculous step was the one where organisms pumped huge amounts of fuel into the atmosphere, enabling large motile organisms to exist, instead of ever figuring out how to burn that fuel themselves.

    • JS Allen

      Interesting point. I just finished Matt Ridley’s “Rational Optimist” where he makes a lot of arguments about humans just “happening” to find ourselves sitting on top of big piles of fuel.

  • rapscallion

    Two possibilities:

    i) Perhaps there’s some horrific anti-life tech that only a few civilizations actually use, but that wipes out even the possibility of life within, say, a 10000 light-year radius.

    ii) Maybe hostile aliens are ideologically committed to eradicating only intelligent life, not all life, which is why they haven’t killed us yet. This possibility is dealt with in some of Alastair Reynold’s books.

    • Eric

      Or perhaps we’re just running as a simulation in some alien computer. It’s a nice solution to the Fermi Paradox, a variant of the “zoo hypothesis”.

      Sadly hard to test. I wonder how one would build an effective refuge inside of a simulation?

      • Finch

        Perhaps a virus of some sort?

      • I’m persuaded the odds are overwhelming that we’re in a simulation (there are many more simulations than the “one true reality” in our observable reality already), but I don’t think it changes much in terms of our local survival strategy.

    • Pre-human life is intelligent, calculating and colonizing. The first salient distinction is intrastellar life and interstellar life. Pre-human life doesn’t seem to me to be robustly interstellar (we don’t see it running amock on the moon or mars, for example, which isn’t even interstellar yet -though I guess “simple” life could be robustly interstellar on more earthlike planets –though Mars wouldn’t be enough?).

  • Evan

    What about post-intelligence filters we might have already passed? Like developing agriculture, science, markets, industry, etc.? Maybe all the aliens out there have been hunter-gatherers for the past few million years and we got lucky.

    • We achieved those things in a very short time after having big brains. Even if we were unusually early on those things, imagine farmers with a hundred million years to keep trying to invent industry.

      • IVV

        Maybe we’re lucky because we’re sitting on a giant battery of energy in the form of fossil fuels. Certainly it was necessary to achieve our first forays into space.

        Then again, Titan would seem to indicate that this isn’t much of a filter, either…

  • Self-organization is a law of the universe. Life is common. Extremely common. Emergent complexity is another law of the universe. Thus, intelligence is (or will be) common. However, someone has to be first. I don’t see why it’s any more or less likely to be us as anyone else. Entropy is a measure of our ignorance of a system — and so is this idea of “filters”.

    • scott

      “I don’t see why it’s any more or less likely to be us as anyone else.”

      Sol is a third-generation star. And a relatively late one at that – generations happen faster nearer to the centers of galaxies. So if life is common, it is extremely unlikely to be us; that very common life should have occurred in the previous two generations of stars well before us.

      • Chris T

        So if life is common, it is extremely unlikely to be us; that very common life should have occurred in the previous two generations of stars well before us.

        Life requires heavy elements which are only known to be produced in supernovas; sufficient quantities probably didn’t even exist until the third generation of stars.

      • I see Chris T already beat me to my explanation. 🙂

  • Origin of life as the only substantial Filter would when it occurred lead to life spreading across a galaxy sooner (via panspermia) or later (via starflight). Panspermia thus doesn’t substantially change the logic. It’s still by far the most likely Great Filter that a life origin event occurs several orders of magnitude less than than once per galaxy per 10 billion years.

    BTW, panspermia requires far, far more than survival of most DNA. It requires the parallel survival of nearly all of the billions of bits worth of sophisticated metabolic structure along with practically all of the millions of bits that code for that structure. A codon half-life of 1 million years means that a viable organism is destroyed in a few days. The longest-lasting viable life forms are seeds and spores that have been evolved for long dormancy, and that requires sophisticated multi-cellular life. Given the cosmic ray environment of space, survival of all the necessary bits over interstellar transport times is more than astronomically unlikely, it is basically impossible. And indeed, wishful thinking about red comets and nanobacteria notwithstanding, we’ve never seen anything that remotely comes close to a viable lifeform in a meteorite or asteroid or comet or lunar sample.

    Wildly distorted speculations on the origin of life is where astrobiology hype (needed to keep the funds flowing to astrobiologists) about the discovery of ET life being just around the corner goes astray. What they want us to forget is that the simplest known life form is astronomically complex. The simplest earth life is full of thousands of very complicated proteins, each of which originated, as far as we can tell, only once in highly improbable events. If they were at all probable in the universe they should have originated multiple times. Thus (a) origins of life must be astronomically improbable, and (b) survival of viable life forms across astronomical timescales as hand-waved by Napier et. al. is practically impossible.

    This we know by observing life on earth itself. We need no grand fancy speculations about Drake Equations and Great Filters to arrive at this conclusion. It does happen to be consistent with the Great Filter and thus by far the most likely explanation for the Great Filter, if one insists on thinking about the origin of life in such a wildly speculative and backwards fashion instead of observing life directly.

    If we want to think about such Filters in more scientific ways, think about them in terms of specific chemical events rather than in terms of vague macro-events like “the origin of multi-cellular life”. It turns out to be rather difficult to define such an event since there are a variety of intermediate phases of cell colonies, genetic interchange and sexual reproduction of a dizzying variety of kinds, and many other complications such that there is no well-defined improbable discrete event “the development of multi-cellular life.” But the origin of each kind of protein in simple cells is a discrete and usually wildly improbable event. Gradual evolution via natural selection can only get us so far back: many of these proteins need to exist together in a metabolism for the functions of the others to make sense, making the improbability of their joint origin a single Filter. In all this we may find many Filters (several sets of improbable metabolic changes) or one Great Filter, a set of metabolic changes that largely had to occur all at once.

    One enzyme that appears to have originated only once in a very improbable event was crucial for overcoming a basic metabolic limitation is nitrogenase. All early life required nitrates or ammonia to build amino acids. It was unable to split the far more abundant and concentrated atmospheric dinitrogen (N2) in order to create amino acids. This was the nitrogen crisis. This far more abundant source of nitrogen was required before life could move out of a very scarce and sparse existence in the very rare areas where nitrates or ammonia existed in abundance.

    • I linked to papers with calculations suggesting panspermia is indeed feasible.

      • nick

        Where did they rebut the argument that a 1 million year half life per codon will destroy a viable organism in a few days? Where did they discover a viable organism in or from space?

      • You won’t even read the source I cite, but instead demand I offer a counter argument to your argument that isn’t even directed at the actual argument made?

      • jsalvatier

        nick’s first arguments seems directly aimed at the argument you made, and he’s asking you to point out where they deal with similar issues. Seems like a very reasonable request.

    • Matt S

      nick, this is a great comment. The panspermia hypothesis seems complicated and implausible, and it seems that in order to believe it there would need to have evidence in favor of it.
      Is there any actual evidence in favor of it? Or is the only evidence these theoretical models suggesting it might be possible theoretically?

    • celebrim

      Thanks for saving me the time by posting your comments.

      Assuming genesis as the great filter presents no real problems unless we assume also that genesis is a weak filter. If we assume genesis is a strong filter, then the number of planets which produce single celled life in our universe is exactly one and we are standing on it. We don’t see anything else in the heavens because there isn’t anything. We don’t see bacteria floating down from distant planets because it never happened anywhere else.

      And the Panspermia hypothesis is then likewise easily dismissed. We don’t even see evidence of bacteria having successfully breached the non-galactic distances between or own planet and its intrasystem neighbors. So there is no reason to suppose that bacteria survives the many millenia long trip through the cosmos to anywhere. It’s easier in fact to believe that genesis isn’t a strong filter than believe the vacuum of space isn’t.

      Besides which, there is no substantial evidence that there is an earth like planet within 80 lightyears, on the grounds that there is no evidence that there is a Sol like star within 80 ly (most yellow giants being notably unstable flare stars, and most other sorts being unsuitable in one way or the other). When you start computing the mass of biological material which would have to escape from the planet to seed the systems within 80 ly of us with living matter (none of which are likely to support earth-like life anyway), you quickly realize that its pretty ridiculous to imagine that much bacteria is being sloughed off by our planet and still living in the vacuum of space. Each cubic ly is 8.5E35 cubic kilometers. If all the bacteria in all the human bodies on the planet were beamed up still living into space, it would be a vapor so diffuse that we might never expect it to ever find a planet to land on much less a cozy spot to multiply when it got there.

    • halvorz

      “A codon half-life of 1 million years means that a viable organism is destroyed in a few days. The longest-lasting viable life forms are seeds and spores that have been evolved for long dormancy, and that requires sophisticated multi-cellular life.”

      The longest lasting viable life forms are bacterial endospores- no multicellular life required. Endospores certainly last longer than a few days:

      “Reliable reports exist of the recovery and revival of spores from environmental samples as old as 105 years (54, 81, 154), and there recently appeared a somewhat more controversial report that viable Bacillus sphaericus spores were recovered from the gut of a bee fossilized in Dominican amber for an estimated 25 to 40 million years (20)!”

      I’m still not convinced that said spores would be able to colonize a new planet. I suppose an experiment attempting to colonize various hypothetical abiotic planetary environments would be relatively easy to do- maybe it’s already been done.

  • haig

    A filtration process is only one possible solution to Fermi’s paradox, based solely on one particular interpretation of it, that the absence of evidence for advanced civilizations is evidence of their absence. This assumes that propagating across the galaxies is a natural progression of advanced civilizations, which is not that obvious. It seems almost certain that a civilization that is capable of doing so would have gone through a technological singularity, and so Fermi’s paradox can be interpreted into the question of whether a superintelligence not constrained by evolutionary goal systems would find it necessary or desirable to spend resources on sending low-entropy information across space.

    • This seems to be the most plausible explanation to me. The idea of a Singularity is gathering the aura of inevitability, so it should be logical to assume that other intelligent organisms would be likely to undergo such a transition as well. Interaction among entities will transpire over ever shorter distances as they will become increasingly impatient with message round trip times. From outside observers, these organisms would simply disappear from the observable universe and seem extinct.

      • Rob Crawford

        “Singularity” is a joke, the pseudo-intellectual equivalent of an economic bubble. Someone posited an imaginary graph and extrapolated that the imaginary line upon it continues towards an asymptote. No evidence, no real measurements, no science.

        I think I’ll start selling bumper stickers that say “In Case of Singularity, This Car Will Be Empty”.

  • If a filter is something that only a few succeed in passing through to (advancement), it seems to me that the Great Dying at the Permian-Triassic boundary should count as a filter. The Cretacious-Tertiary extinction of the dinosaurs and much of their world ought to count as well.

    I’m new to this site. Is there a specialized meaning to the word ‘filter’ here?

  • Ken

    “Yes it is possible that we are the very first, but that hypothesis is of course unlikely by default.”

    Why is this unlikely “by default”? Yes if you take it as an assumption that it is unlikely that we are the very first advanced civilization, then the rest of your post make sense.

    However, the evolution of human intelligence (or humanlike) is very small. If The earth were to reset, to the very beginning of life, I would place money, in fact my entire life savings, that humans would not evolve. It is a fluke of flukes that our race evolved. It is an arrogance to think that evolution converges to human like intelligence.

    In fact complexity is an evolutionary fluke. Most of the biomass on earth is the most basic type of life and it has had just as long to evolve as our animal line. I forget where I saw the picture, but looking at evolution as a graph you can set the root as the very first living organism. This leaf will form branches and ultimately end in leaves. The leafs of all the graphs have nothing close to human intelligence. Look at the non-human branches (look at the graph minus the branch that produced humans and primates) of the evolutionary tree. The leaf with the highest intelligence is the octopus. I don’t think this line will be building radio telescopes any time in the near evolutionary time (I don’t think this line will ever evolve to that point).

    To think that human like intelligence is inevitable is to anthropomorphize evolution.

  • Sardondi

    We’ve now reached such a level of intellectual sophistication that we no longer need religion to help us understand what we are and whence we came. We’re beyond simplistic ideas such as religious faith and belief, having graduated to science and fact.

    And our intellectual superiority has enabled us to finally grasp the true origins of life, which we now know was the result of immense time somehow acting on inert elements, after which that uncertain cosmic action known as “panspermia” spread life to our rock. Thence man.

    Thank G…uh, Gaia we’re above faith, and now deal only in scientific certainty.

  • Gregory Sullivan

    Here are some two alternative explanations:

    Transition to Low Visibility or Migration: A future technology might naturally produce a civilization that is invisible or nearly-invisible to civilizations at the current level of technology. The future technology might allow a migration to a location that is not easily sensed by our current technology. Alternatively the future technology may make the future civilization incomprehensible. In this case the civilization may be present, but we are unable to understand it. Perhaps it can be sensed but it appears to be random background noise. These possibilities may seem fantastical, but I think it is essential to allow freedom of speculation when contemplating future technologies.

    Virtual Reality and Unreliability of External Signals: As our computational resources grow and our control of sensory input improves the richness and fidelity of virtual reality systems grows. Even current VR systems embodied in video games, television, and movies are adequate to entertain most of the populace for hours every day. Future VR systems might include mechanisms for “suspension of disbelief”. When using such an advanced VR system ones beliefs about the external universe would be systematically falsified. Perhaps VR fanciers would enjoy a “simpler, primitive” time when only one advanced sentient life form was present in the universe.

    Robin Hanson has written on the topic VR systems and has valuable insights, so I suspect he has thought of this possibility.

    • nick

      That’s right, all of them, every single one of the many billions of individuals in each of the many billions of civilizations that supposedly exist, want nothing but to live in tiny unobservable holes, just like the dwarves and leprechauns. Not a single one follows the Malthusian/Darwinian laws and grows to the limits of its capabilities:

      Throughout the animal and vegetable kingdoms Nature has scattered the seeds of life abroad with the most profuse and liberal hand but has been comparatively sparing in the room and the nourishment necessary to rear them. The germs of existence contained in this earth if they could freely develop themselves would fill millions of worlds in the course of a few thousand years…Wherever therefore there is liberty the power of increase is exerted….

      (Robert Malthus from his Essay on Population)

      Having left such depressing science far behind, each alien Singularity has resulted in a pot of gold at the end of every nebula. But it’s guarded by very advanced crypto-ETIs that get very nasty if you find them.

      • Gregory Sullivan

        Thanks for your response Nick. Suppose that a future technology allows a type of space-time engineering which includes the ability to construct “basement universes” or sub-universes. The idea of filling up existing space must be reconceptualized when it is possible to construct carefully-tailored segments of space-time. The existing universe may be seen as an inferior region of habitation. Even with more than 6.8 billion humans on this planet there are vast regions with few human domiciles, e.g., desert biomes, tundra biomes, and the sea bottom.

        Perhaps using phrases like “live in tiny unobservable holes”, “dwarves and leprechauns”, and “pot of gold” reflects an overly-constrained vision of future possibilities. Contemplating future psychologies and motivations is admittedly difficult. An advanced civilization need not become a crypto-ETI or “very nasty” to prefer another area of manufactured space-time. An advanced cognitive apparatus may demand a special region.

        One might add another metaphor: Humans are not very visible to fish-life but it is not because humans live in tiny holes. Also, there are few pots of gold on the surface of the planet.

      • The serious points behind my joke about leprechauns are that these “ETI are common but cleverly hiding” theories are (1) extremely unlikely because they assume each of billions or trillions of replicating entities in billions or trillions of civilizations in the galaxies we can do detailed spectroscopy of move in lock-step, all of them hiding and none of them either shouting or expanding in Malthusian fashion in the visible universe, thus never producing any widespread surfaces or radiation transmissions with starkly artificial spectra, and worse (2) they are unfalsifiable. Your basement universe example introduces a third problem, namely that imaginative physics can explain away anything.

        But let’s focus on the first two problems which occur whether the physics is imaginative or otherwise. Why has absolutely no leprechaun ever rebelled and revealed himself and offered some poor human some gold from his pot? Why would absolutely no ETI ever start replicating in Malthusian fashion or shouting in a very visible way across the universe (very, very easy to do even at intergalactic distances)? The aliens may have such amazing technologies or be so “progressive” by some ideological definition that they all either choose (in contrast to all known life forms until at least the 19th century) not to expand in Malthusian fashion, or they can (through imaginative physics) expand their civilization in Malthusian fashion and still remain hidden. These theories require that every single subgroup chooses the same clever hiding approach in lock-step. Thus is the fabulous theory of common ETI (or of leprechauns right here on earth) be kept alive because it’s impossible to do any experiment that would show it to be wrong. Occam and Popper caution us to put far greater weight on the much simpler and falsifiable theory to explain the lack of engineered galactic surfaces: that ETI are extremely rare (far less than one per galaxy) or absent from the universe.

  • Chris T

    One almost certain great filter passed: the assimilation of mitochondria. Complex life cannot exist without it.

  • willis

    There are no filters at all. Life exists only on earth and its invention was absymally simple. Since evolution works on replicating organisms, a host of chemicals self-organized themselves into a brew that could become life, self-replicate, then evolve. Then, within that brew, certain chemicals organized themselves into a data-base containing just enough data for accurate, reliable self-replication. Then other chemicals organized themselves into the machinery with an instruction set that would know to seek out the data-base, retrieve replication data, then interpret and act on it to repicate itself. The whole engine then waited patiently for a bolt of lightning to hit close enough to power it up, but not so close as to blow it up. Right on cue, lightning appeared to accommodate it, bam, a billion years later, here we are! No filters, no nothing but the envitable outcome of chance.

  • Alex Flint

    Your link seems to suggest fertility as a candidate for a future filter. It may well be a worrying trend for humanity, but to me it seems very unlikely that it would be consistently devastating to every species that reaches human-level intelligence: it is caused by many incidental human traits that need not always coincide with intelligence.

  • Karl Hallowell

    willis, the problem here is that you’re assertion is just as unproven as anyone else’s. There’s some precondition that lead to our existence. The problem is that once life becomes space capable, it tends to spread throughout the galaxy. We don’t see evidence of such life. There are a variety of possible reasons: maybe we’re in the first cohort, maybe there’s filters that keep this from happening, maybe it’s out there and we don’t see it yet.

    My view on this is that there existed filters against life in the past that don’t exist now. The universe started with virtually all hydrogen and helium. We know of no interesting structures that can manifest with those two elements alone, unless one attempts to make some geared machine out of frozen hydrogen, perhaps. Stars of any sort didn’t start forming until 200 to 400 million years in with supernova occurring for the big stars shortly afterward.

    The galaxy would have also been a more hostile place for life in the past with an energetic galactic nucleus, possibly a quasar, plentiful supernova, galaxy collisions, etc.

    On planets with the necessary ingredients for life, we still have the issue of high background radiation levels from decay of radioactive elements produced during the supernova in addition to any asteroid/comet bombardments of the sort we’ve seen in our early Solar System.

    Even if these activities aren’t true filters, they would serve to delay the emergence of complex lifeforms. Our own understanding of such things in the Earth’s past is that large organisms fare worse with a sudden collapse or change in the ecosystem than small organisms. They also fare worse with exposure to radiation, though it’s mainly a result of terrestrial DNA being present throughout a organism’s body. Cancerous mutations tend to be proportional to the volume of the organism.


    1) Who made the great filter (or are the metaphysics of this also random)?
    2) You’re right, there is a big filter in the future

    3) If there was a Filter, there would be no liberals.

  • Ron

    Assuming a future-filter pretty much implies that it is impossible or extremely difficult to pass, even with prior expectation of it. After all, if others came before us to at least this step, they too will have had similar thoughts. If everybody so far hasn’t made it, what are the odds we will?

  • Lennart Regebro

    Possible big filters:

    1. Origin of life; seems lees and less likely the more we know of it, but I do not believe in interstellar panspermia as likely

    2. Multicellar life; you are right there.

    3. Continuing life-friendly environment; for example, if Mars had has a large moon, like Earth, it might still be volcanic, and might still have had a thicker atmosphere, and might therefore still have harboured ife. A big moon like ours seem to help.

    4. Land life; It’s may be difficult for sea-creatures to develop high technology. Again the moon, in creating tides, helped to make this happen.

    5. Self-awareness; very likely as a filter, we still know nothing of how it works.

    6. Interstellar travel; It’s likely to be impossible for anything biological. If we can put our consciousness into computers it may be possible for this non-biological entity to travel to other stars, but one question is why it would. Yeah, I’d like to go to other stars, maybe I would even like it if I was transferred to electronic circuitry, but on the other hand, as electronic circuitry living in a practically artificial world, wouldn’t I also like to just like to have my (now completely virtual) pleasure centre stimulated constantly instead?

    This last part seems to me to be the biggest reason that we have seen no interstellar civilization. The only ones that could build ones are robots, and they probably don’t want to. 🙂

  • Here’s another filter.

    Space is just ridiculously, enormously, hugely BIG. And empty.

    Interstellar travel is easy to imagine, but extremely difficult in practice, and not just from a technology standpoint.

    Space travel really sucks. Think about it. You’re alone in a ship, for a journey lasting eons, possibly – you’ll probably die on board – and by the time you get to your destination and then send a message to base and base sends a message back, everyone you knew on the planet will probably be dead (assuming short lifetimes).

    Even going to Mars is going to take a few years round-trip, and Mars is next door. The difference between travelling to Alpha Centuri and travelling to Mars is about 8,000 miles to 1 foot. Our best ship would take 100,000 years. One way.

    Would you want to take that journey?

    Again: much easier to imagine than it is to carry out in practice. There could be advanced planets out there, but doubtful there are advanced star systems.

  • Pingback: RealTime - Questions: "Are we having too much pump?"()

  • Karl Hallowell

    It’s also worth noting here that a filter need only delay not block the development of intelligent space faring life. I’d say in fact that a majority of filters are actually of this kind. 14 billion years sounds like a long time, but if there are many filters lopping off hundreds of millions of years at a time, then you can end up in a situation where it would be hard to develop an intelligence in the time frame.

    For example, in the early Solar System, asteroid impacts are thought to have been very common over the first billion years of the Solar System’s existence. It is reasonable to expect this caused hundreds of millions of years of delay in the establishment of life. If it resulted in a half billion year delay for star systems like ours, then that’s roughly 3% of the life of the universe consumed through this delay. The model of creation of the Solar System also appears to apply to most, if not all, “metal rich” (that is, have a lot of elements heavier than helium) star systems. So it’s possible that any star system in which it is possible to have formed life has experienced this phase in its formation.

    So as I see it, there are three types of filters:

    1) Early universe filters – circumstances that prevent life from forming in the early universe.

    2) Delay filters – slow down the development of life. Many of the filters we describe here are of this sort. For example, mass die-offs which aren’t extinction events (linked with supervolcanoes, plague, large scale nuclear war, collapse of society, etc) are of this form. They would generate centuries or millennia long delays in human colonization of the Solar System, but it’s likely that humanity would shortly return to the previous state of capability. Even extinction of humanity or extinction of multicellular life would still allow for a resurgence tens or hundreds of millions of years later.

    3) Existential filter – life (including manufactured life) ceases to exist. For example, life doesn’t escape Earth before the Sun enters its red giant phase. LHC destroys us and anything in our future light cone.

  • Re: “Unfortunately, the elegantly simple hypothesis that the great filter is mainly a big origin-of-life filter seems at odds with our best evidence. Why? Because if the spread-of-life step had the weakest possible associated filter, then life spreading must be easy. Over billions of years life could have spread to many star systems from its place of origin.”

    Just a moment! The great filter lies on the path between life forming, and life conquering the galaxy. The “spread of life” we expect on that path takes the form of technology-based space travel. The “spread of life” in the statement quoted above apparently refers to some other kind interstellar microbe diffusion. These types of “spreading” employ different mechanisms – and it seems quite possible for one type to be easy and the other type to be difficult.

    Difficulty of microbe diffusion may well be a barrier – to life spreading – but it isn’t on the path we are most interested in – the one that contains our future.

    I am sure that difficulty of teleportation also prevents life spreading – but do we *really* want to claim that as part of the great filter? I would say “no” – because it isn’t on the projected path in the first place. It is the same with microbe diffusion.

  • Re: “It seems to me that if the great filter is to consist of just one big step, the only plausible possibility is the development of multi-cellular life. All the steps before that one seem able to spread to other star systems via single-celled life hidden in dust.”

    …*if* you grant the mechanisms of panspermia. Anyway, besides what we know of panspermia difficulties, we have no evidence that this hasn’t happened. Zero. Maybe our solar system is spreading a trail of primitive bacteria through the universe. Or maybe we are emerging from a trail of bacteria being which was spewed from another solar system long ago – now long-since obliterated. We just don’t know.

  • Abelard Lindsey

    Guys, you want a “great filter”? I’ll give you one.

    Here’s mine:

    Read them carefully.

    The evolution of the Eukaryote is probably such a single, rare event that it only happened once in the Milky Way. That’s us. Then, it takes the Eukaryote to make enough Oxygen in the atmosphere such that we can breath it. Prokaryotes can’t do this.

    Even if we get the FTL or wormholes, we still have to terraform whatever planets we go to out there so that we can live on them and relax on the beaches without having to use breathing gear just like someone with COPD.

    I really hope I’m wrong. Not because I like aliens (I don’t), but because I want to get the hell out of here and I would like some decent place to go to.

    My ideal universe is where there’s lots of Earth-like planets (ones that I can actually breath the air and hang out on the beach) but where there are no intelligent aliens. I want the galaxy for myself and the rest of you.

    Hamilton’s commonwealth (Pandora’s Star, Judas Unchained), but without the aliens, is my ideal universe.

    • Buck Farmer

      How probable is a eukaryote that can survive (perhaps dormant) a transplanetary or interstellar journey?

      If it’s non-negligible then this doesn’t explain the radio silence as you still get panspermia and almost no filters preventing development of civilization.

      That is unless you assume the Earth is singular…which leads to sort of empty arguments.

  • Tom Cone

    Another point to keep in mind: Even assuming that we could identify intelligent life if we saw it, we can’t actually say there is no intelligent life in the universe now. Rather, given the lightspeed lag with which we observe the distant universe, we must say there was no intelligent life as of X years ago.

    E.g., if we’re observing a galaxy 100 million light years away, we can only say there was no intelligent life there as of 100 million years ago. Maybe right now it’s got intelligent life up the wazoo.

  • Tom Cone

    Another point that’s typically neglected in discussion of this topic is the opportunity costs of very advanced beings. E.g., “Advanced races could re-configure the galaxies themselves. Why don’t we observe that?” (How do we know we’re not observing it, for one thing?) But suppose a civilization is so advanced that it could reshape its home galaxy. Yeah, but there are other uses of their resources that might be much better, precisely because they’re so advanced.

    Suppose I have the technological capacity to turn a desert into a forest in 10 years. But what if I can simply reconfigure myself so that I’m perfectly suited for a desert in 1 day?

    Analogously, expending time and other resources reshaping galaxies is not necessarily worthwhile for a very advanced civilization with tons of other options.

  • Pingback: Overcoming Bias : Bad News: Kant Plugs Bets()

  • The great filter is ahead of us.

    To produce a galaxy spanning civilization, the problem of xenophobia has to be solved. Unless it is, then as populations on different planets genetically drift apart (as they must), and they become different species, they will exhibit xenophobia and make genocidal war on each other.

    It only takes one new species exhibiting genocidal xenophobia to wipe out all the non-xenophobic species. Then as that xenophobic species drifts and forms new species, some of them will be xenophobic too.

    The patriarchal alpha male phenotype has to express the equivalent of xenophobia to be able to compete against other males for females (or for anything). The alpha male has to take resources for himself and deny them to other males or he isn’t the alpha male. When those taken resources are used to take and maintain more resources, the alpha male has to exhibit xenophobia to deny resources to others because those resources might be used against him.

    If we do find a galaxy spanning civilization, it won’t be a patriarchy, or any other kind of oligarchy. Unless we find a way to permanently and irreversibly shape human society into a non-oligarchy, we will filter ourselves out.

    • Hedonic Treader

      Controlled evolution is probably the most likely solution to this type of problem, starting with a singleton on earth which cares about the utility function of all sentient life. Interstellar colonization is undertaken only if it can be designed so that free evolution is impossible.

      This can be done by creating colonziation systems such as self-replicating spacecraft in a way that effectively prevents spontaneous mutation. Colonization would then follow a coordinated long-term plan that focuses on keeping peace and quality-control of all existing sentient life. Xenophobia and free random speciation would not be factors in such a plan.

      Similarly, the singleton could control evolution on earth in order to minimize suffering and prevent a hedonistic “race to the bottom”. Such a singleton could come about via integration of human brains and specialized AI modules through wireless communication implants connecting to an enhanced version of the internet, enabling extremely efficient global coordination, quality-control of all sentient minds, sustainable management of the ecosphere (without wildlife suffering), control of populations and genomes, longevity and universal basic rights of all existing persons, and complete global transparency of all agency to prevent effective attempts to undermine the system.

      I would expect such an approach to be less efficient than open darwinian evolution in terms of growth and resource efficiency, but far preferable from an ethical hedonistic perspective, especially in the long run.

      • Jeffrey Soreff

        Very much agreed – we certainly need some form of control to avoid a “race to the bottom”, to avoid a malthusian dystopia – particularly if ems are developed, and the timescale for selection abruptly drops from decades to minutes.

      • If a civilization such as you describe, with “controlled evolution” were to meet a “wild-type” civilization (such as ours), their only recourse would be to destroy and sterilize it. That is the only approach such a civilization can have to things that it does not understand and which it does not control. That is exactly what xenophobia is, the compulsion to hate and destroy that which one does not understand and what one does not control.

        I don’t think that many humans today would allow that sort of control, i.e. would voluntarily submit themselves to it. To institute that type of control, those individuals would have to be destroyed first. How are the initial stages of that transformation distinguishable from any other form of absolute life or death dictatorial centralized control?

        What would compel those who first institute the destruction of those who would not submit to the “quality-control of all sentient minds” (i.e. mind control far beyond what 1984 and Brave New World ever imagined) to submit to it themselves? In other words, once they have mind control over everyone else, why would they surrender themselves to mind control?

        The problem is in a top-down control scheme. Every top-down control scheme has to prioritize its own survival over all other things. It has to destroy what ever it does not understand because what is not understood might be or become an existential threat.

      • Hedonic Treader

        daedalus2u, I acknowledge all the problems you describe. By “quality control” of minds, I meant something like telepathy implants + a metric to measure well-being, request immediate help, cognitive pain relieve, fix mental disorders etc.

        But you’re right, due to the top-down approach, and due to the darwinian nature of memetics, mind control would probably involve thought control too at some point. OTOH, the singleton would still need to be a very complex functional decision-maker, and without singular super-AI, that would require the global collective of transhuman minds to think rationally (and “slave hive” collectives aren’t rational decision-makers).

        Despite the critical points you raise about this scenario, it seems all possible alternatives are equally gloomy. Roughly, I see only these three:

        1) Singleton, then controlled evolution (Inefficient, existential risk in building it, essentially totalitarian in nature)
        2) Open darwinism (Unpredictable outcomes, high expected value of suffering, torture, wars, massive desire frustration, “race to the bottom” dynamics; only with gradients solution as improbable, but possible remedy)
        3) Collapse/extinction.

        Now calculate expected ethical value. There is no really good solution that I can see. If you can point me to one, I’m very interested.

  • Hedoniic Treader

    I gave some more thought to the possibility of a “mind collective” singleton that is as non-totalitarian as possible (with regards to personal freedoms) while enforcing the rules and control necessary to prevent the ramifications of open evolution.

    Imagine a world in which people’s minds are connected to each other and to the internet at all times for coordination and communication purposes. Add to this biological immortality and safety back-ups of each person’s connectome. In such a world, everyone would have a potentially unlimited personal life span and therefore a strong self-interest in creating and maintaining a peaceful, sustainable world.

    Reproduction would no longer be necessary to maintain population – it would only be needed to replace lives that decide to end their existence or that happen to be completely destroyed by accidents (including backup). Additional reproduction could be used to expand population to whatever limit will be sustainable, but this could be done as a collective decision, not as a decision of individuals like today (hence no selection pressure of individuals to reproduce).

    Traditional families and their generations would be replaced by a global collective of immortal individuals deciding what number of additional immortal individuals should come into the world. Defectors to such rules would be detected quickly (due to global integration and communication of all minds) and punished accordingly.

    Uploads, dangerous AI projects or other existential threats could be outlawed. Alternatively, existing persons could upload, but copies of uploads would be outlawed unless the collective decides to allow a given number of them. People (or ems) who prefer to reproduce could be given the option to concede biological immortality and\or accept scheduled death to keep population stable and prevent selection pressure.

    Alternatively, the nature of reproduced individuals could be influenced by the collective so that no selection pressure results in undesirable evolutionary paths (from the collective’s perspective). Population numbers could end up lower, or significantly higher than today while sustainability and evolutionary stability is guaranteed. Memetic evolution would run slower due to longevity (no generational shifts).

    At the same time, all existing persons could enjoy strongly enforced personal rights and proper above-subsistence lifestyles. Digital post-scarcity could provide an abundance of VR experiences and other types of creative work whose creation does not rely on wages because people have free time on their hands and the products can be copied without incremental costs.

    This scenario is still improbable given how different people’s world views and preferences are currently. But imagine the memetic drive for convergence in a globally connected world like this, combined with the self-interest of immortal individuals for sustainability and security. Collective personal self-interest has always gone a long way, and this could be the key to enforce the rudimentary ground rules to prevent undesirable evolutionary shifts, wars, bioterrorism etc. in a world of globally connected immortals.

    tl;dr: If everybody’s immortal and globally connected, maybe there can be the democratic decision to centrally plan and control reproduction, and therefore keep darwinian dynamics in check.

  • I am quite sure that the entire idea of a hyper strong AI that can increase its own intelligence and so become exponentially more intelligent is wrong. The problem isn’t that electronic systems can’t support intelligence, I am quite sure that they can. The problem is how does the AI recognize an intelligence greater than its own and then adopt changes to produce such an intelligence?

    The only way to recognize something is via pattern recognition. To recognize a pattern, a pattern recognition system must have a facsimile of that pattern to compare against. How can an AI tell that parameters that do not map onto its existing intelligence facsimile are superior intelligence and are not madness? I don’t think there is any way to distinguish between them, even in principle.

    I am not worried about existential threats to future human interstellar civilizations. I am worried about existential threats right now, global warming being a gigantic one, environmental endocrine and genome disrupting chemicals is another, global conflict due to mass starvation is another. The reason all of these problems are happening is because individuals prioritize their own wants ahead of other peoples needs, and enormously prioritize those wants against the needs of future generations.

    They are prioritizing their wants ahead of solving human race existential threats. In any kind of competition, the group that devotes zero effort to solving human race existential threats will have an advantage over those who devote some effort. The easiest way to ignore existential threats is to be unable to perceive them. Then, being ignorant of existential threats makes one a superior competitor to those who are aware of existential threats. Unless we find some way to change that dynamic, those who are unaware or choose to ignore existential threats will be the most successful, and so will be those at the top in charge of any top-down power structure. Looking at current politics I am not encouraged.

  • People who become addicted to drugs usually initially choose to take them. Watch out! That’s what i want to say here.

  • Pingback: Overcoming Bias : Hurry Or Delay Ems?()

  • Pingback: Overcoming Bias : Our Quiet Galaxy()

  • Evan Grantham-Brown

    If life can so easily spread between stars, then it should have no trouble at all going from planet to planet. By this logic, our own Solar System ought to be teeming with Earth-descended life. Yet we see no sign that this is the case. Granted, our explorations have been limited so far, but we have yet to find the slightest hint of life on any other planet or moon in the Solar System.

    The obvious conclusion is that panspermia is much harder than you make it out to be. Abiogenesis remains a plausible candidate for the Filter.