Hope For A Lumpy Filter

The great filter is the sum total of all of the obstacles that stand in the way of a simple dead planet (or similar sized material) proceeding to give rise to a cosmologically visible civilization. As there are 280 stars in the observable universe, and 260 within a billion light years, a simple dead planet faces at least roughly 60 to 80 factors of two obstacles to birthing a visible civilization within 13 billion years. If there is panspermia, i.e., a spreading of life at some earlier stage, the other obstacles must be even larger by the panspermia life-spreading factor.

We know of a great many possible candidate filters, both in our past and in our future. The total filter could be smooth, i.e. spread out relatively evenly among all of these candidates, or it could be lumpy, i.e., concentrated in only one or a few of these candidates. It turns out that we should hope for the filter to be lumpy.

For example, imagine that there are 15 plausible filter candidates, 10 in our past and 5 in our future. If the filter is maximally smooth, then given 60 total factors of two, each candidate would have four factors of two, leaving twenty in our future, for a net chance for us now of making it through the rest of the filter of only one in a million. On the other hand, if the filter is maximally lumpy, and all concentrated in only one random candidate, then we have a 2/3 chance of facing no filter at all in our future. Thus a lumpy filter gives us a much better chance of making it.

For “try-try” filters, a system can keep trying over and over until it succeeds. If a set of try-try steps must all succeed within the window of life on Earth, then the actual times to complete each step must be drawn from the same distribution, and so take similar times. The time remaining after the last step must also be drawn from a similar distribution.

A year ago I reported on a new study estimating that 1.75 to 3.25 billion years remains for life on Earth. This is a long time, and implies that there can’t be many prior try-try filter steps within the history of life on Earth. Only one or two, and none in the last half billion years. This suggests that the try-try part of the great filter is relatively lumpy, at least for the parts that have and will take place on Earth. Which according to the analysis above is good news.

Of course there can be other kinds of filter steps. For example, perhaps life has to hit on the right sort of genetic code right from the start; if life hits on the wrong code, life using that code will entrench itself too strongly to let the right sort of life take over. These sort of filter steps need not be roughly evenly distributed in time, and so timing data doesn’t say much about how lumpy or uniform are those steps.

It is nice to have some good news. Though I should also remind you of the bad news that anthropic analysis suggests that selection effects make future filters more likely than you would have otherwise thought.

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

    Clever! But doesn’t the subjective (knowledge-dependent) nature of this smoothness definition make it a bit hard to reason about? Maybe next year there’s going to be a huge government program for futures research, and scientists invent 50 new filter candidates in our future. Or vice versa with paleontology.

    Admittedly, only for pretty extreme, artificial-looking distributions for filter timings can it happen that lumpy gives us worse survival chances than smooth. But it can happen. Maybe because of some constraint we happen to live in such an artificial-looking universe. (…artificial-looking probability distribution of universes, to be more precise.)


    “As there are 280 stars in the observable universe, and 260 within a billion light years, a simple dead planet faces at least roughly 60 to 80 factors of two obstacles to birthing a visible civilization within about ten billion years.”

    Are you suggesting we know for a fact that there are no alien civilizations within a billion light years? If effective FTL travel is not possible then intergalactic colonization is pretty much impossible, or at least, very, very slow. If some form of effective FTL is possible it may still be too slow to cover hundreds of millions of lightyears between the alien civilization’s emergence (which might very well not have been that long ago since second generation solar systems haven’t been around that long compared to the time it took intelligent life to evolve on Earth) and today and intergalactic colonization may still be prohibitevely costly (and perhaps seen as unnecessary by a civilization that already controls a whole galaxy).

    Great point about a lumpy filter though, nice to see someone doing the math on scenarios like that.

    • Cahokia

      Bear in mind – visible civilizations do not equate to civilizations capable of FTL travel.

      • IMASBA

        Where did I say that they did equate? But remember that a civilization 100 million lightyears away from us has to have been around for at least 100 million lightyears for us to receive any non-FTL message from them and I don’t even want to know what the power requirements would be for a signal that can still be read 100 million lightyears away (which would also have to just happen to be aimed right at us because if they just broadcasted in all directions a signal that powerful would disrupt their own day-to-day communication systems.)

      • david condon

        Given that our technological growth rate leads to greater safety, the chances of a mass extinction decrease over time. If we live for a hundred thousand years, it is quite unlikely we would go extinct thereafter over any time frame. The lack of aliens indicates we will go extinct. Technology’s impact on safety indicates this will happen soon.

      • IMASBA

        Not sure why everyone keeps assuming alien life had to have existed long before we did. It is entirely plausible alien civilizations from other galaxies simply have not existed long enough to reach us. It’s more logical to assume aliens evolved around the same time we did since they need heavy chemical elements (which were not available in the first solar systems) and time to evolve just as much as we do. Sure there could be some a few million years ahead of us (and if they live in another galaxy it would take them millions of years to reach us without FTL) but the fact that we don’t find remnants or descendents of multi billion-year old civilizations everywhere we look shouldn’t be taken as a sign that we are doomed or alone in the universe.

        Also, technology increases civilian safety but at the same time leads to bigger bombs. The extinction of a highly advanced civilization will depend on how stable their social organization is (having no, or only limited, ritualized warfare probably helps) and the more socially stable they are the more likely they are to be nice enough to instruct their seed ships to leave alone those planets that already have life on them.

      • truth_machine

        That’s hogwash, though the known facts do indicate a short-term demise.

    • truth_machine

      “If effective FTL travel is not possible”

      It’s not, despite what you may have seen on Star Trek.

  • arch1

    “Thus a lumpy filter gives us a much better chance of making it.”
    This posting seems to regard passing the Great Filter as essential for long term civilizational success. I am guessing that this in turn is based on the notion that any successful civilization will inevitably turn ever increasing amounts of conventional mass-energy to its ends – in the process inevitably becoming visible to (shortly before destroying? hmm) incidental rabble such as ourselves.
    Is that your assumption, and if so a) what is your certainty c of its truth, b) shouldn’t your analysis incorporate a ‘probability cutoff’ of something like 1-c to account for the possibility you are wrong – i.e. that prosperity doesn’t necessarily imply visibility?

    • I’m not arguing for or against goals here. I’m trying to draw inferences from our data of no visible civilizations.

  • truth_machine

    Yeah, let’s just ignore the known facts about global warming and ocean acidification … no need to look in the horse’s mouth when we can just sit around and theorize about how many teeth it has.

    • IMASBA

      Climate change will not wipe out the human race, no matter how bad it gets there will always be places left where millions of humans can survive and eventually thrive.

      • Curt Adams

        I agree there will be humans, but in the worst current estimates humans would only survive on high mountains. That could well knock out our chances of ever being a space-faring civilization, which would be a valid part of Robin’s Great Filter. I agree it’s hard to see climate change as the main lump – seems *some* worlds would manage OK with all fossil fuels burnt, and *some* would get emissions under control.

      • IMASBA

        “in the worst current estimates humans would only survive on high mountains.”

        Nah, that’s ridiculous. There are vast areas of the world that lie, say 1km, above current sea level (like Iran, most of Mexico and the US state of Colorado). There’s not even enough water on the planet to flood such areas. Even in the warmest periods of the Earth’s history there were still vast landmasses left. Given that humans surviving in such a world can simply tap into existing knowledge space travel would only be delayed, probably by only a century or two.

        I agree with you that assuming other worlds would get in big trouble for burning their fossil fuels is nonsensical, their surfaces and climate will be very different, not to mention there could well be many aquatic intelligent species.

      • Curt Adams

        It’s not the water, it’s the heat. Worst estimates are a 30 degree F increase in temperature along with a substantial increase in humidity. Along with the loss of polar icecaps and resulting high summer temperatures in the arctic, that means any sea level area is going to have wet bulb temperatures well above 98.6F on a regular basis, meaning even somebody outside in the shade, with a breeze and unlimited water will die of heatstroke. This is a much harsher climate than anywhere on the planet currently. Such conditions happen today only on a very occasional basis in the hottest deserts.

        Incidentally, it will be much hotter than any previous period with advanced life on the planet as a) the sun is gradually getting hotter and b) we’ve never had the sequestered carbon in oil, coal, and geological gas released before, just the methane hydrates.

      • IMASBA

        It won’t be worse than previous periods: if it gets really bad it will probably be comparable to the Permian era. That’s hotter and wetter than today, with twice as much CO2 in the air, but still with lush vegetation and abundant animal life outside of the equatorial zone. Humans were built for hot climates, so we’ll do fine, we might not be able to support 7 billion of us, but here can still be enough of us left to maintain advanced technology.

        If climate change kills off 2 billion people it would be the biggest disaster ever, but it still wouldn’t prevent the numerous survivors from going into space one day.

  • Friendly-HI

    What seems like a reasonable estimate for the likelihood that a space-colonizing civilization can avoid being detected by even the most sophisticated instruments at the cutting edge of what is physically possible if it chooses so? Radio waves don’t propagate indefinitely, eventually they fade into the background noise, so us not detecting anyone isn’t really strong evidence for nobody being out there.

    To me it seems like the best strategy is to avoid being seen and trying to conceal your true size while you’re colonizing your galaxy. There might be older civilizations out there that will come to stomp you out before you spread too far and waste “their resources” on your utility function instead of theirs. There may even be failed AI’s afoot churning out paperclips at the limit of what is physically possible and they might reason that ending your existence is the reasonable thing to do. With the advent of strong AI it is not clear to me what benefit any contact with another civilization might gain you. If your strong AI is doing its job properly you won’t gain much valuable additional knowledge about things that matter to you from meeting anyone “out there”. In fact they are competitors by default and the only thing that might keep them from curb-stomping you is the outcome of their cost-benefit-risk analysis of a war. Meaning it makes sense to conceal the true size and power of your civilization, at least initially.

    • IMASBA

      “Radio waves don’t propagate indefinitely, eventually they fade into the background noise, so us not detecting anyone isn’t really strong evidence for nobody being out there.”

      Exactly: there’s no way we can detect even a civilization on the other side of our own galaxy, let alone in another galaxy (and there are quite a lot of galaxies within that billion light year radius).

      “There might be older civilizations out there that will come to stomp you out before you spread too far”

      Well, actually there is a significant home advantage in a hostile-seed-ship scenario. The seed ship is limited in mass, for propulsion reasons, while the defending civilization has a lot of local resources and industry, they just have to stop the seed ship before it can set up an industrial base in the solar system.

      • Friendly-HI

        I was rather thinking in terms of us being the seed-ship colonizing another system while being detected by another civilization that has a very significant time and resource head-start on us, choosing to take us on anyway.

        Let’s explore the premise that “technological advancement” is not going differ between our own and a much older civilization with a head-start over us. I find this plausible as a consequence of strong AI which should enable both civilizations to operate at the cutting edge of what is physically possible.

        Let’s turn the tables and imagine in one million years we have the time- and resource-advantage on a close-by more recent civilization that is on par with us in terms of technological advancement but lags behind significantly in resources. Considering that they too want to grab as much of the universe as they can, it may still make a lot of sense for us to fight a war that has a win-to-loss ratio of 1 to 100 in their favor. The universe is big and it will stick around for a long time, so taking huge losses in the short-term may still be worth it in terms of future potential. An even better alternative could be to surround them entirely in a “bubble” and prohibit them from expanding any further, keeping the peace while being able to grab the resources they would have otherwise settled on first in the future – and they get to live. A third alternative might be merging together, surely depending on how different and (un)reconcilable our utility-functions are.

        And that’s why I think it makes a lot of sense to keep your size hidden, because it seems very plausible to me that you can’t size up your opponent just by looking at their technological advancement, because that’s probably not where competing civilizations will differ much.

      • IMASBA

        I think only the containment scenario would make sense, although a merging could in some situations be negotiated. War is too costly because a siege could last for millions of years (it’s like using pea shooters against a solid rock fortress), but if the defending world were to attempt a counter offensive they’d run into the same problem, so a stalemate is the natural outcome (this of course assumes there’s no wormhole connecting the two parties).

        As to technology, yes, I agree that it will probably only take a few centuries to catch up to even the oldest civilizations, so to learn everything that can be learned, and therefore there won’t be a big technological gap.

  • matthoboken

    Your implicit assumption in discussing this topic always seems to be that technologically advanced civilizations would be interested in being “cosmologically visible.” Why? There’s nothing to learn from traveling the stars that can’t be learned from simulations. (This point holds even apart from your ancillary assumption that technologically advanced civilizations would have the wayfaring instincts of humans.) Nor is there any clear reason to think that the energy or resource demands involved in exploration via simulation would be visible at great distances.

    • My assumption is just that there is a substantial chance that a tech advanced civ would become visible. The smaller that chance, the larger that last filter step. There are many reasons that civs might do things besides what they might learn as a result.

      • IMASBA

        Visibility will most likely be limited to a few light years beyond their actual physical presence. There’s now way in hell we’d be able to detect a civilization in even the nearest galaxy (actually even a civilization 1000 light years away in our own galaxy would most likely be invisible to us) unless they sent a physical object to our galaxy. And there are a lot of galaxies within a radius of 1 billion light years… I just cannot see how this would not bug you when thinking about the great filter.

      • I’m not talking about visibility of a tech civ when it first appears. I’m talking about the visibility of its descendants after a billion years of continued evolution.

      • Dan Browne

        Actually Penn State researchers are looking exactly for large artifacts in other galaxies and they have some candidates that need explaining away.


      • Of course. That has been the case for many decades; we see new things, we puzzle over them, and eventually explain them satisfactorily in terms of dead stuff.

      • Dan Browne

        It will, nevertheless be extremely interesting to find out exactly what large solar system scale infra-red emitting objects could be.

      • Thanks for the link – I made it a center of my next post.

      • Dan Browne

        You’re very welcome. I very much enjoy your blog so I’m glad to be able to contribute.

      • IMASBA

        The same rules still apply to a billion year old civilization: they need to have a physical presence very near to Earth for us to detect them. I know your assumption is that such civilizations would (successfully) send self-replicating probes to galaxies other than the one they evolved in, everything kind of stands or falls with that (and with the probability of intelligent life elsewhere evolving a long time before us).

  • Dan Browne

    Here’s an arxiv link showing that gamma rays seem to make 90% of the galaxy uninhabitable. That’s part of the fermi paradox right there and a massive, massive filter for complex life (bacteria would likely survive it). Perhaps it already happened once before on Earth in the past. http://arxiv.org/abs/1409.2506

    • IMASBA

      It could also mean most civilizations are aquatic…

      • Dan Browne

        Yeah. Good point.

      • IMASBA

        Attack of the space octopi 😉

        Actually this might add a new dimension to the debate: it may be that it typically takes aquatic lifeforms much longer to evolve into a spacefaring civilization (they can’t invent fire early on in their development for example) and therefore there simply are no billion year old civilizations anywhere near the Milky Way galaxy. There then might be some that are a few million years older than us but haven’t yet conquered local space.

      • Dan Browne

        Yeah. Running with it a little… if it’s not intelligent life that’s difficult but intelligent land-based-life that’s difficult, that definitely has implications for the projected quantities of spacefaring civilizations. We ourselves have dolphins and whales, both of which have large, evolved brains, but not necessarily intelligence much greater than that of a dog.