Stock Vs. Flow War

When our farmer ancestors warred, they often went about as far as they could to apply all available resources to their war efforts. This included converting plowshares into swords, ships into navies, farmers into soldiers, granaries into soldiers on the move, good will into allies, and cash into foreign purchases. When wars went long and badly, such resources were often quite depleted by the end. Yet warring farmers only rarely went extinct. Why?

The distinction between stock and flow is a basic one in engineering and finance. Stocks allow flows. A granary is a stock, and it can produce a flow of grain to eat, but that flow will end if the stock is not sufficiently replenished with every harvest. A person is a stock, which can produce work every week, but to make that last we need to create and train new people. Many kinds of stocks have limits on the flows they can produce. While you might be able to pull grain from a granary as fast as you like, you can only pull one hour of work from a worker per hour.

Natural limits on the flows that our stocks can produce have in the past limited the destructiveness of war. Even when war burned the crops, knocked down stone buildings, and killed most of the people, farmland usually bounced back in a few years, and human and animal populations could grow back in a few generations. Stones were restacked to make new buildings. The key long-term stocks of tech and culture were preserved, allowing for a quick rebuilding of previous professions, towns, and trade routes.

Future technologies are likely to have weaker limits on the conversion of stocks into flows. When we have more fishing boats we can more quickly deplete the stock of fish. Instead of water wheels that must wait for water to come down a stream, we make dams that give us water when we want. When we tap oil wells instead of killing whales for oil, the rate at which we can extract oil grows with the size and number of our wells. Eventually we may tap the sun itself not just by basking in its sunlight, but by uplifting its material and running more intense fusion reactors.

Our stronger abilities to turn stocks into flows can be great in peacetime, but they are problematic in wartime. Yes, the side with stronger abilities gains an advantage in war, but after a fierce war the stocks will be lower. Thus improving technology is making war more destructive, not just by blowing up more with each bomb, but by allowing more resources to be tapped more quickly to support war efforts.

This is another way of saying what I was trying to say in my last post: improving tech can make war more destructive, increasing the risk of extinction via war. When local nature was a key stock, diminishing returns in extracting resources from nature limited how much we could destroy during total war. In contrast, when resources can be extracted as fast and easy as grain from a granary, war is more likely to take nearly all of the resources.

Future civilization should make resources more accessible, not just to extract more kinds of slow flows, but also to extract fast flows more cheaply. While this will make it easier to flexibly use such stocks in peacetime, it also suggests a faster depletion of stocks during total war. Only the stocks that cannot be depleted, like technology and culture, may remain. And once the sun is available as a rapidly depletable resource, it may not take many total wars to deplete it.

This seems to me our most likely future great filter, and thus extinction risk. War becomes increasingly destructive, erasing stocks that are not fully replenished between wars, and often taking us to the edge of a small fragile population that could be further reduced by other disasters. And if the dominant minds and cultures speed up substantially, as I expect, that might speed up the cycle of war, allowing less time to recover between total wars.

GD Star Rating
loading...
Tagged as: , ,
Trackback URL:
  • http://www.gwern.net/ gwern

    Improving tech may also mean wars do not reach the same tempo of destructiveness because too much wealth is lost when wars go further or the increasing complexity of the economy means the damage ends effective resistance faster or military imbalances accumulate faster (stone-age warfare if you have 10 more fighters: you have 10 more fighters; modern warfare if you have 10 better fighters: you destroy the opposing air-defense system and proceed to bomb unopposed). Tech may make relative flows faster, but that’s never the only dynamic at play…

    Do we see your claims about stock/flow in historical warfare? Is the devastation of WWII in Germany far worse than the wasteland of the Thirty Years’ War 300 years and most of an industrial revolution before? And the Thirty Years’ War in turn far worse than the Germanic tribes left behind of the Roman empire? Or further afield, was the devastation of Persia after the Mongols wrecked the country and its delicate water systems really less problematic than the much higher-tech Iran-Iraq total war, featuring human-waves to clear minefields, 700 years later? Was WWII worse for the US engaged in total warfare than the American Revolution with its 162k casualties/fatalities/refugees out of a population of ~3m? Shouldn’t we see a generalized trend of wars causing ever higher casualty percentages and immiserating countries ever more greatly, rather than the opposite (comparing WWII against the Yanomano eg) if these claims about flows are correct?

    • http://overcomingbias.com RobinHanson

      I am NOT proposing a one factor model to explain most variation in war damage. I am identifying a factor that matters and that we can reasonably expect to robustly move in a particular direction into the distant future. If you want to test the effects of this factor in a few cases you’ll have to find cases that more clearly differ mainly in this factor.

      • http://www.gwern.net/ gwern

        If it matters and you have the direction of the effect right in the first place (I gave 3 equally plausible opposing effects), then why don’t we see any trends like predicted? Your stock vs flow paradigm makes a clear prediction about trends which is falsified by pretty much all long-term data series about war costs and casualties I’ve seen (like Pinker’s stuff).

        Or is everything in OP unfalsifiable?

      • http://overcomingbias.com RobinHanson

        Humans are a stock, and we have yet to find better techs for converting them into flows. In fact, they live longer and so are larger stocks, with smaller % flows. So my account does NOT predict higher casualties so far.

      • http://www.gwern.net/ gwern

        You *already* explicitly predicted higher casualties!

        > *Our stronger abilities to turn stocks into flows can be great in
        peacetime, but they are problematic with war*. Yes, the side with
        stronger abilities gains an advantage in war, *but after a fierce war the
        stocks will be lower.*

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        You *already* explicitly predicted higher casualties!

        No, he did not retrodict this for the past–he only predicted it for the future.

        [What precisely is the historical trend Robin expounds? Is it (as I supposed above) that technology increases the flow/stocks ratio? Apparently not!

        [So, he’s apparently saying that em technology, in particular, will raise this ratio. For the first time, “we” will be able to increase the flow of a human’s labor power. But “we” will also dramatically increase the stock of “humans.”

        [Is there any historical evidence for any trend in the flow/stock ratio? If not, why should this technology be any different? That it can increase a flow for the first time doesn’t speak to the ratio when stocks are also expected to increase. If the human stocks increase, as I think Robin expects, by a factor of a thousand, will flows increase to the same extent? Bottlenecks with the operation of the robots would seem to preclude speed up of that magnitude.

        [The most impressive result of ems is how many humans it allows. This might well overshadow the increased flows. If stocks are predicted to increase faster than flows, isn’t Robin’s premise contradicted? Shouldn’t the prediction–to the extent one is possible–be exactly the opposite?]

  • Don Reba

    we make damns that give us water

    dams 🙂

    • IMASBA

      But we can also render water undrinkable for decades through nuclear or biological weapons. We can basically salt the entire Earth.

  • http://juridicalcoherence.blogspot.com/ Stephen Diamond

    You seem to argue from the premise that technology creates faster flows to the conclusion that stocks will be more quickly depleted. But I don’t see any argument that technology doesn’t equally mean the creation of greater stocks.

    [Consider the Internet as a stock, and comparative it to the size of one’s hard drive in pre-Internet days. The stock of information available to folks has increased far faster than our ability to access it–to turn it into flow.]

  • IMASBA

    “When our farmer ancestors warred, they often went about as far as they could to apply all available resources to their war efforts.”

    Robin, I think this was actually pretty rare: most warfare was seasonal and could be sustained almost indefinitely, even empires engaged in wars that lasted decades. However total war did occur in some emergencies when the stakes were particularly high for the common man (the Caledonians’ successful efforts to avert conquest by the Romans, or the unsuccessful last stand of the Southern Song against the Mongols).

    In more recent times your point holds: total warfare has become common for nation states going to war with each other. Rapid depletion of resources would indeed be an existential threat to a technologically advanced civilization fighting a civil war. Even our current world knows this threat (weapons of mass destructions that can easily decimate the global population, destroy food and water sources, or both).

    • IMASBA

      I am skeptical though that this explains the great filter fully. Through chance alone there should be a few civilizations who are peaceful (possibly after altering their own nature through mind-reprogramming or genetic manipulation) or have managed to not descend into civil war*. These civilizations should have an easy time expanding if the more aggressive civilizations tend to destroy themselves early on.

      * perhaps humanity is one of these: we’ve managed to not get into nuclear wars for quite a while now and our forager period proves we can live peacefully in the right environment, perhaps a future Earth can combine advanced technology with channeling forager instincts and ideas.

      • http://overcomingbias.com RobinHanson

        I’m not remotely proposing this to explain the great filter “fully.”

      • hypnosifl

        Just a note, your earlier comment links to a paper predicting the time for the Earth to remain in the habitable zone of the Sun where oceans can still exist, but the time left for complex multicellular life is probably significantly less, somewhere between 500 million and 1 billion years, because of a long-term trend of decreasing carbon dioxide which will doom photosynthetic plants. This is predicted due to the increasing lumosity of the Sun, which increases the rate of weathering of rocks that can trap atmospheric carbon dioxide in carbonate. Plants with C3 photosynthesis, which includes all trees except bamboo, need higher CO2 levels so they will die first, while plants with C4 photosynthesis like grasses and bamboo may survive a few hundred million years longer; but the first extinction may decrease biodiversity enough so that if intelligent life hadn’t evolved by then, it would be much less likely to in the time remaining. And the extinction of C3 plants is only projected to happen in about 600 million years–wikipedia’s Timeline of the far future cites this paper which notes towards the end (the pages aren’t numbered, but it’s a couple pages before the references section) that a study on the C3 plant Marchantia indicates its CO2 requirements are such that it “would be unable to grow at all 0.6 Gyr from now”. They also note on that page and the previous one that the productivity of plants drops before they completely die off, so that forests might disappear before C3 plants actually went extinct (which again would presumably decrease the chances of intelligent life evolving if it hadn’t already).

        I’ve read about this issue in various places, it’s noted for example in an interesting book about our planet’s long-term future, The Life and Death of Planet Earth, by Peter Ward and Donald Brownlee (they’re also the authors of the book Rare Earth, required reading for anyone interested in possible explanations for a past great filter, since it cites a number of plausible rare conditions in planetary systems needed for a planet in the habitable zone to be a stable home for the long-term evolution of multicellular life). In their comments about the future of photosynthesis in The Life and Death of Planet Earth, on p. 109 they say they say that among researchers “there seems to be a convergence on a time, somewhere between 500 million and a billion years from now, when land life as we know it will end on Earth, due to a combination of CO2 starvation and increasing heat. Moreover, global productivity is already in decline, and has perhaps been declining for the past 300 million years, as carbon dioxide levels have dropped as plants struggle to maintain a habitable temperature.”

        So, I think all of this provides some possible support for the scenario in your Great Filter paper of evolutionary “hard steps” that are approximately evenly-spaced (explained in terms of your safecracker analogy) with a spacing about equal to the time between the origin of multicellular animal life and the evolution of humans, say 500-800 million years. In that case there might be as many as 8 hard steps in our past (ignoring Rare Earth style rare conditions in the formation of our planetary system itself), so if each step had a probability of 1 in 1000 or less, that would give a total probability of less than 1 in 10^24–and I’ve seen the number of stars in our observable universe cited as 10^22 – 10^24, so this seems like a pretty plausible answer to the Fermi paradox without needing any hard steps in our future.

      • http://overcomingbias.com RobinHanson

        Thanks for a very thoughtful and relevant comment. I wonder how robust is this prediction of falling CO2 due to more weathering due to more luminosity. I accept luminosity will go up, but I thought many things influenced weathering and CO2.

      • hypnosifl

        Yes, I’d be interested to read a review article on the subject to see how much agreement there is, and how many different researchers have looked at the problem–Ward and Brownlee cite this 1999 paper in particular, and then say “This paper was by no means the last word. Other articles with slight refinements have appeared since, but there seems to be a convergence on a time, somewhere between 500 million and a billion years from now”. The authors of that paper are also part of a research group that has a number of other related papers listed on this page.

      • IMASBA

        Red dwarfs are more common than yellow stars. Red dwarfs have lifespans in the tens of billions of years and recently some have been found to have rocky planets in their habitable zone.

      • hypnosifl

        Red dwarfs are indeed a lot more common–this book says G type stars like the Sun make up about 4% of stars in our galaxy, while the estimates here suggest red dwarfs make up more than 70% of stars in most galaxies. But if you accept the usefulness of Bostrom’s self-sampling assumption, that very difference, combined with the fact that we nevertheless find ourselves in a G star’s system, should lead you to suspect that the evolution of intelligent life may be less probable in red dwarf systems. This article suggests one possible reason why that might be–gravitational interactions between a red dwarf and a planet in its habitable zone would tend to erode the planet’s axial tilt fairly quickly, so that the planet would no longer have seasons, which creates an extreme temperature gradient between the equator and the poles with only a narrow range of latitudes that would be habitable for complex multicellular animals (presumably decreasing the planet’s biodiversity and the pool of animal groups that might eventually give rise to an intelligent species), and in addition there’s a possibility that “In a worst-case scenario, however, the entire atmosphere of a zero-obliquity planet could collapse, Heller said. Gases might evaporate into space around the planet’s blazing middle and freeze to the ground in the bleak north and south.”

      • IMASBA

        “the fact that we nevertheless find ourselves in a G star’s system, should lead you to suspect that the evolution of intelligent life may be less probable in red dwarf systems.”

        Yes, but our sample size is kind of small.

        Tidal locking could be a problem for life but we just don’t know that for sure (thriving ecosystems might exist in underground lakes or deep oceans (Earth’s cephalopods show that dexterous intelligent beings can live in such places) and atmospheric heat circulation might greatly expand the livabl surface of a tidally locked planet.) Of course moons wouldn’t have the tidal locking problem at all.

      • IMASBA

        It is also entirely possible that primitive life is common in red dwarf solar systems today but that life on Earth has evolved uniquely fast. In other words that we’re (almost) alone today but that in a billion years life on Earth (and in other yellow star solar systems) has gone extinct/is primitive while the universe will be teeming with red dwarf solar system civilizations that have finally overcome all the difficult evolutionary steps.

      • Curt Adams

        That doesn’t sound right. If the planet is tidally locked like the moon is to the earth, then you’d get extreme climate variation. But just a low axial tilt produces a perpetual spring. Hadley circulation is quite adequate to keep most of the earth habitable in a perpetual spring. Maybe what you read was referring to tidal locking.

      • hypnosifl

        I included a link to the article where I read this claim in the comment you’re responding to, click the words “this article” there to read it–the article is definitely talking primarily about axial tilt, although eventually it adds that ‘To make matters worse for any beleaguered life forms, in the tilt erosion process a planet’s spinning on its axis slows as well. Given enough time, besides losing its seasons, a world becomes “tidally locked”‘. Googling a bit, I see another article here which focuses more on the fact that planets in the habitable zone of a red dwarf would become tidally locked, and that climate simulations indicate this would produce extreme temperature variations.

  • Hedonic Treader

    A singleton to the rescue, then.

    • Hedonic Treader

      Upon reflection, extinction would probably be better anyway.

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        Give me liberty or give me death! [But Robin didn’t think this way when he justified selling a child into slavery to avoid its death.]

      • Hedonic Treader

        I don’t care about the self bias of individuals. I expect it to be human nature to put oneself first, in most cases for most people.

        The real problem might be that extinction might be objectively better. And people would never get or acknowledge that, because it’s a low status belief that is socially undesirable – whether it is true or not.

        The irony is that the positive potential is enormous, but that doesn’t mean it will be unlocked, or that humanity avoids the critical mass of negtives that would outweigh it.

        When I look at the way politics works and the quality of arguments and attitudes people use to justify easily avoidable real-world suffering, I am rountinely shocked by how low the quality is. To expect a positive future from this, combined with how hard coordination is (Moloch), requires an abstract low-probability-high-payoff thinking (e.g. H > D ) or an absurd amount of optimism about how human nature will fix its flaws with technology.

  • ipencil

    Yet warring farmers only rarely went extinct.

    What’s your evidence for this? Going back past 1500, there is scant, if any, information. Since then, your observation may be survivor bias, and in many places there is still a dearth of information.

  • Pingback: Outside in - Involvements with reality » Blog Archive » Filtration

  • Silent Cal

    Faster flows also incentivise the intentional non-extractive destruction of enemy stocks.

    • http://overcomingbias.com RobinHanson

      Yup, an added problem.

  • arch1

    “‘I know not with what weapons World War III will be fought, but World War IV will be fought with sticks and stones.” –Albert Einstein, 1949.

  • http://juridicalcoherence.blogspot.com/ Stephen Diamond

    This is another way of saying what I was trying to say in my last post: improving tech can make war more destructive, increasing the risk of extinction via war.

    They’re independent arguments, aren’t they?

    • http://overcomingbias.com RobinHanson

      No, increasing flow access to stocks is another way of saying that returns to predation don’t diminish as fast with declining prey populations.

      • http://juridicalcoherence.blogspot.com/ Stephen Diamond

        The predation argument is premised on a limitation on stocks (to the solar system), whereas the second argument says nothing (that I can discern) about the trajectory of stocks.

  • arch1

    Is there a way to remotely determine whether a star has been partly artifically harvested? (Presumably such harvesting would remove material from near the star’s surface. BTW I seem to recall that type 1A supernovas reach critical mass by gravitationally harvesting their stellar companion)

  • Curt Adams

    I don’t see wars as *particularly* relevant to our propensity to exhaust stocks. Two very important stocks – fish in the ocean and reasonably safe potential CO2 release – have been mostly used up without any war being involved. Sure, war can increase the pressure to use up a stock, and can absolutely mess up cooperative arrangements to protect them. But, on the whole, I think it’s a secondary effect.

  • unclear

    I was at university the other day when I saw a waitress shove a cafe customer hard. The customer and her boyfriend were startled. Then a waiter turned on the boyfriend as he got between them. The waiter was a big muscle bound guy. He shoved the boyfriend to the ground. There was shouting. I ran across the road to settle things. I’ve done some jujutsu and muay thai training. More recently I had done MMA training. I’ve done some self-defence training from local self-defence schools too. I’m a small guy and physically weak but I’m a decent figher. I haven’t been in a street fight.

    I threw my bag aside and stood over the boyfriend and the waiter. He had him in a wristlock and was giving him some pain, not just immobalising him. The girlfriend was keen to leave and distressed but the boyfriend couldn’t get up. The waiter for shouting at the guy, along with the waitress. It was obvious that I didn’t have the physical strength to remove the waiters hand. I thought about all the techniques I could have used and realised that any crude action could mean the boyfriends hand breaks if the waiter holds on, but in all likelihood, it wouldn’t do much. My fighting strength is grappling, but even if I was in the boyfriends position, I would probably just escalate the situation since after arriving, the waiter had stopped escalating (and by bf was in no position too). The waiter would intermittently threaten me to back of but I didn’t give a shit. From experience with friends, I can generally handle anyone without fighting experience nomatter how big or muscly they are. In hindsight, the fact that he had got the guy in a pretty good wrist lock and his size should have hinted that maybe he is a fighter, perhaps a bouncer at night. In any case, all that I did, at best, was help keep things from escalating because I was a witness who had signalled my preparedness to get involved. I tried reasoning with the waiter but it didn’t make a difference.

    If the waiter had escalated, I don’t know what I could have done. Unless I was attacked myself and escalated, I had no means of protecting anyone. Eventually campus security arrived, but proved to be a bunch of wimps. They were scared of the big guy. There were 10 or so security staff and the guy was still on the ground in a wristlock. Eventually the police arrived and the waiter responded to their authority.

    This experience has made me wonder about how better I can reason with someone like the waiter in the future, or how I could prepare/train for a situation to help the guy on the ground. I highly value my ability to handle aggressors cause I had lots of physical abuse as a child and nobody who knew about it did anything. One idea I can think of, is to become physically big and strong, or signalling authority like the police officers.

    • http://juridicalcoherence.blogspot.com/ Stephen Diamond

      The question to me is whether you could have launched a knockout first strike against the waiter.