Is it reasonable to apply historical lessons about the rise and fall of civilizations to the present case of a single integrated civilization?
Historically those rapid rises and falls were enabled by inter-civ competition: The Romans stole market share from others, held onto it for a while, and in turn it was stolen from them. C'est la vie.
That all changes when there's a monopoly. Maybe that monopoly rots from within over time, or maybe it doesn't, but I don't see how lessons learned from prior civs would necessarily pertain. Unless that is we have a civ of AIs or aliens or Mars-based humans coming to eat our lunch.
It is almost irrelevant which scenario we choose unless we attach an at least approximate timeline to it. Will another 200 years pass like the last 200, ending in collapse? By then, GDP per capita in the world’s leading economy could reach around one million dollars. (In 1825, it was the equivalent of $3,500 in the UK; in 2025, it is $60,000 in the US — a 17-fold increase in two centuries.) Or will AI first accelerate growth, and only then comes the decline? In that case, it would be many times higher. And when collapse does come in 2200, will it be the whole of humanity, or “only” Earth?
Nothing meaningful can be said in either direction without specifying what causes the change in the growth rate. In the absence of this, the most reasonable thing is to assume that it will remain as it has been so far — a base rate of annual productivity growth of 1.2–1.3 percent, with occasional accelerations.
"But if we see our world today as a single integrated civ, then the civ rise and fall trend suggests that we will suffer a much larger decline soon."
In the 1990s, a great deal was discovered in a short time about the dynamics of complex systems. One robust result was that systems which continually self-organize to have greater complexity, usually? always? do so at least partly via changes which can be seen at some level of abstraction as a series of discrete events which have a power-law distribution of effect sizes, such that the average effect size approaches infinity as the system's size approaches infinity. E.g., if you look at gene fixations at the higher level of evolutionary cascades in an ecosystem, they're almost certainly a series of punctuated equilibria with that size distribution. If you look at language changes due to migrations, conquests, and anatomical evolution, those are huge, infrequent events, which might also have a power-law distribution approaching infinite average size.
The average effect size is infinite because as the system grows, it becomes more interconnected and complex, and the average effect size encompasses a greater fraction of the system. In civilizations, I think this would manifest as the odds of a collapse which takes down N% of the connections in the system increasing as the civilization grows larger and more-complex. So, once a civilization is above the size at which it can be taken down by random noise, it should become more-likely to have a total collapse as it grows more-complex. I think. (One catch is that "is large enough not to be taken down by random noise" is not really a Boolean proposition, but has a probability, which may scale at a rate which makes my claim invalid.)
An important consequence of the average effect size approaching infinity is that any such system is doomed to eventual collapse, because the probability that an effect which changes the entire system at once will be a total collapse of the system, probably approaches 1 as the system's size increases. (This "doom" would not necessarily come in the form of a total collapse, but might come as an endless series of partial collapses, each of which eases the growing pressure towards total collapse. Think of forest fires in California, which have the same distribution: if you had no forest fires for hundreds of years, there would be so much accumulated dead wood that the entire west coast might be one fully-connected woodpile, and burn down in one great conflagration.)
That doesn't mean collapse is inevitable for /us/, because we're smart, and can do the math, and might be able to figure out how to manage event size, randomness, and selection so as to provide continued growth without an infinite average effect size.
But selection seems to me to be the potential Achilles' heel. We can definitely prevent this inevitable collapse by doing selection ourselves, intelligently and deliberately. BUT, doing so puts us in a hermeneutic circle. We would be doing the selection according to our current values. Which means we would be deliberately trying to cut off every possible future with values which turn out to be /better/ than our own.
I agree with the perspective that it is at least decently likely we will have a significant decline sometime soon (i.e. in the lifetime of current young people).
I would be interested in seeing an analysis/scenario of how someone would practically prepare for this. Even if the Cultural Drift problem is intractable, it seems true that different people historically have taken more or less damage from their civilization falling
Also, is there definitely no trend in the percent reduction from a civlization fall? I would think a culture collapsing would have been closer to 100% in the past (extinction for the group) and a smaller reduction now as we are getting better at preserving knowledge.
1) Prior golden era empires were built primarily from a zero sum conquest model (combined with increasing internal scale). They died when they could no longer conquer, as they were never good enough at generating internal growth.
2) Prior golden era city states such as Florence and Athens were able to generate positive sum internal growth, but they tended to fall from external states combined with internal sclerosis.
3) The proven internal growth potential of modern places with liberal markets, fossil fuels, science and divided representative government is incomparably more robust than anything prior. By an order of magnitude.
4) We still face the major threats of external war and internal sclerosis, but now add the threat (or is it benefit?) of AI.
In reviewing my own comment, a better word for what I intended was incomparably more “powerful” or “successful”. I suppose you are right that we can’t know if it is more robust if that means resistance to collapse.
My own view is that we'll get AGI (or ems, but probably not as soon) before our civ loses the ability to make progress, and then all bets are off.
Even if the US fails, there are other places that will make progress for a while (China...).
And if Musk builds his independent Martian settlement that'll keep the flame alive (it has to - if civ collapses there everyone dies very quickly), but I think it'll take 50+ years to be able to survive independently even with a directed effort, and we may not have that much left.
Given our massive prosperity and innovation, combined with a global set of potential opportunities, once we get full scale entrepreneurial markets in law, governance, culture, and community, a fundamentally different dynamic process will come into being. Some of the results will resemble some features of the present, others might be very different. But the most likely prediction is that entrepreneurial markets in law, governance, culture, and community will, over many decades, bring the benefits of entrepreneurial capitalism to those domains which have been stagnating due to the absence of entrepreneurial capitalism.
This comparison is silly. Contemporary human society is fundamentally different from any societies that existed before say 1800. How? Look at energy consumption.
How much was per capita energy consumption in medieval Europe, or the ancient Middle East, or imperial China (excluding food)? The main form would be firewood for cooking and heating - no more than 5 kg/day (12 Mjoules) for a household of say six. So 2 Mjoules/day-person x 365 = 730 Mjoules/person-year.
World energy consumption in 2024 was about 600 exajoules (6 x 10^20) for 8 x 10^9 people; 7.5 x 10^11 = 750,000 Mjoules/person-year. That's three orders of magnitude. Energy is what humans use to manipulate external conditions.
Thus current ability to manipulate is enormously greater; humans can routinely overcome or negate external conditions that would be fatal to pre-industrial societies.
Could there be a societal collapse anyway? I suppose so, but the mechanism would have to be novel.
Re: "some try to deny that our world today counts as an integrated civ, as it doesn’t have a single ruler over it all. The US dominating the world for as long as it had doesn’t count, in their view" - that seems like the obvious response to me. A cultural monopoly could reasonably be taken to consist of one language, one currency, one religion and one nation. That doesn't seem to be especially close today.
One of my professors, an Assyriologist, remarked that in the four millennia BCE of written history in Mesopotamia, the region was ruled by a single government for a combined period of about 300 years. In the Old Testament period, approximately the last 1000 yrs. BCE, we see the rise and fall of the Assyrian, Babylonian, and Persian empires and the rise of the Roman.
Rome is probably the best example, splitting into East / West. The Western Roman Empire fell fast and violently, then mixed with other influences and evolved into something new (that we all know and love). The East hung around with a much slower decadent decay. I suspect the global civ decay would be similar, with some regions falling fast, but others, maybe the larger part?, with a more gradual decadent glide path to the bottom (unless something energetic and new is able to replace it before it gets to the bottom).
I suppose one civilization must logically be strongest if any exist, but I wouldn't say that Europeans were able to conquer so many other civilizations during the Age of Exploration because they all happened to be "weakened" coincidentally around the time those Europeans showed up.
Rome, Maya, Khmer, Han, Byzantine, Mesopotamia, Inca, Aztec, Egypt, Indus are all much more similar to eachother than they are to our civilization. Ten different breeds of horse, cattle, mule, goat vs a locomotive. Industrial civilization is special and uses its own rules.
The British Empire fell and Britain is a pale shadow of its former self in many respects. Clearly social decline is possible and a serious issue. But modern Britain is still much richer and more powerful than peak Imperial Britain. It is not comparable to peak Imperial Rome vs Dark Ages Rome.
Also I don't think the US dominates the world like a true hegemonic empire. Pre-WW2, it was just one of many great powers. Post WW2, the Soviet Union and now China. US collapse merely hands the keys to China. China has quite different legal system, politics, economic model, social structure... a different civilization.
Um. Because it's the wealthiest, safest, fairest civ in human history? It's far from perfect, but compared to what came before...
I mean we live longer than ever, even poor people are fat from too many calories, we can travel across the entire globe at will, and have very nearly abolished slavery. We have *morphine*. And cheap recorded music. Seems pretty good.
I strongly agree with Dave. Why would you not want it to continue, Artep? There has never been a time in history people flourished as they do now. Not even remotely close.
Is it reasonable to apply historical lessons about the rise and fall of civilizations to the present case of a single integrated civilization?
Historically those rapid rises and falls were enabled by inter-civ competition: The Romans stole market share from others, held onto it for a while, and in turn it was stolen from them. C'est la vie.
That all changes when there's a monopoly. Maybe that monopoly rots from within over time, or maybe it doesn't, but I don't see how lessons learned from prior civs would necessarily pertain. Unless that is we have a civ of AIs or aliens or Mars-based humans coming to eat our lunch.
A lot of prior civ collapse was due to internal decay processes, not external rivals.
It is almost irrelevant which scenario we choose unless we attach an at least approximate timeline to it. Will another 200 years pass like the last 200, ending in collapse? By then, GDP per capita in the world’s leading economy could reach around one million dollars. (In 1825, it was the equivalent of $3,500 in the UK; in 2025, it is $60,000 in the US — a 17-fold increase in two centuries.) Or will AI first accelerate growth, and only then comes the decline? In that case, it would be many times higher. And when collapse does come in 2200, will it be the whole of humanity, or “only” Earth?
Nothing meaningful can be said in either direction without specifying what causes the change in the growth rate. In the absence of this, the most reasonable thing is to assume that it will remain as it has been so far — a base rate of annual productivity growth of 1.2–1.3 percent, with occasional accelerations.
Earth and off Earth economy and culture would be quite linked, so rise and fall together.
"But if we see our world today as a single integrated civ, then the civ rise and fall trend suggests that we will suffer a much larger decline soon."
In the 1990s, a great deal was discovered in a short time about the dynamics of complex systems. One robust result was that systems which continually self-organize to have greater complexity, usually? always? do so at least partly via changes which can be seen at some level of abstraction as a series of discrete events which have a power-law distribution of effect sizes, such that the average effect size approaches infinity as the system's size approaches infinity. E.g., if you look at gene fixations at the higher level of evolutionary cascades in an ecosystem, they're almost certainly a series of punctuated equilibria with that size distribution. If you look at language changes due to migrations, conquests, and anatomical evolution, those are huge, infrequent events, which might also have a power-law distribution approaching infinite average size.
The average effect size is infinite because as the system grows, it becomes more interconnected and complex, and the average effect size encompasses a greater fraction of the system. In civilizations, I think this would manifest as the odds of a collapse which takes down N% of the connections in the system increasing as the civilization grows larger and more-complex. So, once a civilization is above the size at which it can be taken down by random noise, it should become more-likely to have a total collapse as it grows more-complex. I think. (One catch is that "is large enough not to be taken down by random noise" is not really a Boolean proposition, but has a probability, which may scale at a rate which makes my claim invalid.)
An important consequence of the average effect size approaching infinity is that any such system is doomed to eventual collapse, because the probability that an effect which changes the entire system at once will be a total collapse of the system, probably approaches 1 as the system's size increases. (This "doom" would not necessarily come in the form of a total collapse, but might come as an endless series of partial collapses, each of which eases the growing pressure towards total collapse. Think of forest fires in California, which have the same distribution: if you had no forest fires for hundreds of years, there would be so much accumulated dead wood that the entire west coast might be one fully-connected woodpile, and burn down in one great conflagration.)
That doesn't mean collapse is inevitable for /us/, because we're smart, and can do the math, and might be able to figure out how to manage event size, randomness, and selection so as to provide continued growth without an infinite average effect size.
But selection seems to me to be the potential Achilles' heel. We can definitely prevent this inevitable collapse by doing selection ourselves, intelligently and deliberately. BUT, doing so puts us in a hermeneutic circle. We would be doing the selection according to our current values. Which means we would be deliberately trying to cut off every possible future with values which turn out to be /better/ than our own.
(The same problem applies to genetic evolution.)
I agree with the perspective that it is at least decently likely we will have a significant decline sometime soon (i.e. in the lifetime of current young people).
I would be interested in seeing an analysis/scenario of how someone would practically prepare for this. Even if the Cultural Drift problem is intractable, it seems true that different people historically have taken more or less damage from their civilization falling
Also, is there definitely no trend in the percent reduction from a civlization fall? I would think a culture collapsing would have been closer to 100% in the past (extinction for the group) and a smaller reduction now as we are getting better at preserving knowledge.
Good question re pattern in % decline over time.
>how someone would practically prepare for this
Save lots of money. Store it in many diverse places.
Be ready to run on short notice.
The main lesson I took from all the Holocaust films I was made to watch as a child was "get out while you can"; don't wait until it's too late.
A couple of random thoughts.
1) Prior golden era empires were built primarily from a zero sum conquest model (combined with increasing internal scale). They died when they could no longer conquer, as they were never good enough at generating internal growth.
2) Prior golden era city states such as Florence and Athens were able to generate positive sum internal growth, but they tended to fall from external states combined with internal sclerosis.
3) The proven internal growth potential of modern places with liberal markets, fossil fuels, science and divided representative government is incomparably more robust than anything prior. By an order of magnitude.
4) We still face the major threats of external war and internal sclerosis, but now add the threat (or is it benefit?) of AI.
I don't think you can know how internally robust is our current civ culture.
In reviewing my own comment, a better word for what I intended was incomparably more “powerful” or “successful”. I suppose you are right that we can’t know if it is more robust if that means resistance to collapse.
Johan Norberg's most recent book is about lessons from past civ declines - https://www.amazon.com/Peak-Human-Historys-Greatest-Civilizations/dp/1838957294 Maybe not a lot there that's new to you.
My own view is that we'll get AGI (or ems, but probably not as soon) before our civ loses the ability to make progress, and then all bets are off.
Even if the US fails, there are other places that will make progress for a while (China...).
And if Musk builds his independent Martian settlement that'll keep the flame alive (it has to - if civ collapses there everyone dies very quickly), but I think it'll take 50+ years to be able to survive independently even with a directed effort, and we may not have that much left.
Interesting times.
Even an AI/em civ could fall.
Sure, but observations about human civs are probably less applicable to them.
Sure. And if one does, the fall may be harder and more permanent than past falls.
But the mechanism and causes may be new and hard to anticipate from here.
Given our massive prosperity and innovation, combined with a global set of potential opportunities, once we get full scale entrepreneurial markets in law, governance, culture, and community, a fundamentally different dynamic process will come into being. Some of the results will resemble some features of the present, others might be very different. But the most likely prediction is that entrepreneurial markets in law, governance, culture, and community will, over many decades, bring the benefits of entrepreneurial capitalism to those domains which have been stagnating due to the absence of entrepreneurial capitalism.
This comparison is silly. Contemporary human society is fundamentally different from any societies that existed before say 1800. How? Look at energy consumption.
How much was per capita energy consumption in medieval Europe, or the ancient Middle East, or imperial China (excluding food)? The main form would be firewood for cooking and heating - no more than 5 kg/day (12 Mjoules) for a household of say six. So 2 Mjoules/day-person x 365 = 730 Mjoules/person-year.
World energy consumption in 2024 was about 600 exajoules (6 x 10^20) for 8 x 10^9 people; 7.5 x 10^11 = 750,000 Mjoules/person-year. That's three orders of magnitude. Energy is what humans use to manipulate external conditions.
Thus current ability to manipulate is enormously greater; humans can routinely overcome or negate external conditions that would be fatal to pre-industrial societies.
Could there be a societal collapse anyway? I suppose so, but the mechanism would have to be novel.
Re: "some try to deny that our world today counts as an integrated civ, as it doesn’t have a single ruler over it all. The US dominating the world for as long as it had doesn’t count, in their view" - that seems like the obvious response to me. A cultural monopoly could reasonably be taken to consist of one language, one currency, one religion and one nation. That doesn't seem to be especially close today.
One of my professors, an Assyriologist, remarked that in the four millennia BCE of written history in Mesopotamia, the region was ruled by a single government for a combined period of about 300 years. In the Old Testament period, approximately the last 1000 yrs. BCE, we see the rise and fall of the Assyrian, Babylonian, and Persian empires and the rise of the Roman.
I think we can expect our global civ to fragment as it decays, which should be cause for hope in your model.
I don't see why we should expect that to happen anymore more now than it did in past civ falls.
Rome is probably the best example, splitting into East / West. The Western Roman Empire fell fast and violently, then mixed with other influences and evolved into something new (that we all know and love). The East hung around with a much slower decadent decay. I suspect the global civ decay would be similar, with some regions falling fast, but others, maybe the larger part?, with a more gradual decadent glide path to the bottom (unless something energetic and new is able to replace it before it gets to the bottom).
Since invasion by others really is a hugely important cause of civilizational collapse (I agree with Steve Sailer's critique of Jared Diamond https://entitledtoanopinion.wordpress.com/2012/12/29/the-collapse-of-complex-societies/ ), that is reason to be relatively complacent.
Invasion is a big proximate cause, but less of a distal cause. Once other factors weaken a civ, it becomes ripe for invasion.
I suppose one civilization must logically be strongest if any exist, but I wouldn't say that Europeans were able to conquer so many other civilizations during the Age of Exploration because they all happened to be "weakened" coincidentally around the time those Europeans showed up.
That was a very unusual case.
I wouldn't even say it was unique for the Americas: they appear to have been inhabited prior to the arrival of most of the ancestors of the Amerindians of Columbus' time, but the people there earliest got wiped out. https://westhunt.wordpress.com/2022/01/07/the-first-team/ https://westhunt.wordpress.com/2025/08/30/king-of-the-beasts/
Rome, Maya, Khmer, Han, Byzantine, Mesopotamia, Inca, Aztec, Egypt, Indus are all much more similar to eachother than they are to our civilization. Ten different breeds of horse, cattle, mule, goat vs a locomotive. Industrial civilization is special and uses its own rules.
The British Empire fell and Britain is a pale shadow of its former self in many respects. Clearly social decline is possible and a serious issue. But modern Britain is still much richer and more powerful than peak Imperial Britain. It is not comparable to peak Imperial Rome vs Dark Ages Rome.
Also I don't think the US dominates the world like a true hegemonic empire. Pre-WW2, it was just one of many great powers. Post WW2, the Soviet Union and now China. US collapse merely hands the keys to China. China has quite different legal system, politics, economic model, social structure... a different civilization.
Why on earth would anybody with a heart and soul wish for this “civilisation” to continue?
Um. Because it's the wealthiest, safest, fairest civ in human history? It's far from perfect, but compared to what came before...
I mean we live longer than ever, even poor people are fat from too many calories, we can travel across the entire globe at will, and have very nearly abolished slavery. We have *morphine*. And cheap recorded music. Seems pretty good.
I strongly agree with Dave. Why would you not want it to continue, Artep? There has never been a time in history people flourished as they do now. Not even remotely close.
Total utilitarianism. Also, souls don't exist.
You may not have a soul, but I highly doubt it. Still, I won’t argue with you
Do read The Dawn of Everything about what’s come before and what’s possible. This capitalism is not the best ever and it’s built on exploitation
I have read that.
Because we’re trashing the Earth. It’s not just about humans. What about all the other creatures. It’s so typical that people just think about people.