”Just like plants improved the efficiency of photosynthesis by moving the Rubisco reactions to a separate part of the leaf, James Watt doubled the efficiency of steam power by separating the hot and cold parts of the engine. This pattern extends to the modern day. We get more food from less land, more energy from less fuel, and more computation from smaller circuits.“ - this is absurd. We do get more food from less land but that is because we pump ungodly amounts of fossil fuels in the food production process. Numerous studies have shown that the energy efficiency of the modern agricultural sector is much worse than slash and burn or many other forms of primitiv agriculture.
An increase in efficiency does not imply a decrease in resource consumption. This has never happened in past and never will, each increase in efficiency is matched with an instantaneous increase in the overall consumption. Yes we made the electric bulbs more efficient, but we also increased their reach, making the overall consumption of electricity go up. Think of any tech and this is true, there has not been one single year when there has been a decrease in resource consumption. And this is keeping in mind that most of the world still lives in poverty. What we have managed to do for them is to feed them, and that has ruined ecosystems around the world, just to feed us. Imagine the amount of resources that will be required in making their standard of living at par with the west.
But this argument still ignores the main reason growth is pursued. Technological progress, efficiency maximisation, resource extraction, all of these are mere means to an end. And the end is the maximisation of return on capital invested. It is not inherent in technology that it will seek indefinite and explosive growth, but it is the fiduciary duty of a chief executive to his shareholders that he maximise the return on their capital, and failing to do so, he will be sacked and another pawn will be brought in his place who can ensure growth. Technology is just the means to deliver these return. Efficiency plays the game of capital when it makes chasing these returns easier, in no other context is efficiency beneficial to our society that thrives on planned obsolescence.
"Since each human is born with this invention-machine in their skull, the rate of invention increases with our population size. This positive feedback loop defines the arc of human history and secures the prospects for unlimited future growth in human prosperity."
I have to ask then, Maxwell, what do you think our prospects are after 2050? Population stagnation and decline appear to be a sure bet at this point. Does this spell the end of progress or will AI pick up the slack?
This exact question is one of the things which got me into this line of research. The best macro-economic models of economic growth suggest that everything comes down to population growth in the end. But it does seem possible that AI can increase the effective population of researchers by extending the power of each individual.
Seems like invention doesn't quite increase with population -- more like that it increases with the population that's making use of a functioning economy. I guess really it's productivity times population.
This is why it's so great when high function countries like the US and Canada have immigrants, who come from, say, India, which also has a functioning economy, but at a lower productivity.
And this is why it's so great when more countries establish functioning economies in the first place.
Oct 18, 2023·edited Oct 19, 2023Liked by Maxwell Tabarrok
I don't think it's fair to say biologists see "collapse." In logistic growth - as opposed to Malthus- growth slows gently until you reach a relatively stable population size (carrying capacity). You do not see a crash as Malthus hypothesised. There are some classic models like Lotka–Volterra which have booms and busts, but this is a predator prey model and doesn't apply to us.
From an ecological viewpoint, carrying capacity is fundamentally limited because, although in theory new technology can raise it to some degree, at the end of the day it's at the very least limited by the number of atoms in the universe. As far as I know we live in a finite universe with finite (and indeed, a decreasing amount of) energy. Perhaps this is a rather weak claim but nevertheless important to point out.
My null hypothesis for the pattern of human growth that we see today is that we are simply reaching our carrying capacity. If so, any attempts to artificially raise fertility rates through policy will be dynamically resisted in perhaps unpredictable ways. So to me when economists suggest some policy or the other to raise fertility rates, to me it seems a bit ridiculous... like expecting the ocean to rise because you threw a pebble in.
Yes the basic logistic model is a gentle slow down near the maximum. But when there is some finite stockpile which fuels exponential growth above the long run maximum, then there is a collapse. This is exactly what Ehrlich predicts in the population bomb.
And yeah I agree that I leave the question of final physical limits unanswered. If we agree that thousands of years of growth are possible then that's a big update from the standard ecological view which I think is important.
I didn't talk about this in the post but I don't think that it's obvious that there are finite limits to growth in the universe. There are singularities everywhere we look!
Equating current slow down in population growth to reaching the limits of carrying capacity is highly dubious.
Firstly, we are a vast distance from the "all the atoms in the universe" limit.
Secondly, in typical biology setups, the population is limited by starvation/disease. Here it's limited by cheap contraception and people deciding that reproduction is too much work.
Yeah, population stops growing because of the 'demogrqphic transition', where much of the world has now reached a level of education and wealth that leads them to choose to have smaller families. You can see on the numbers of any of these rich countries that they're nowhere near any limits on available food, much less atoms, but they are shrinking, by the life choices (behavior) of individual people.
If we saturate efficiency gains against natural laws , and there are plenty to choose from ..., Chambadal-Novikov, Carnot, Betz, Shockley–Queisser ,etc ... , you just end up back to the sane idea that there are limits.
Overcoming some limiting factors does not generalize infinitely to all. There is no evidence to say we will be able to make leaps over limiting factors continuously against diminishing returns to complexity and there is absolutely proven thousands of known limits that cannot be overcome if you are a believer in physics . In the case you aren't it would just go back to being a pseudoscientist like most economists.
And not all humans are born with invention machines in their skulls, most invent nothing. Yet again population increases will butt up against diminishing returns as total factor productivity of inventiveness declines which is already measurable and studied. However my argument doesn't rely on this because maybe AI delimits us there to another level constrained by energy.
Economics and ecology are not aligned in a vision of unlimited growth and prosperity. Relative decoupling of "growth" from the physical reality can get more growth but absolute decoupling is impossible. So you still hit boundaries, get into intensification traps and cause mass extinctions. This fictitious Absolute-decoupling even if it was achieved would look like someone just typing in lots of numbers in a database or a math equation that = ♾️ and then pretending that is infinite growth. Which is not meaningful.
>"This episode in evolutionary history shows that there is more to population dynamics than the usual story told by ecologists. "
No sir , your strawman ecologist only exists in the mind of you and economists. Please read Howard T Odum. To get a holistic understanding .
Since you like the biological metaphor, Your arguments are equivalent to you positing something like an infinite number of efficiency gains like C3 to C4 to Crassulacean Acid metabolism to and unlimited number of more innovations being possible . Unfortunately the reality is that that isn't possible , reality isn't infinitely malleable and doesn't work off your hopes and dreams of techno-hopium pseudoscience.
I would suggest rather than pumping out there puff peices like the rest of the tech-bro-infinity crowd we try to normalize accepting the limits of reality and and accept not ignoring them because they are scary. It's been moving towards a pseudoscientiific secular religion.
If people adopt this religion then it allows them to ignore the very real limiting factors, waste absorption rates, and the fact we haven't transhumanned ourselves out of dependence upon earths biosphere yet. Sometimes growth must be abandoned for survival , so the blind faith in growth full speed ahead is insane if not examined to make sure we are going to get escape velocity rather than sputter and crash. Better to throttle biosphere destroying growth and take a safer slower approach to get the singularity rather than kill ourselves with stupid ideology.
I used "infinite" "indefinite" and "unlimited" sloppily throughout the piece. I think that most of your objections are about making growth rates faster and faster because this quickly hits physical limits.
The Ehrlich position is that that growth rates must go to zero soon. I am trying to respond to that and show at least that growth can continue at a constant rate for thousands of years.
Howard T Odum looks really interesting, thank you for the reference. But I don't think what I said was a straw man. Odum does not represent "the usual story told by ecologists."
Who would you be using as reference ecologists? My experience with ecologists has been they are generally good at science and not full of shit . But maybe you are talking about some hippy that got an associate degree and a blog or something. Maybe the ecologist equivalent of the techno-hopium community. Pushing the Despairoin vs Hopium drug of choice false dichotomy continues unabated in the blogosphere I'm sure
Where does erlich say growth must go to zero? I want to see his full arguments. I haven't kept up with anything he says because he ignores Wrights law and discounted the spread of de-limiting from green revolution inputs when making his prognostications. Much like people still do with Africa where we still have low penetration of green revolution package.
Any argument for growth continuing for thousands of years can be proven wrong with math unless we are talking about, as I had mentioned in previous comment, these types of growth like calling printing AI generated book pdf files and then printing money to buy the pdf files and calling it growth and boosting GDP. But those types of growth become fairly meaningless Sagan "garage dragons" when it comes down to it.
Start here. To save me having to phone type comments.
There is also this thing economists do where they are ecology blind by saying we can have all this growth and carrying capacity is infinite, while the ecologists point out that we are already in a mass extinction because of it and it comes at an increasing cost. Even if ultimate stock of resources is large the flow speed of acquisition of them can be to slow and discharging waste can exceed boundaries. We are already above the sustainable growth rates measured empirically. Top soil going down, species going down, toxins going up, etc ..
What does it mean to say "growth" when it comes with the contraction of everything else via the externalities.
The language around this entire subject is disturbing like when Stephen Pinker says we are getting richer and × people have been lifted from poverty using this dollar number but simultaneously the actual number of people that are food insecure increased by a hundreds of millions. It's just this weird propaganda machine and there is a whole dedicated group propagating the nonsense instead of trying to develop a real sense of where we are. Clearly funded by moneyed interests of people that want to see the status quo continue, just don't look over there at the extinction behind the curtain, everyone remain happy with the progress across the dimensions and metrics we feed you . Ok I'm just ranting now bye.
At the 2.9% growth rate in energy use, 1000 years is a few stars power output.
And I don't think all low energy economic value is meaningless rubbish.
We could imagine a world where everyone was a brain in a vat, living in a virtual utopia run on very efficient computers. That world could have a tiny energy use per person, and yet very pleasant lives.
In the 1910's we were running out of Nitrogen containing minerals. And then we invented nitrogen fixation. So X going down doesn't mean we are past the limits, just that we haven't invented a way to fix that problem quite yet.
Maybe in a few years we can use GMO algae to remove old pollution.
I see no laws of physics against disassembling stars for fusion fuel.
That gives a finite amount of energy, but pragmatically a very large amount.
So if we think all sorts of tech on that level is possible, it's a game of "can we build the high tech stuff before the low tech runs out". Ie can we build enough solar before the coal runs out. And yes we can.
This is very insightful, thank you! If you look at history its also the case that places with higher populations have proportionally higher population growth too. Having more humans around isn't more mouths to feed but seems to generate more ideas etc (ie Romer's nobel), and this leads to more prosperity. Now, perhaps at very high population levels like ours today something fundamentally changes, but someone would need to have a good argument for what that thing might be!
Would you by any chance be interested in talking with me about some of your interests on a podcast, as I run one? I would love to talk about some of the topics on your substack!
”Just like plants improved the efficiency of photosynthesis by moving the Rubisco reactions to a separate part of the leaf, James Watt doubled the efficiency of steam power by separating the hot and cold parts of the engine. This pattern extends to the modern day. We get more food from less land, more energy from less fuel, and more computation from smaller circuits.“ - this is absurd. We do get more food from less land but that is because we pump ungodly amounts of fossil fuels in the food production process. Numerous studies have shown that the energy efficiency of the modern agricultural sector is much worse than slash and burn or many other forms of primitiv agriculture.
An increase in efficiency does not imply a decrease in resource consumption. This has never happened in past and never will, each increase in efficiency is matched with an instantaneous increase in the overall consumption. Yes we made the electric bulbs more efficient, but we also increased their reach, making the overall consumption of electricity go up. Think of any tech and this is true, there has not been one single year when there has been a decrease in resource consumption. And this is keeping in mind that most of the world still lives in poverty. What we have managed to do for them is to feed them, and that has ruined ecosystems around the world, just to feed us. Imagine the amount of resources that will be required in making their standard of living at par with the west.
But this argument still ignores the main reason growth is pursued. Technological progress, efficiency maximisation, resource extraction, all of these are mere means to an end. And the end is the maximisation of return on capital invested. It is not inherent in technology that it will seek indefinite and explosive growth, but it is the fiduciary duty of a chief executive to his shareholders that he maximise the return on their capital, and failing to do so, he will be sacked and another pawn will be brought in his place who can ensure growth. Technology is just the means to deliver these return. Efficiency plays the game of capital when it makes chasing these returns easier, in no other context is efficiency beneficial to our society that thrives on planned obsolescence.
The US and many other countries consumed less fossil fuels in absolute numbers (and much less per person) than it did in 2000.
https://ourworldindata.org/grapher/fossil-fuel-primary-energy?tab=chart
It also uses less farmland in total.
https://ourworldindata.org/grapher/cropland-use-over-the-long-term
"Since each human is born with this invention-machine in their skull, the rate of invention increases with our population size. This positive feedback loop defines the arc of human history and secures the prospects for unlimited future growth in human prosperity."
I have to ask then, Maxwell, what do you think our prospects are after 2050? Population stagnation and decline appear to be a sure bet at this point. Does this spell the end of progress or will AI pick up the slack?
This exact question is one of the things which got me into this line of research. The best macro-economic models of economic growth suggest that everything comes down to population growth in the end. But it does seem possible that AI can increase the effective population of researchers by extending the power of each individual.
This is a good piece on the topic
https://www.openphilanthropy.org/research/could-advanced-ai-drive-explosive-economic-growth/
Seems like invention doesn't quite increase with population -- more like that it increases with the population that's making use of a functioning economy. I guess really it's productivity times population.
This is why it's so great when high function countries like the US and Canada have immigrants, who come from, say, India, which also has a functioning economy, but at a lower productivity.
And this is why it's so great when more countries establish functioning economies in the first place.
Yep, this is a super fascinating topic. Thanks for sharing this, probably will form the basis of a future write up.
I don't think it's fair to say biologists see "collapse." In logistic growth - as opposed to Malthus- growth slows gently until you reach a relatively stable population size (carrying capacity). You do not see a crash as Malthus hypothesised. There are some classic models like Lotka–Volterra which have booms and busts, but this is a predator prey model and doesn't apply to us.
From an ecological viewpoint, carrying capacity is fundamentally limited because, although in theory new technology can raise it to some degree, at the end of the day it's at the very least limited by the number of atoms in the universe. As far as I know we live in a finite universe with finite (and indeed, a decreasing amount of) energy. Perhaps this is a rather weak claim but nevertheless important to point out.
My null hypothesis for the pattern of human growth that we see today is that we are simply reaching our carrying capacity. If so, any attempts to artificially raise fertility rates through policy will be dynamically resisted in perhaps unpredictable ways. So to me when economists suggest some policy or the other to raise fertility rates, to me it seems a bit ridiculous... like expecting the ocean to rise because you threw a pebble in.
Yes the basic logistic model is a gentle slow down near the maximum. But when there is some finite stockpile which fuels exponential growth above the long run maximum, then there is a collapse. This is exactly what Ehrlich predicts in the population bomb.
And yeah I agree that I leave the question of final physical limits unanswered. If we agree that thousands of years of growth are possible then that's a big update from the standard ecological view which I think is important.
I didn't talk about this in the post but I don't think that it's obvious that there are finite limits to growth in the universe. There are singularities everywhere we look!
Equating current slow down in population growth to reaching the limits of carrying capacity is highly dubious.
Firstly, we are a vast distance from the "all the atoms in the universe" limit.
Secondly, in typical biology setups, the population is limited by starvation/disease. Here it's limited by cheap contraception and people deciding that reproduction is too much work.
Yeah, population stops growing because of the 'demogrqphic transition', where much of the world has now reached a level of education and wealth that leads them to choose to have smaller families. You can see on the numbers of any of these rich countries that they're nowhere near any limits on available food, much less atoms, but they are shrinking, by the life choices (behavior) of individual people.
If we saturate efficiency gains against natural laws , and there are plenty to choose from ..., Chambadal-Novikov, Carnot, Betz, Shockley–Queisser ,etc ... , you just end up back to the sane idea that there are limits.
Overcoming some limiting factors does not generalize infinitely to all. There is no evidence to say we will be able to make leaps over limiting factors continuously against diminishing returns to complexity and there is absolutely proven thousands of known limits that cannot be overcome if you are a believer in physics . In the case you aren't it would just go back to being a pseudoscientist like most economists.
And not all humans are born with invention machines in their skulls, most invent nothing. Yet again population increases will butt up against diminishing returns as total factor productivity of inventiveness declines which is already measurable and studied. However my argument doesn't rely on this because maybe AI delimits us there to another level constrained by energy.
Economics and ecology are not aligned in a vision of unlimited growth and prosperity. Relative decoupling of "growth" from the physical reality can get more growth but absolute decoupling is impossible. So you still hit boundaries, get into intensification traps and cause mass extinctions. This fictitious Absolute-decoupling even if it was achieved would look like someone just typing in lots of numbers in a database or a math equation that = ♾️ and then pretending that is infinite growth. Which is not meaningful.
>"This episode in evolutionary history shows that there is more to population dynamics than the usual story told by ecologists. "
No sir , your strawman ecologist only exists in the mind of you and economists. Please read Howard T Odum. To get a holistic understanding .
Since you like the biological metaphor, Your arguments are equivalent to you positing something like an infinite number of efficiency gains like C3 to C4 to Crassulacean Acid metabolism to and unlimited number of more innovations being possible . Unfortunately the reality is that that isn't possible , reality isn't infinitely malleable and doesn't work off your hopes and dreams of techno-hopium pseudoscience.
I would suggest rather than pumping out there puff peices like the rest of the tech-bro-infinity crowd we try to normalize accepting the limits of reality and and accept not ignoring them because they are scary. It's been moving towards a pseudoscientiific secular religion.
If people adopt this religion then it allows them to ignore the very real limiting factors, waste absorption rates, and the fact we haven't transhumanned ourselves out of dependence upon earths biosphere yet. Sometimes growth must be abandoned for survival , so the blind faith in growth full speed ahead is insane if not examined to make sure we are going to get escape velocity rather than sputter and crash. Better to throttle biosphere destroying growth and take a safer slower approach to get the singularity rather than kill ourselves with stupid ideology.
I used "infinite" "indefinite" and "unlimited" sloppily throughout the piece. I think that most of your objections are about making growth rates faster and faster because this quickly hits physical limits.
The Ehrlich position is that that growth rates must go to zero soon. I am trying to respond to that and show at least that growth can continue at a constant rate for thousands of years.
Howard T Odum looks really interesting, thank you for the reference. But I don't think what I said was a straw man. Odum does not represent "the usual story told by ecologists."
Who would you be using as reference ecologists? My experience with ecologists has been they are generally good at science and not full of shit . But maybe you are talking about some hippy that got an associate degree and a blog or something. Maybe the ecologist equivalent of the techno-hopium community. Pushing the Despairoin vs Hopium drug of choice false dichotomy continues unabated in the blogosphere I'm sure
Where does erlich say growth must go to zero? I want to see his full arguments. I haven't kept up with anything he says because he ignores Wrights law and discounted the spread of de-limiting from green revolution inputs when making his prognostications. Much like people still do with Africa where we still have low penetration of green revolution package.
Any argument for growth continuing for thousands of years can be proven wrong with math unless we are talking about, as I had mentioned in previous comment, these types of growth like calling printing AI generated book pdf files and then printing money to buy the pdf files and calling it growth and boosting GDP. But those types of growth become fairly meaningless Sagan "garage dragons" when it comes down to it.
Start here. To save me having to phone type comments.
https://dothemath.ucsd.edu/2011/07/can-economic-growth-last/
https://dothemath.ucsd.edu/2011/07/galactic-scale-energy/
https://dothemath.ucsd.edu/2012/04/economist-meets-physicist/
There is also this thing economists do where they are ecology blind by saying we can have all this growth and carrying capacity is infinite, while the ecologists point out that we are already in a mass extinction because of it and it comes at an increasing cost. Even if ultimate stock of resources is large the flow speed of acquisition of them can be to slow and discharging waste can exceed boundaries. We are already above the sustainable growth rates measured empirically. Top soil going down, species going down, toxins going up, etc ..
What does it mean to say "growth" when it comes with the contraction of everything else via the externalities.
The language around this entire subject is disturbing like when Stephen Pinker says we are getting richer and × people have been lifted from poverty using this dollar number but simultaneously the actual number of people that are food insecure increased by a hundreds of millions. It's just this weird propaganda machine and there is a whole dedicated group propagating the nonsense instead of trying to develop a real sense of where we are. Clearly funded by moneyed interests of people that want to see the status quo continue, just don't look over there at the extinction behind the curtain, everyone remain happy with the progress across the dimensions and metrics we feed you . Ok I'm just ranting now bye.
At the 2.9% growth rate in energy use, 1000 years is a few stars power output.
And I don't think all low energy economic value is meaningless rubbish.
We could imagine a world where everyone was a brain in a vat, living in a virtual utopia run on very efficient computers. That world could have a tiny energy use per person, and yet very pleasant lives.
In the 1910's we were running out of Nitrogen containing minerals. And then we invented nitrogen fixation. So X going down doesn't mean we are past the limits, just that we haven't invented a way to fix that problem quite yet.
Maybe in a few years we can use GMO algae to remove old pollution.
There are likely to be limits.
The question is if we are near them.
I see no laws of physics against disassembling stars for fusion fuel.
That gives a finite amount of energy, but pragmatically a very large amount.
So if we think all sorts of tech on that level is possible, it's a game of "can we build the high tech stuff before the low tech runs out". Ie can we build enough solar before the coal runs out. And yes we can.
This is very insightful, thank you! If you look at history its also the case that places with higher populations have proportionally higher population growth too. Having more humans around isn't more mouths to feed but seems to generate more ideas etc (ie Romer's nobel), and this leads to more prosperity. Now, perhaps at very high population levels like ours today something fundamentally changes, but someone would need to have a good argument for what that thing might be!
So glad you enjoyed it! Definitely huge influence from Romer's work here.
Would you by any chance be interested in talking with me about some of your interests on a podcast, as I run one? I would love to talk about some of the topics on your substack!