I do not blindly believe in the word 'exponential', as I have alluded to in my piece about the delusions of the myth of progress. Many people throw that word around, talking about how everything in modernity is an exponential curve, but I think it is very misleading. Nothing in the world of physics and ecology is exponential forever. A population might grow exponentially for a short while, but eventually it runs into the boundaries of its environment. 1
Things in the human world are similar: in the short-term things can grow exponentially, but eventually you run into a limit to how many people buy you things, or how many users get on your social media platform. The problem is that 'short-term' is a very contextual and frame-dependent term. For most modern people, 5 years is a long time, and 15 years is an incredibly long time, the kind of thing that they call "long-term". Thus if they see that a company has kept growing for 5 to 10 years, they will consider this to be a hard truth etched into the fabric of reality, because their perspective on time is very, very narrow.
This is not helped by the fact that modern 'culture' is incredibly biased towards novelty, as it only pays attention to the stuff released recently, and maybe to things released a few years ago for a cheap nostalgia trip. You don't see people read books from the 17th century, and even that period could be said to be fairly 'modern' in the scheme of civilization as a whole.
In fact, modernity could only be the way it is because of its inherent bias towards novelty. The constant series of technological changes necessarily change how people live and what opportunities are present, meaning that young people have to keep up otherwise they wouldn't be able to fit in this constantly changing society. Older people complain that young people are only interested in the new things, which is a fair complaint when it comes to culture because modern media tends to be awful, but think of how employable a young person would be if they did the exact same things, and thought the same way, that their parents, or even grand-parents did, to earn a living and secure prosperity for themselves. Such a person would be hilariously anachronistic to how their current society functions, because new options are made, displacing the old ways of the past, constnatly. Thus, the modern world necessarily has to be neotenous in order to accommodate for the constant changes brought about by technology.
Here is a general model to understand the different stages of growth, to give an alternative to the naive view of perpetually exponential processes. The function I have in mind is basically the sigmoid curve, which applies for a single cycle of growth.
Things get more tricky if we think about a succession of cycles. Do the sigmoids stacks in such a way that, together, they produce a linear trend? Or is the resultant an exponential trend? Or is the resultant itself a sigmoid trend? Or perhaps the collapsologists are right and everything eventually comes down. Hard to tell, but let's start with the basics.
In the beginning of any cycle of growth is a period of slow start. This happens for a variety of reasons: little starting capital, little attention on the idea, (here I am not just examining technologies and products, but also how ideas scale) inability to take advantage of network efforts and economies of scale, too expensive for most people to afford—very typical of technologies: the real revolution is not when they're invented, but when they become affordable—this type of stuff.
It takes time for momentum to build up, but when it does, the resulting phase can be very explosive, especially in our social media-driven world. This is the phase of growth, the proper 'exponential' part that most people pay attention to, where the profits of a company can be reinvested back, when the extra attention to an idea allows it to gather funding and be made significantly better as a result, when a new market starts to gather tons of people, when a movement becomes big enough that the public at large knows about it, etc.
Everything however ultimately stalls out in growth, leading to a third phase of diminishing returns. Why is that? Why can't things just keep growing? Well one of the main reasons, as mentioned above, is that the number of users and consumers is finite, not just because human population is finite, but because the pool of people who would be interested in a product, or the number of scenarios where a technology could be used, are limited, and because many products, ideas and services also run into competition.
It's tempting to imagine for instance, especially for very naive people who haven't been confronted with the harsh realities of how ideologies wrestle with one another, that a worldview, or a method, aimed at improving people's lives could scale to the entirety of the world. This is what some Buddhists hope for instance, that because their own life has been turned around through their practice, that the entire world would benefit from doing the same. This naivety sounds absurd when it is stated plainly, yet every movement that tries to 'save the world' is essentially attempting that, whether they come from a religious perspective or not.
The naive view that your movement could grow exponentially until it captures the entirety of the planet ignores just how different people can be, across social classes, generations and different cultures. Typically, when a movement spreads like wildfire, it does so within a certain cohort of people: middle class, young Western people typically, the type who come from a rootless world and thus adopt new ideas and new worldviews much more readily, and who have enough money and safety that they can be idealistic in how they view the world. As soon as it exhausts that cohort of people however, it is rarely able to "jump" over to new groups: lower class people find the movement to full of privileged people, older people find it to be too rootless and catering to youth culture, and different cultures find the political and spiritual alignment strange, for example.
This is another reason why class inequality is so problematic. Not only does it lead to tons of people having to scrounge for money, while a few essentially treat money as a game to make more money and fulfill their wildest fantasies, but it leads to distinct classes of people who can only live in parallel social realities. We happen to live at a time where the lower class tends to ally itself with the upper class through what is called 'the right-wing', for a variety of reasons, mainly related to mass-migration and the complete inadequacy of the left to really solve any problem, but this is a transient alliance if there ever was one. The interests of the upper and lower class are, predictably, completely misaligned at the end of the day, but the same is true between the lower and the middle class.
Returning to the diminishing returns on growth, another example lies in the inevitable competitiveness of a market or a field. A product or service might initially find great success while it is growing, the second phase of the sigmoid curve, but when it runs into competition, it will have to be more strategic and have to make tradeoffs in order to remain in the game. This is why products and services tend to be much better when they are fairly new, but as soon as they become part of the mainstream players, they degenerate into the same slop and compromises as everything else, pushing the need for profits above everything else, because if the companies didn't, they wouldn't be able to remain in the game. Competition has a way of phasing out quality over time, because competition creates its own logic of self-perpetuation at all costs.
With competition in mind, it's quite obvious why the biggest tech companies are all oligarchic, or essentially monopolistic in our times. If they ran into competition like most sectors do, their growth would be stunted by the rest of the field, meaning that they wouldn't become the biggest tech companies in the first place. The reason why they don't run into competition is because of the nature of network effects: infrastructure (Amazon's distribution, or social media platforms, or a payment service) is valuable to spread as much as possible because it allows processes to be more seamlessly integrated, allows the scope of the infrastructure to reach more and more of the planet, and allows users to all be reconciliated in the same service, which is particularly important for social media, because no one wants to keep switching between them just to talk to a few people at a time.
Another factor for diminishing returns is something I have mentioned in the post on the delusions of the myth of progress: science and technology also run into them, even when they do not run into resource limitations yet, because the ideas we come up with tend to follow the path of least resistance. We work on the lowest hanging fruits first, because they are the easiest to reach for, because they provide us with social benefits in the short-term, and because we often need a scaffolding of ideas to work on the more subtle, abstract ones.
Thus things get harder over time simply because we exhaust the easy solutions and ideas first. This is a trend, this is not an absolute, yet I would say that it is a reliable principle. As a result, it is quite irrational to expect that millions upon millions of people haven't reached for a low-hanging fruit when it comes to difficult societal problems. If a huge societal problem remains in place after many, many years, it signals that there are good reasons for it being so.
A last phase which might occur after the phase of diminishing returns is the decline part, where a technology is phased out by something superior or more convenient, where a service has stopped being useful enough for people to want to spend money on it, that type of thing.
In fact, the way I see the general trend of society is in such a rise and fall fashion too, which is why I write about societal collapse again and again. My intuition is that, as a collective, we are in the decline phase of our society. One of the main limiting factors that any society depends on to build and maintain itself is energy, which is needed for basically everything. Our demand on energy keeps increasing year after year, as the past shows, and as the modern development regarding data centers also confirm, but our supply mainly comes from the cheap, energy-dense fossil fuels, especially oil, which are inherently non-renewable. We are starting to see how we cannot simply wish energy and progress to manifest out of thin air however, as our declining supply meets increasing demand.
One of the main counters to the peak oil view is that the recent problems with regards to oil are due to politics and not the fundamental nature of oil. This is very misleading, because political decisions are made based on resource constraints. It's not a coincidence that all of the notable conflicts, or attempts of conflicts, in the global stage, from 2020s onwards all had to do with fossil fuels of some sorts: Russia, Greenland, Venezuela, and now Iran. Energy is king, which is why politics tends to reflect it.
But that counter does reflect something important: even if energy is the limiting factor, and its non-renewable nature at the scale we use it at dooms our civilization to decline eventually, what will actually lead to societal collapse is up for grabs. Who knows what will cause our downfall? Nuclear warfare perhaps, or severe droughts, or internal political implosions and civil wars, or an actually deadly pandemic? Who knows, but the trend certainly isn't going up. Limiting factors tell you what cannot happen, such as how it's impossible to build a perpetual motion machine, but they cannot tell you what will happen.
So the general trend of our society is downwards for now, reflecting the rise and fall patterns of all the previous civilizations before us. As for the specific technologies that we depend on for our economy, by and large my assessment is that most of the crucial ones are already in the phase of diminishing returns, in that they won't become noticeably better in the coming decades. Think of the technologies that our supply chains depend on: trucks for transportation, excavators for mining, drill rigs, pumpjacks and refineries for oil—the technologies for fracking and oil sands are newer, but they also have a significantly lower yield due to the products they have to work with; it is not an innovation in any meaningful way, it is simply more advanced technology desperately thrown to keep the bottom line where it is at—furnaces for smelting ores, cement kilns, glass kilns, assembly lines, all of those 2 crucial aspects of our supply chains come from many decades ago, and they won't magically become better, because if they could, the incentive structures for massive companies to make them better would have kicked in.
All of the new shiny toys that we have, such as computers, solar panels, wind turbines, electric vehicles, touch screens, and all of the software that runs on those, are certainly not useless, though even then it is worth noting how they aren't that new all things considered, and run into tons of bloat when it comes to software and computers, but they are fundamentally reliant on the 'old' supply chains, the latter being only viable because they run on cheap and energy-dense fossil fuels. You can't mine with electricity, nor can you transport all of the materials cheaply, and smelt ores, and make glass, or cement. To the extent that you can solve those problems technically, the viability of such a 'transition' in terms of energy and scalability are something else entirely. Just because it is technically possible to mine and transport iron, and make steel from it, all with electricity, doesn't mean that it is profitable, and it doesn't mean that it is scalable.
So this is my view on the diminishing returns on growth. There is an entire set of arguments that come from Joseph Tainter's book, the collapse of complex societies, where the idea of diminishing returns on complexity is at the forefront of the book, but I don't really feel the need to add more arguments because the conclusion remains the same for me. Sure, I could analyze how governments become bloated over time, how institutions become ineffective, and how complexity makes it impossible for anyone to even understand what is going on in their society, but again, the conclusion remains the same: our society won't keep growing forever. Who could have predicted that you can't keep growing forever on a finite planet?
1 This aspect of the limits of growth made itself very apparent to me when I came across some hilariously bad figures related to migration made by random Facebook users. They were tracking what would happen to a country if the rate of migration, and the rate of reproduction of migrants were constant, and it led to an absurd conclusion, along the lines that in 30 years from now, the population of a single European country would be in the ballpark of hundreds of millions of people. The figures I am presenting here are nonsensical because I don't remember nonsensical ideas to begin with, but you get the gist: assuming that things grow at the same rate forever leads to tons of absurd predictions about the not so long future.
2 It could be argued that batteries also fall in this category of technologies that are in the phase of diminishing returns, because of how old the foundational idea is, though the specific batteries that we find around us are indeed newer and different. But again, it's difficult to imagine that they could become 10x to what they currently are, but such massive improvements would be needed to make them competitive in terms of weight-to-energy ratio compared to stuff that runs on fossil fuels.
Go back to the list of blog posts
Collapse Diminishingreturns Mythprogress Exponential Inequality Struggleclass Echochamber Peakoil
2026-06-27