#17 If there is a limit to technological progress, we have to become responsible spacemen
...and the evidence is piling up
I'm sorry for the newsletter delay. I've been down with a cold, fever, and jaw inflammation. It might result from pushing myself too hard—I'm sure you've been there. Thanks for your patience.
In one of my fever dreams, I was trapped in a vast universe, hovering in a spaceship through a seemingly infinite galaxy. As a spaceman, I was on a quest to explore technological boundaries and extraterrestrial civilisations. Alongside me were economists shouting, “There are no limits!” and “We can stand on the shoulders of giants to make progress!” I was the ship's captain, nervously watching our depleting resources. Despite the endless view outside, our ship was running out of everything. “Don’t be so negative; we’ll find a solution! Let’s see what AI tells us”, my fellow travellers commented. My fevered mind suggested that boundlessness might be the clearest sign of limits.
Like in most of these fever dreams, it went on and on with no solution, no conclusion, dragging on whether there were limits. Furthermore, there was much confusion (and sweat)—like every day. But my fellow travellers commented: “Our theoretical models are clear, elegant and consistent. It can not be ignored by reality.” “We can substitute everything as long as the price is right.” “Human ingenuity will help us out!”. But it sounded more desperate than ever. Especially when food and water shortages were more felt than ever, there was no solution, no new civilisation, no life, no water, no living anything visible.
When I woke up, I knew where all this had come from. I have been reading this paper about universal limits to technological developments (I know you shouldn't do it when you are already feverish). Mind-boggling. And very convincing. It might sound contradictory, but the basic idea is that you can conclude that technological progress has an upper limit from a sheer unlimited universe without signs of living intelligence.
There are three reasons for an Upper Limit to Technological Development: Fewer benefits from complexity, higher maintenance costs, and catastrophes that disrupt progress. We see all this happening. Incremental progress requires vastly more resources, and from time to time, we see such significant setbacks (or catastrophes, wars, etc.) that progress is not a given.
So, as the author argues, we might be close to our upper limit.
I think no one at the IMF reads this kind of paper. I was reading through the World Economic Outlook (obligatory reading for economists), and my eye caught chapter 3 about “How to increase growth.” The economic mantra is constantly increasing productivity growth. And, as many studies have noticed, we expect a lot from Artificial intelligence. It is very questionable if that is possible. Another recent paper also makes this case. AI alone may not be sufficient to accelerate the growth rate of idea production indefinitely.
But you might blame my fever state. But describing economic growth as an end instead of a means and an end that is unconditionally good is a pointless exercise. The IMF chapter reads as being so detached from our earthly reality. It pursues growth in reductionist isolation from everything happening in the world—again, temperature records, wars, whatever— a closed-end system. This is not the case, as Kenneth Boulding already wrote in 1966. Our economy is an open-ended system embedded in social and ecological realities. It is still the Cowboy Economy, where resources are seemingly limitless, often leading to exploitation and environmental degradation. And the only way to pursue, to make the system sustainable is to make from that Cowboy Economy the realisation that we are on Spaceship Earth: Earth is like a spaceship with finite resources. This concept requires cyclical ecological systems and careful management of energy and materials.
“The closed economy of the future might similarly be called the 'spaceman' economy, in which the earth has become a single spaceship, without unlimited reservoirs of anything, either for extraction or for pollution, and in which, therefore, man must find his place in a cyclical ecological system which is capable of continuous reproduction of material form even though it cannot escape having inputs of energy” (Boulding, 1966, P7-8)
So, where do we end (or begin)? First, for me, get better. Unfortunately, I still think that my fever nights were not so much besides our reality. But for our global economy (or, better, society), it would be very helpful if we realised that productivity growth is not ‘manna from heaven’ (as the residual in the Solow-growth models was called). It is not a logical consequence of investments in ‘productivity-enhancing’ factors like education, research and development, trade openness or a regulatory agenda aimed at deregulating markets. It might be the case that we can not bet on more macroeconomic productivity increases. Our agenda should be about the realisation that we live on a spaceship, a closed system within our biosphere, and that we have to take care of that spaceship — otherwise, it will crash.
The IMF's old-fashioned agenda
Global growth has been declining since the 2008–09 global financial crisis. Advanced economies started to slow down in the early 2000s while emerging markets and developing countries saw a downturn after the crisis. This ongoing slowdown is mainly due to decreased total factor productivity (TFP), a key measure of how efficiently resources like capital and labour are used. TFP accounts for more than half of the decline in growth.
According to the IMF, several factors contributed to this slowdown:
Misallocation of Resources: Capital and labour are not efficiently distributed among businesses, leading to decreased productivity. This misallocation is due to various frictions within sectors.
Less Investment: Since the global financial crisis, private capital formation (investment) has dropped in many advanced and emerging market economies.
Fewer Working-Age People: The population aged 15-64 has been growing more slowly, leading to a smaller workforce. This decline is visible in most advanced economies and some emerging markets.
Okay, this might be true. But what does it imply? Without changes, global economic growth is expected to be about 2.8% by the end of this decade, much lower than the pre-pandemic average of 3.8%. And according to the IMF, this is a problem. I agree; if you have a growth-dependent system, lower growth means higher debt (especially in emerging markets), more poverty, more inequality, et cetera. That is why we have to change the system.
However, the IMF gives (of course) solutions for higher growth. The chapter suggests policies that can improve capital and labour use and increase labour force participation. For instance, focusing on labour force participation—especially for women and older workers—could boost growth. The chapter also discusses the impact of artificial intelligence, suggesting that its proper use could improve productivity and raise economic growth.
Another set of policies should address structural issues like high public debt and geoeconomic fragmentation (dividing the world into economic blocs). Policies that promote market competition, trade openness, financial accessibility, and labour market flexibility can help overcome these barriers.
Mmm, I think we know this recipe. Did it help us in any way????
If you read the text carefully, they also acknowledge that it is a little utopian to be optimistic about growth (or dystopian, for that matter). They say about technological progress:
Meanwhile, the growth in efficient TFP, which reflects the rate of technological progress, is expected to slow in the baseline scenario, following its long-term trend. Factors such as the increasing difficulty of generating new ideas, slower growth of research employment, a plateau in educational attainment, and the slower catch-up process are expected to play a role. The net effect is a decline in the TFP growth rate by 0.1 percentage point from its two-decade average prior to the pandemic. However, major technological advances, particularly in AI, could increase TFP growth substantially.
This is my bridge to the other two papers I read this week: the universal limit on technological progress and the limits to AI as progress.
An upper limit to progress
The paper "Is there a universal limit to technological development?" might not be the most likely read for an economist. It explores whether technological progress has a ceiling impacting astrobiology, sustainability, and policymaking. It starts with why, although we can technologically leave our planet, we still do not find technosignatures (signs of advanced technology from other civilizations) from other advanced extraterrestrial civilisations. One of the biggest questions. However, it gives insights into smaller but relevant questions about limits to technological advancements.
The core idea of a Universal Limit to Technological Development (ULTD) suggests that every civilization eventually reaches a point where technological progress stops. This limit could explain why we haven't detected any signs of advanced extraterrestrial civilizations. The absence of detectable technosignatures indicates that technology might have a universal boundary.
The paper discusses several reasons for this upper limit.
Decreasing technological returns on societal complexity: As societies become more complex (requiring more people, resources, and structures), the benefits from new technology start to diminish. As the infrastructure and resources needed to support complex systems grow, the costs can outweigh the advantages of further technological innovation.
Increasing maintenance costs: The upkeep of existing technology can become prohibitive. For example, maintaining large scientific projects like the Arecibo Observatory or ensuring nuclear reactors' safety involves significant costs. This rising maintenance cost can limit a society's ability to invest in new technology.
Civilization-damaging catastrophes: Catastrophic events, like asteroid impacts or volcanic eruptions, can cause severe damage to civilizations, leading to technological stagnation or even regression. Recovering from such disasters can divert resources away from technological advancement.
Regarding the Fermi Paradox—why, with so many stars and planets in our galaxy, we haven't found advanced extraterrestrial civilizations—the ULTD concept suggests that interstellar communication or colonisation becomes impossible if civilizations hit a technological limit.
Regarding sustainability and policy, the paper questions the idea that technology alone can solve our problems. Grand projects like nuclear fusion, asteroid mining, or terraforming Mars might not be the answer to global challenges. The conclusion is that Earth is our only home, and we must manage it responsibly. Our spaceship, just like Kenneth Boulding, concluded 58 (!!) years ago. So, our purpose is to become responsible spacemen on this ship.
Overall, the paper indicates that technology might have a universal limit, raising essential questions about the future of human civilization and our place in the universe. It calls for careful management of our planet and suggests that technology might not always be the ultimate solution.
AI is also not the panacea.
It was getting a little (or a lot) philosophical. I like it. As a social science, I think economics should not forget its philosophical foundation (and especially not forget to reflect on it). In the end, it is about the most difficult questions of life. What is a good, fulfilling life? What is humanity’s responsibility to all other inhabitants of this planet? What is our responsibility for the future of humanity? What is just? These are a few big questions that economists operationalise as if they were physics. It is not.
Ideas drive progress, not economic growth, but the production of more goods and services. Conventional economic wisdom is that there is no limit to ideas. But, as explained above, there can be limits to their execution.
This paper stresses another factor: the combination of lower demographic growth (where more people always equated more ideas) and AI. When I was reading it, I thought of it as some micro-application as the paper above.
The paper "Artificial Intelligence and the Discovery of New Ideas: Is an Economic Growth Explosion Imminent?" examines whether Artificial Intelligence (AI) could boost economic growth. This idea comes from AI's ability to generate new ideas essential for economic progress. To explore this, the paper presents three models that show how AI might influence the production of new ideas, called the ideas production function (IPF).
AI is believed to drive economic growth by automating tasks and helping researchers discover new ideas. To understand this better, the paper outlines three models: AI as a technology that allows researchers to work faster, AI as a tool that can handle more tasks simultaneously, and AI as a facilitator that encourages innovation. The paper's simulations suggest that AI, on its own, might not be enough to cause a continuous and dramatic rise in economic growth.
The paper also discusses the role of population growth in economic progress. Traditionally, more people meant more ideas, leading to more economic growth. The question is whether AI can compensate for slowing population growth. Although AI holds promise, there are concerns about its limitations. It could create "stepping-on-toes" effects, where researchers' work overlaps, leading to inefficiency. AI might also increase "mundane knowledge work," tasks that consume time without necessarily advancing discovery.
The simulations show that AI alone might not be enough to create a significant and sustained economic boom. The conditions required for this explosion might be too specific or unlikely to happen. Therefore, the paper concludes that while AI can contribute to economic growth, it probably won't lead to an explosion of new ideas and economic development by itself.
I think that is more than enough for this week.
Take care
Hans
Wonderful article, Hans. Something I find interesting to reflect on is the potentiality of the human civilisation when we finally get a handle on peace. There are signs from philosophy, psychology and brain sciences that the collective de-traumatised human experience could create inventive power that itself is infinite or shall we say very very large. I intuit that it is only under these conditions that certain breakthroughs will occur e.g. efficient and effective space flight and exploration. On this planet however, the many necessary competent characteristics for every human being required to achieve a peaceful planet, will also provide the necessary applications to ecosystem details and flourishing while perfecting new more subtle energy technologies and resource farming. Going by the economic growth formula, this may also show a declining growth. The real question then is, if there is a flourishing ecosystem with a flourishing human planetary society but a declining economic growth, then maybe the whole model is transformed and we are not even using those measures to determine how we are doing. I have no answer for the future except in peace and human collaboration and deep consultation. I do have a sense of the great possibility that emerges from such a future. There is for this the possibility of a state of human designated by the idea that, when we think of 'who I am', could it be that who I am is the showing of everything and everyone in my experience. This leads to an idea of a 'shared brain'. I intuit the human future is infinite but not in the sense that we think of as resource infiniteness.