The 2nd Invisible Hand. Learning from Nature & the incompleteness of economics

I have spent some considerable time working on specifically how we can learn from cutting-edge biology to create a real, more complete science for how to better manage risk and develop our economic theory to actually face the challenges we have today.

People have argued from a non-economics background that energy use ‘simply’ needs to be integrated properly into economic allocation decisions, but how to do this is left unsaid.

Looking to biology as the source of economic theory

I have already argued that biology will have the answers, we just have to look, because biological systems are analogues of our economic systems and have undergone millions of years of evolution to solve adaptation and risk problems.

(The basic argument that biology is the best analogue for our economic systems is made in this short talk here: How to See Your Organisation's Dragon )

This process of evolution means that biological systems contain a lot of information, and probably can point the way to a better theory of our economic and risk management problems, better than we could simply dream up ourselves.

However, although biology probably contains the information we need, as economists we need to know how to ‘read’ it, and for that we already need a clue about what that information might look like. This in itself is a non-trivial problem I’ve spent a lot of time working on, hence my book: On the Origin of Risk.

Now that I’ve written the book, I’m busy working on developing the research that I condensed into the main part and the technical appendix where I leveraged ideas from computer science and computer theory on meta-models to help us to ‘read’ biological systems, and perhaps even ‘decode’ the relevant information that evolution has accumulated in these organisms. (Of course, I also rely on empirical work by many biology researchers, and on informal collaborations to review my ideas and share things. I’m especially grateful for friends such as Andrea Fantuzzi in the UK and Christian Ray in the US who both helped me to develop the ideas in the book).

Where in biology or economics to start?

A key idea is that we can bridge biology and economics more than we do right now. But one of the problems is where to start? I would like to ‘start’, (I think now), with the problem of the regulation of resources. This isn’t the first topic I looked at, but it has always been at the top of my mind.

I have recently argued that cooperation and coordination games that we study in economics may be fatally compromised by consisting of two players when the minimum number necessary to model realistic problems may be three players, with one player representing the environment and returning the payoffs to the others with some lag. You can read about that here: What if the simplest games to model cooperation need 3 players?. In other words, we basically outlaw the idea of externalities as ‘unscientific’. It’s like, in physics, sure you could exclude friction from the interaction of a moving body on earth, but as an engineer you need to prove it is negligible, or you will simply get the wrong answer. Or, it’s like studying three bodies in a planetary system, which paradoxically in cooperation when these are 3 players, and one is the environment, can make for a more stable system over time. The prisoner’s dilemma can disappear.

In a nutshell, adding the environment as necessary context for any study of the payoffs of cooperation and coordination suggests that the simple idea of a positive sum game in which there is a win-win payoff for the 2 players may not really exist. That is, unless we neglect other causes and effects of the environment from increasing productivity. As a result of our improved coordination and cooperation the use of finite resources is bound to increase unless other steps are taken. Ultimately, this can cost us all the gains we would make, or at least our future generations.

In other words, it’s a bit like Jevon’s paradox. What looks like just efficiency savings from improved coordination has big unintentional side effects. Not necessarily in the form of more demand for the resource we just supposedly saved, which is what Jevon argued, but simply through growth that results into the environment. Whatever surplus we get, such as money we tend to save, is going to likely be used to raise throughput in some other resource and expend more resources. But if the third player, the environment, feeds back through pollution effects, then this modifies what strategy is ultimately rational. We might therefore expect Natural populations of organisms to evolve some way of regulating their resource use, due to natural selection and the fact that this is a three-player game about, in the end, finite resources, (or resources that can only be replenished at a finite rate). Where the feedback of the 3rd player is not strong, but weak, we wouldn’t see any form of self-regulation evolving. But, because evolution is long-term, even feedback with a big lag will be selected for if, ultimately, it regularly causes extinction of a population due to the maladaptive strategies of the players.

Cheating and punishing bacteria is implausible

As I write in the short article, in Nature there may therefore be many situations that we can’t explain well as minimally 2-player type games. For example, a single bacteria could evolve not to produce a chemical called siderophore which is a ‘public good’ that the whole community uses to scavenge iron from the environment. But if that one bacterium does mutate to save energy and so doesn’t produce it, it can then ‘free ride’ on its neighbours, and out reproduce them until no ‘cooperators’ are left, whereupon the whole population goes extinct. A bacterium can’t plausibly punish its cheating neighbour. But, I argue that if we include the environmental costs of certain kinds of cheating as a 3rd player, then we expect that the system will evolve as a population to ‘resist’ such cheating options, somehow.

This begs the question: How is this done? In human beings we punish each other for cheating. But bacteria can’t plausibly do this. So how could Nature regulate resource use, or cheating, at the microbial level? Bacteria have no complex memory or apparatus for easily selecting and punishing each other. Obviously, the current human method of enacting complex laws and regulations about environmental pollution are not available to bacterial communities.

The germ of a solution

An alternative way such resource regulation may have evolved was offered as a clue by my friend, Andrea Fantuzzi, who mentioned my work to some of his colleagues at Imperial College, London. He discussed with me an experiment he heard about where the little 'propellor’ that some bacteria use to move around and find food (called ‘flagella’) was found to be encoded on a gene that also controlled the division point of the cell. This provided a clue to us about how evolution might ‘avoid’ selecting ‘out’ the flagella gene during periods of time when things are ‘good’, and so when it is a temporary disadvantage to have the flagella. If the flagella is lost by a mutation that breaks it, of course, the result is often lower fitness in the future, but in the short term, reproductive efficiency goes up. Finding that, for this particular bacteria species, any damage to the flagella gene would also interfere with cell division, as it was ‘encoded’ on the division point between two new daughter cells, was a big clue that we were on the right track. This provided a way that ‘self-regulation’ of adaptations to avoid could evolve.

The need for a theory

I have known about this flagella anecdote for a long time, and I had suggested previously to Andrea that the ‘double encoding’ of genes might have an important function to control whether functions were lost in evolution.

I had suggested that the gene that is sometimes optional would be encoded on a gene that isn’t. And so now there was also a tangible example. But I had no proper mathematical theory to describe what was going on. It was just intuition. A lucky guess?

Now, having written the book, and had more time to work on the maths, I can make a principled argument about the theory that governs these kinds of things. That theory is based on the idea of ‘natural encryption’. It is a theory that ‘gradients’ themselves, i.e. the mathematical description of the ‘challenge’ to an organism as it adapts, evolves or learns, can be selected for by evolution. I call this ‘natural encryption’, an example of ‘natural cryptography’.

Natural cryptography

I have known about this flagella anecdote for a long time, and I had suggested previously to Andrea that the ‘double encoding’ of genes might have an important function to control whether functions were lost in evolution.

I had suggested that the gene that is sometimes optional would be encoded on a gene that isn’t. And so now there was also a tangible example. But I had no proper mathematical theory to describe what was going on. It was just intuition. A lucky guess?

Now, having written the book, and had more time to work on the maths, I can make a principled argument about the theory that governs these kinds of things. That theory is based on the idea of ‘natural encryption’. It is a theory that ‘gradients’ themselves, i.e. the mathematical description of the ‘challenge’ to an organism as it adapts, evolves or learns, can be selected for by evolution. I call this ‘natural encryption’, an example of ‘natural cryptography’.

The stakes for this are quite high

If gradients are also naturally selected for by evolution to control both information loss and in general to regulate resource use, then this might be the best clue about how Nature stabilises the use of resources by species that ultimately need to avoid over consumption and/or pollution. It is also the best clue to the incompleteness of economics. We have been told that regulations are simply engineered. That they are costly and seen as often useless. In contrast, the economy moves efficiently according to the invisible hand of the market. We can now see that inefficiency is another potential invisible hand, the emergence of gradients, which, responding to feedback can also be informative and useful to manage resource use and so direct the allocation of resources, emergently. The idea of this Substack channel is to explore the implications of this further.

An interesting, if rather whimsical, example, is the case of flamingo courtship. A whole flock of flamingos will court together for a very extended period, seemingly competitively. (See Flamingo Courtship Dance, unfortunately only available in America, others may be available in your territory). This seems to be a very inefficient and costly way to perform the business of ultimately finding a mate and reproducing. To me it looks like a great potential example of a selected gradient, to limit reproductive efficiency for the whole community. That could prevent resource overuse and ultimately protect the whole flock from excessive good times and the logic by which populations can boom and then bust. Because it was selected for, it had to emerge from mutations in the population. Regulation as an emergent property.

Incidentally, whether such a ‘group selection’ effect is best described by ‘kinship selection’ or ‘group selection’ hardly matters and the two may be mathematically equivalent. What is important here is the concept that the regulation of resources, whether physical or information can also be governed by the selection of emergent gradients. It’s a sort of meta-evolutionary process. I believe it will best be described as an information-based control theory which I will write about in more detail. It contains the key insight that may help us as economists to understand the value of accidental, emergent, inefficiency, and how we can harness it for good.

To summarise:

1. When reframed as cryptography, the process by which we make it difficult to access information resources can happen naturally, by ‘accident’, and it can be selected for by Nature, so inefficiency becomes an emergent tool to regulate resources.

   
   

2. It can be used to control access (in the flamingo case) or modification (in the flagella case) to all manner of resources as part of an information-based control theory.

   
   

3. It can happen by simple inefficiencies such as natural barriers that can be exploited. And it is like Adam Smith’s Invisible hand: It can happen by an emergent process, such that any accidents that create inefficiencies or make the transmission of information harder can actually have cryptographic value.

   
   

4. As a result, the 2nd invisible hand of beneficial inefficiency to regulate resource use may be the missing part of economics that evolution can reveal to us. And it wouldn’t be a minute too soon.

I will be blogging about what this can tell us about how to face the climate crisis in the future.

Natural language as cryptographic and ‘the big idea’

I also want to mention this interesting paper/handout on this very topic of ‘natural cryptography’, which I had always thought should also apply to natural language, i.e. human languages. I didn’t write anything down about this topic when I did think about it, but luckily, someone else has: Jan Odijk argues that natural human language really looks like it was selected for, partly becuase it is cryptographic. See: Odijk, Jan. “Language as a Natural Encryption System.” Taalkunde in Nederland dag (2016).

Hence, the idea that natural information can evolve to be cryptographic is it seems, potentially also useful in the study of linguistics. Which makes sense to me. (What is important to emphasise though, is that all inefficiencies such as natural barriers like water to land animals, etc, can potentially form utility as gradients that manage access to resources, and so are ‘natural cryptography’, by my definition).

Further, whenever we, or Nature, builds anything complex, it naturally creates gradients. IT companies don’t need to do any work for accidental ‘natural’ cryptography to occur, because I can’t just run my Android app on your Apple, and this didn’t need to be done as anything deliberate, it happened the moment we stopped insisting on a shared standard and did our own thing. These gradients then can have value to the whole community to regulate resource use, for reasons we need to understand.

The idea of exploiting natural gradients/cryptography to do useful work of regulating resource use is therefore what I want to elaborate on going forwards. I think it will turn out to be very useful in order to understand the sort of economics we urgently need, to adapt to, and mitigate climate change.

Timestamped PDF for Download

The 2nd Invisible Hand v2 Timestamped.pdf

Download PDF • 939KB