World3 is the model used by the limits to growth to draw its conclusions. It is a program that ran on old computers and had certain constraints programmed that combined with the input data produced a simulation with future outcomes.

[[2024-06-02-20240602081511-brian-hayes-on-world3-model]]

In Daniel Susskind book “Growth” he echoed one of the criticisms that the readers of the New York Times had for the model:

‘[it] takes arbitrary assumptions, shakes them up and comes out with arbitrary conclusions 
 best summarized not as a rediscovery of the laws of nature but as a rediscovery of the oldest maxim of computer science: Garbage in, Garbage out’

[[2025-10-17-20251017154237-growth-daniel-susskind-review]]

This got me thinking. It bothered me too. I thought that the assumptions that limits to growth are necessary, unavoidable. There are many relationship we don’t know their causality, when do they operate linearly, human behavior, outside boundaries, etcetera.

How could it be any different for a book written about growth? Growth causes are, by definition, outside the realm of economics. Is it technology making us more efficient? Or are we extracting more materials? Is it a good weather year all around? Or maybe we found new oil? A new way of working?

Economics register the outcome of growth, but can’t fully explain it. Invisible hands, energy surpluses, productivity increases. They all are vague and they can’t possibly be any clearer.

To achieve certainty over growth we need to include the material world, physics, nature, anthropology and psychology. This country grew X because it consumed Y amount of extra energy, the rate of losses hasn’t changed and this pattern of behaviour has increased perceived utility by Z.

That last paragraph sounds like Garbage in, Garbage out too. It requires many assumptions to be coherent, and it might create something chaotic that can’t predict anything. Maybe anything written about growth will suffer the same fate.

Software is Garbage in, Garbage out

I have seen the same type of arguments while working developing software. Some people want to keep computation in the realm of first order logic, algebra or Turing machines. A lispian model of computation where all the types are well defined, all the code is proven.

While anything outside that realm such as software methodologies, patterns, systems thinking, Domain driven design, team topologies is somewhere between uninteresting and counterproductive.

Here is the same argument than the growth argument: Is software a method to instruct the machine to do things, a science? Or is it a method to build theory as Naur described it?

https://pages.cs.wisc.edu/~remzi/Naur.pdf

This is, in my opinion, the source of technical debt. It is the build up of discrepancies of theory over time. Computer science is not equipped with methods to deal with it because the theory and reality are beyond its boundaries. In the same way that the causes of growth are beyond economics.

It is Garbage all the way down

In both economics and computation I have deliberately drawn a line between the certainty of science and linear thinking against the loose reasoning of systems thinking and its variants. They are two sides of the same coin.

What are axioms if not the recognition of the limitations of a science. It is quite common to hear about “rational choice” as a very questionable truth of economics. Game theory, collaborative behaviour are at odds with it quite commonly.

On computation we can go all the way down to maths and Gödel’s incompleteness theorems. Less assumptions, almost no axioms. But inevitably axioms.

In both cases it is a forced perspective that is more coherent than the broader view, but does not guarantee truth. The more uncertain and broad the problem is, the less reliable are these scientific approaches. Correctness diffuses.

How to deal with all this Garbage

When facing the complexity of the real world, both approaches are flawed. Systems thinking makes global assumptions. Linear thinking narrows down to the limits of its axioms. Neither can get us closer to the truth.

All left is politics, narratives and beliefs. Which is the best we have. Why is growth a contentious topic? I believe it is because nobody really fully understands it and there isn’t a self coherent theory for it.

Is it that growth is the outcome of our intelligence? Or is it mainly that we found a lot of oil? Is it modern economics? Or is it the amount of immigration? Is it globalisation?

Is it that the software system is too coupled? Or is it that we are not following this methodology? What if we move everything outside, or everything inside? Should we make this a service?

I persuade you, you persuade me. The aggregate of opinions drive the outcome, despite its unevenness. Someone will try to formalise it and change that outcome, moving the discussion to its axioms. But it is all “Garbage in Garbage out”, even the fanciest proven axiom dependent theorem.