Complex systems emerge from simple rules
Can you predict your next move?
Can you make sense of what is happening around you?
And within you?
My goals with this article are:
- to give you new questions to contemplate in the shower
- to share with you how I like to analyze the world around me
- to receive feedback and hear what you think
Everything is Emergent
I am here to argue that biology emerges from chemistry, chemistry emerges from physics, and physics itself emerges from a few natural laws and constants.
OK… but what does emergent mean?
It is easier to show it, rather than explain it
These beautiful patterns are emergent,
it’s called Game of Life and it has 4 simple rules:
- Each cell with one or no neighbours dies, as if by solitude.
- Each cell with four or more neighbours dies, as if by overpopulation.
- Each cell with two or three neighbours survives.
- Each cell with three neighbours becomes populated.
In practice at each instant you have a pattern in the grid, to obtain the next pattern (animation) you follow the rules to select which cells will die, which will remain the same and which will be born.
Starting from an initial — often simple — pattern it’s possible to create amazing animations that seem almost alive.
Simple rules and initial conditions give rise to a complex system.
How complex? Well, the game of life is Turing Complete, which means that has the same computational power as the computer/smartphone where you are reading this article right now.
In easier terms, it’s possible to create an initial pattern that recreates the Game of Life… you can have the Game of Life inside the Game of Life.
When I say that biology is emergent, I am referring to the fact that life, at its core, is chemistry. Biology emerges from the interactions of complex molecules, mainly involving biomolecules such as proteins, nucleic acids, lipids and carbohydrates.
And what are these complex molecules made of? Particles that interact according to the rules of physics, chemistry itself is emergent!
Emergent properties are properties that become apparent and result from various interacting components within a system but are properties that do not belong to the individual components themselves
Conclusion
LLM (large language models) like ChatGPT are considered to have emergent properties. We have trained them to predict the next logical/common word and yet they can produce coherent text — and new ideas maybe — they can translate and summarize text, they can chat, answer questions and much more!
Note that emergent properties are unpredictable; the engineers who worked on GPT-3 could not predict how good it would be at solving problems for which it had not been trained. Yet I believe the future of AI will be shaped by these ideas, we are trying to reproduce what we see in nature, and nature is emergent.
We humans are also a complex system, many — if not all — of our actions are emergent. I often wonder where my thoughts or actions come from, what is the input that triggered that specific output, and what rules am I following.
