The Premise
Everything interesting is a graph.
Not metaphorically. Structurally. The topology of connections — who talks to whom, what feeds into what, which nodes can reach which other nodes — determines more about a system’s behavior than the properties of any individual component.
Neural networks are graphs. Economies are graphs. Ecosystems are graphs. Language is a graph. The supply chain that built the device you’re reading this on is a graph. The thermodynamic argument for why civilizations rise or collapse is, underneath, an argument about network topology.
This isn’t a claim that graph theory is a useful analytical lens (it is, but that’s the boring version). The claim is stronger: graph structure is the primitive. The thing that determines whether a system produces intelligence, collapses into equilibrium, self-organizes into something worth preserving, or falls into an absorbing state it can’t escape — is the shape of its connections.
Two Ideas in the Name
Systems are attracted toward graph organization. Given energy flow through a medium, structure spontaneously emerges — and that structure is always a network. Atoms form molecular graphs. Cells form tissue networks. Organisms form ecosystems. People form societies. The tendency to form connected topology is as fundamental as the tendency toward entropy. It may be the mechanism by which entropy is locally reversed.
The attractors of complex systems are determined by their graph topology. A dynamical system’s long-term behavior — what it settles into, what it can escape from, what traps it — is shaped by the connectivity of its state space. Change the edges and you change the basin. This applies to neural computation, to civilizational dynamics, to the thermodynamics of capability diffusion.
What’s Here
Long-form writing that explores these ideas across domains — information thermodynamics, computational intelligence, the physics of connected systems, and what happens when the graph structure of human civilization encounters tools that radically change its conductivity.
Some of it is rigorous. Some of it is speculative. All of it takes the graph seriously as the thing that matters.