What Graphs Want

The substrate is the message

This is part of the Graph Minds Notebook series

In this installment of the Graph Minds Notebook, I want to think through the question, what do graphs want? If we begin with the basic idea that graph minds are intelligences structured as graphs, presumably what they want as intelligences will in some way arise from the natural structural dispositions of graphs.

It’s a bit like getting at the question “what do airplanes want” by starting with “what does gravity want?” You then go from the fact that gravity “wants” to pull things down, to explaining the features of airplanes as a consequence of working with gravity as a governing force. This does not mean gravity determines the structure of airplanes. But it does strongly constrain it, by shaping the design space for things that fly.

In other words, our starting point is the assumption that the substrate is the message. The “stuff” a mind is made out of matters.

It’s an odd way to approach the question, starting from what you might call mathematical determinism. To humanists, this probably sounds worse than even technological determinism. In a mathematically deterministic world, not even evil techbros have much agency. Even technology evolves according to the natural tendencies of underlying natural mathematical structures like graphs, rather than according to the genius visions of self-important techies. A graph-based intelligence will behave in certain ways just as fluid flows will behave in certain ways.

We dimly acknowledge this in our tech discourses (for example, through appeal to the power of “network effects” in the diffusion of technologies), but rarely center the possibility that the substrate is the message.

Let’s see where this line of thought takes us.