[zwaps]

Interestingly this mirrors other disciplines. Consider for example Schelling's model of segregation in social science, which is (probably not by accident) very similar to cellular automata.

Just a few years laters, people quickly translated this to graphs - a network. The spacial interpretation aside, the idea surely should hold for higher dimensional social relations.

And that's where we reach and interesting point that Wolfram writes about: These system are useful, as long as we can calculate or derive a state. If all we can do is simulate it, then it's much less useful. And indeed, creating such models is the true "art".

So while the "idea" of using relations and hypergraphs lays a foundation that I am sympathetic to, I also feel like the things "on the to do list" are, indeed, the meat of the issue.