Here’s a pair of (formal) papers on rewriting string diagrams, which could help solve problems in graphical programming languages like “how do you nicely display an arbitrary set of merged functions?” (ie level of detail, though “nicely” isn’t defined in these papers, it could be…) https://arxiv.org/abs/2012.01847 https://arxiv.org/abs/2104.14686

Really interesting paper, though a lot of the category theory was hard to decipher for me. I love hypergraph rewriting systems! The flexibility of hypergraph structures, along with their ability to easily represent higher-order structure is great. They're portable between models of computation, can represent semantics, and express structure. They also happen to be incredibly simple. I think that makes hypergraphs (and their mathematical siblings) an incredibly compelling primitive.

Hypergraph rewriting? You follow the Wolfram Physics Project at all @Orion Reed?

@wtaysom I’m aware of it but don’t follow it, I wouldn’t be surprised if they come up with some interesting findings, but I’ve always found Wolfram to be.. idk.. a bit egotistical? Could never figure him out.

It would be cute if they found some really robust correlates. As for the ego, I read "A New Kind of Science" back when it was published. Found the automata interesting, but the prose intolerable. A few years later, I met Wolfram. A little conversation helped me... how to describe it... fashion a pair of neuroatypicality glasses whose filter has helped me enjoy his writing since.