don't quite get the general motivation, is the question "what types of monadic / effectful computations make sense to embed in a dataflow computing paradigm"?
in particular, most baggage requires baggage polymorphism and baggage subtyping to be practically useful, and those are two concepts that always seem to confuse everyone. certainly that's what everyone complains about re: lifetimes in rust
no shade to these languages, my go-to language is full of such baggage (rust), but idk if people are ready for how complicated our types are going to get. and this isn't even dependent types
"X% of code written by AI" seems like a silly metric without asking: - how much of your code is written by other people - how much of your code is written by the compiler
"we need a manhattan project for AI" no what we really need is a manhattan project for porting every math TeX package to Typst (and please start with mathpartir)
