> but it's nothing like the code generating models we're all using today for code and automation. Is this true? Code is structured output. At the very least it

> but it's nothing like the code generating models we're all using today for code and automation. Is this true? Code is structured output. At the very least it seems like a question of degree rather than kind. While the LLMs we're using today are limited to sequenced text, it seems that a model like Jev could excel at coding on a more structural level (factoring, controls) by working within the constraints of an actual language specification and supplemental domain model. I don't know, though -- maybe that's too deep and complex.