Hello there.
I was thinking of making a text based scripting library that can be accessed via Python with the task to compile a LabVIEW vi to a kind of pseudo-code text based scripting language.
The first proof of concept would then be to generate an identical copy of that VI from that pseudo code in a python environment.
Not by using the copy function, but by re-building it from scratch using the pseudo-code script.
Additional I would like to explore how AI tools can work on that code (a text based language) to adapt the diagram, where the object locations in the diagram, and wire routing, location, proportions etc, define programming style context.
That style context is the most interesting part, it's the connection between the strict code functionality and the personalized human perception model.
It is a thought experiment to allow the community to play around and experiment with AI potential where the graphical environment is not a disadvantage for AI but delivers an advantage.
(beside the Nigel initiatives of NI itself, which cannot use the full community potential)
Although the programming style doesn't seem to have functional benefits, it could bring AI generated code closer to the human by respecting the human style that makes use of the graphical presentation style to improve clarity depending on the developer's preference.
Representing the programming style into pseudo-code could then be a topic open to the community that could open up a world of innovation that is not possible in other text-based programming languages.
And who knows, this might open up a new added-value of using LabVIEW above text-based languages. After all: Wouldn't it be crazy to leave the graphical programming concept just because the current baby-AI's are not grown up yet?
(I hope I don't cross any propriarity boundries with this concept)
