I am training codex (an AI model available to chatgpt subscribers, similar to Claude Code) to learn how the VFi Toolkit works.
One advantage of codex is that it can generate comments to document better the toolkit functions and also the replication and example files.
@robertdkirkby: Whenever you have time, could you please save in the pseudo code folder also the tex source file to generate the pdf? In this way codex can proofread the tex and modify it. Then I can do a pull request and you are free wether to accept it or not. Same goes for the documents in lifecycle models, olg, etc.
An alternative to tex is markdown, which one you prefer, they are both equally good for AI.
I feel like people only want to see the pdf, so I set up a separate repo which is all the tex files. This repo is largely just intended for AI, doesn’t seem like something a human wants to see.
If you are setting up pull requests on Intro to Life-Cycle Models or Intro to OLG Models docs, please do ‘lots of small pull requests’ rather than ‘one big pull request’. Makes it much easier for me to read and approve them all (plus if there is one small thing wrong, I reject just that one change and not the whole lot).
You can do whatever you want to ‘Documentation’. It is so out of date it is largely junk, I’ve been meaning to delete it as I don’t think it actually serves any useful purpose anyway.