Idea: Fund TeXmacs development via an OpenAI copilot service

See for example this copilot service for programming:

https://copilot.github.com

For TeXmacs, the dataset could consist of all research papers freely available on the web and perhaps wikipedia as well.