I am interested in projects that liberate the knowledge contained in the big models, which have been trained on our collective output, into formats that are accessible and useful for everyone (and on which humans and machines can collaborate)

Simplest possible version: a website where people post their conversations with AI -- does that already exist? (other than reddit ...)

Working on a system that allows people in an enterprise to create knowledge and AI can help organize and retrieve the knowledge. Is that something you are thinking about?

it's not the purpose of RAGs as they extends the capabilities of LLMs to specific domains / knowledge base ?