Did a presentation yesterday about the LLM stuff I've been doing in Agregore at a meetup focused on indigenous usage of AI and local training/inference.

I think it's helped me clarify some of the "why" on this work internally so I'm gonna work on a blog about it.