it’s becoming increasingly more feasible to have an assistant which has absorbed the entirety of the internet’s knowledge
Yeah, that’s never going to happen, I’m afraid. The models do get denser and better at transformative tasks, but you will simply never be able to ask that 22B 4-quant for the birthdates of obscure but historically important Bolivian politicians. That’s simply about information density and that’s not a useful application for models.
It’s going to be irrelevant, of course, once there’s a convenient 1-click way to integrate your local Kiwix-server into your model’s Open-WebUI’s knowledge base. There’s no need to waste VRAM on Wikipedia and Stackoverflow knowledge.


Been running a server for my friends for over a decade now. Can recommend. It’s just one apt-get to set up, runs on a Pi Zero for a dozen people, has clients available for pretty much any platform and doesn’t really require any maintenance. Latency will depend on the routing between you and your friends’ ISPs, of course, but the whole purpose of the software itself was to provide a low-latency voicechat server for gaming.
But: That’s it. You don’t get anything else. It’s a barebones voice chat server. You can set up rooms and have basic text-functionality, but you don’t get any fancy user management, no full-fledged chatrooms, no persistence beyond the room setup and only limited backend options. Keep that in mind.