Self-hosting AI has quietly become part of my daily workflow. I already run multiple local LLMs through Ollama for writing, research, and quick experiments, mostly because I prefer control, privacy, ...