A grassroots nonprofit working on the alignment problem from two directions: transparency for military AI and a psychotherapeutic approach to LLM fine-tuning.