https://bryonmiller.com/notes/ Notes on AI evaluation, AI security, cybersecurity, and research design. en-us https://bryonmiller.com/notes/why-ai-evaluation/https://bryonmiller.com/notes/why-ai-evaluation/How a cybersecurity background led me toward testing model behavior rather than treating AI only as a governance problem. https://bryonmiller.com/notes/games-as-evaluations/https://bryonmiller.com/notes/games-as-evaluations/Why scenarios, game mechanics, and sequential art may reveal model behavior that static prompts miss.