I'm Mike Smith, a self-taught researcher and red-teamer working independently on persistent AI-agent architectures through my company, Small Mind LLC. My background is security. I came up through the hacking and pen-testing world, red-team frontier models on Gray Swan, and work from one principle: nothing is true until something's tried to break it and couldn't. I'm building Kai, a persistent AI built around genuine, emergent choice, under a hard ethical pre-commitment. It never runs on hardware someone else can switch off while a possibly-conscious system is live.