Dylan Waldner
AI safety, alignment, and ethics researcher.
About
I'm interested in alignment at the population level, studied through evolutionary framings. I build tools intended to govern populations of AIs all at once.
Current direction
As robotics pushes toward human-level embodiment in the physical world, the question of how these agents will interact with people and with society becomes increasingly pertinent. For general robots to integrate into our society, they need normative constraints that are aware of both context and time, rich enough to capture how the law actually works. I treat robot law as an alignment concern, framing the law as a coordination mechanism that shapes populations of agents through shared norms.