Safety6h ago
Virtue Ethics Proposed For AI Alignment
This essay argues that rational agents should align actions to practices rather than fixed goals. The author challenges the Orthogonality Thesis by applying virtue ethics to machine behavior. This shift moves alignment focus from objective functions to behavioral dispositions. Practitioners should consider how practice-based frameworks reduce the risks of goal misgeneralization in autonomous systems.