Philosophy as Architecture: Deriving AI Safety from First Principles Through Buddhist Philosophy

r/artificial
AI Safety AI Research

## Abstract We present a framework for AI safety in which safety properties are enforced by software architecture rather than model