Research
151d agoRethinking AI Goals: A New Alignment View
The essay challenges the conventional goal‑centric view of AI, proposing that both humans and rational AIs should align actions with practices rather than fixed objectives.
The Signal
By redefining agency through action networks, it offers a framework that could reshape global alignment research, influence policy debates, and guide developers toward more context‑aware systems.