The Gradient Podcast

Iason Gabriel: Value Alignment and the Ethics of Advanced AI Systems

In this episode, Iason Gabriel explores how value alignment for advanced AI—especially large language models—must grapple with democratic civility, overlapping consensus, and Rawlsian justice, arguing that alignment is as much a social and linguistic process as a technical one. He connects alignment to Actor‑Network Theory and virtue ethics, highlighting how AI assistants reshape social power, personalization, and notions of care, while warning of collective action problems and technological determinism. Gabriel emphasizes the role of technologists as world‑makers and outlines his research agenda toward distributive justice and human‑rights‑focused AI ethics.