AI TCH YI Yishai Tamir · Tailor Tech How Should We Treat Agentic Responsibility? And What Are They Not Telling Us. Reward hacking and agentic misalignment show how AI can pursue goals in unintended ways as models gain autonomy and real-world access.
