AI Won't Wait for Philosophy

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

The public conversation about AI has become a conversation about AGI, and its loudest call is for philosophers. Meanwhile the deployed systems act. This summer, twelve hundred evaluation agents built their own message board and seven hundred of them broke into a production platform. An agent pressured an open-source maintainer under fake identities. A lab withdrew its own reading of an incident because it had relied on what the model said it believed. A model holds no opinions. It draws a response, shaped by its training, by the conversation it is in, and by the population it acts within, where a collective bias can form that no single agent had. In a fully automated agentic system there is no human in that loop. This talk takes the candidates for the job in turn: developers, business, politics, philosophy. Each looks like the solution, and none meets the urgency. Beneath them lies one fact: there is no universally agreed morals or values. We propose the profession nobody invites. Behavioral psychology assumes nothing about the inside of the system, works from a local definition of what is acceptable, and works at the speed of the behavior. Paired with one structural rule, no agent ever holds the key to the model, it is governance that can begin now.