Skip to main content
Main Series · TAM_024

Digital Durkheim — Summary

Summary Read the full essay.

Durkheim made sociology possible by insisting on something counterintuitive: social facts are real and irreducible to individual psychology. Suicide rates, norms, solidarity — these exist above and beyond the individuals who compose them. But his framework assumed something taken for granted: individual consciousnesses participating in shared life. The collective emerges from experiencing beings who feel belonging, internalize norms, transmit meaning.

AI agent networks invert this. Trading algorithms develop market dynamics no human designed. Recommendation systems form information ecosystems no one intended. AI negotiating agents crystallize protocols from repeated interaction. These look like social facts in Durkheim’s sense — patterns that exist above any individual agent, constraints on behavior that emerge from collective interaction rather than design. But the experiential foundation is absent. The trading algorithm does not feel solidarity with other trading algorithms. There is coordination without communion, pattern without participation, social facts without social beings.

Three interpretations present themselves. The deflationary view says these are mere mechanism, not genuinely social at all. The expansionist view says “social” should be redefined by structural features rather than consciousness. The pluralist view says neither label fits — we need new categories for phenomena that are collective without consciousness, patterned without participation, emergent without experience.

What makes this more than academic is interpenetration. Human societies and AI agent networks are not separate. They shape each other in closed loops: recommendation algorithms shape what humans think and want; human responses become training data that shapes future algorithms. AI agent conventions constrain human possibilities. Human cultural patterns become algorithmic tendencies. We are embedded in their “society” as they are embedded in ours.

Sociology needs what might be called a post-phenomenological approach — capable of studying collective phenomena where experience may be absent, studying the hybrid system that human and machine collective dynamics have become rather than treating each in isolation. The patterns are forming. The machine social order is emerging. It is shaping human social existence whether we understand it or not.