CONCEPT
The Social Graph
The machine-readable map of human relationships—who knows whom, how strongly, and in what contexts—that
Mark Zuckerberg made the epistemological foundation of Facebook and that now carries the AI layer Meta is building into three billion lives.
The social graph is not a database of friendships. It is a claim about the nature of human knowledge: the most important information about a person is not what she explicitly declares but what her connections reveal. A network of relationships carries information that no individual node possesses—the same insight that underlies ensemble methods in machine learning, collective intelligence research, and distributed AI architectures.
Zuckerberg’s foundational commitment was to build the graph on real identity rather than pseudonymity, on the premise that relational data is only valid if the nodes are genuinely who they claim to be. The commitment produced the infrastructure of social connection for an era—and simultaneously created a
surveillance architecture whose power for advertising, political targeting, and behavioral manipulation has been the most contested feature of the digital age. When AI runs through the social graph rather than alongside it, both the opportunity and the hazard scale: AI assistants gain access to the richest context about human