Complexity theory, also called complexity science, is the study of complex adaptive systems: collections of many interacting parts whose local interactions give rise to macro-level behavior that cannot be read off from the parts in isolation. Its core phenomena are emergence (global patterns arising from local rules), self-organization (order forming without a central controller), and adaptation (the system reorganizing how its parts interact in response to a changing environment). Because the feedback among components is nonlinear, such systems are frequently unpredictable, sensitive to initial conditions, and capable of abrupt shifts between regimes. This is a distinct field from 📝computational complexity theory, which concerns the resources required to solve computational problems; the overlap in name masks a genuinely different object of study, and it is worth keeping the two separate.
Societies, cultures, and religions are paradigmatic complex adaptive systems. They have no single controller, they coordinate through countless local interactions among individuals and groups, and they exhibit emergent, self-organizing structure that adapts over time.
That framing is exactly why I build the models I do. If a society behaves like a complex system, then the way to understand it is to simulate the agents and let the macro-behavior emerge, rather than assume it top-down. My 📝agent-based and 📝multi-agent AI work treats belief, identity, and group dynamics as properties that arise from interaction, not as fixed inputs. Complexity science is the lens that makes that approach not just defensible but necessary.