Social scientists long held that religious groups were too complex to forecast. A body of computational work argues otherwise: that fundamentalism, extremism, and radicalization can be predicted on near-term time scales. The approach uses 📝Multi-Agent AI (MAAI) to model the psychological mechanisms of the Information Identity system — the dynamics of 📝Social Identity and 📝Fusion Theory — inside artificial societies. One such system, MERV, was built by a team I led and showed how those mechanisms generate patterns of religious violence that mirror real-world conflict. The work was published in the Journal of Artificial Societies and Social Simulation and covered by major outlets including the BBC. What makes this matter is leverage: if you can simulate how identity, threat, and group dynamics tip a population toward violence, you can test interventions before they play out in the world rather than after. That premise runs through the conflict modeling work at 📝CulturePulse, from Northern Ireland to the Balkans — using simulation to anticipate where social stability fractures and why.
References
I came to this as a cognitive scientist who got tired of explanations that could only ever describe the past. Prediction is a harder, more honest bar. If a model of religion can't tell you something about what happens next, I'm not sure it's telling you much about religion at all. That's the standard I hold my own work to.