This book chapter discusses the evolution of social simulation, starting from classical agent-based models where agents interact based on explicitly defined behavioral rules. It progresses to AI-enhanced simulations using Large Language Models and ultimately leads to Social Digital Twins: high-fidelity, data-driven representations of real-world socio-technical systems.
Along this trajectory, we explore the main methodological foundations, applications, advantages, and limitations of each paradigm, highlighting the gradual shift from abstract models designed to investigate general social mechanisms to increasingly realistic computational representations of specific social systems.
This evolution not only enhances the accuracy and usability of simulations but also provides new perspectives and tools for understanding complex social phenomena.
Blogger's Review: With technological advancements, the research direction of social simulations is expanding towards more complex and realistic systems. The emergence of digital twins marks a significant step in simulating social dynamics. Future research will further integrate AI technologies to enhance the intelligence and application scope of these simulations.