Home | Abstracts | Accommodation | Call for Papers | Details | Keynote | Organizing Committee | Photos | Presentations | Program | Sponsors | Venue
Archived page — originally published at https://igssworkshop.wordpress.com/. Archived on 25 July 2026 with permission.
Home
Online • June 8th-10th, 2021
What AI can do for Agent-Based Modeling
Motivation
The agent-based model is recognized as the principal scientific instrument of generative social science, the (necessity) motto of which is, “if you didn’t grow it, you didn’t explain it.” In other words, given an unexplained observed macroscopic social pattern—a wealth distribution, a disease time series, a spatial segregation pattern—we seek a micro-to-macro account.
Specifically, we design agents (the micro-scale) intended to generate the macro target, assessing the fit between model-generated and real-world macro structures by means of statistics. This method of agents has been successfully applied in many spheres, from epidemiology to anthropology to economics. Typically, however, agent modelers handcraft the agents, and in particular, the agents’ rules of behavior.
Even when a particular model (e.g., the artificial Anasazi model) succeeds in growing the target, it is only one explanatory candidate, leaving open several questions: Is this solution unique? Harking back to the motto of ABM, though we have explained it, there may be many ways to grow it. Can we find a more complete set of rules? Relatedly, how robust is the solution to a small change in the agent rules?
If we could discover a “neighborhood of” agent models whose members all generate the target, the result would seem less ad hoc and unstable. These are the core concerns of the nascent field of Inverse Generative Social Science. The essential difference from traditional ABM is not to craft entire agents, but rather, to encode the space of possible agent constituents (rules, parameters) and possible mathematical and logical concatenations, and search this large space for the fittest agent architectures using Genetic Programming, Decision Trees, Causal State Modeling, Associative Rules and other techniques from Machine Learning and AI. Agents thus become outputs of the model, standing the prevailing “paradigm” on its head.
If the vision of Inverse Generative Social Science is achieved this will be a watershed for agent-based modeling and for social and biological science more generally.
Featured Article
Introducing Inverse Generative Social Science
Karl Naumann, 18 June 2021
For three days this week I have been glued to Zoom, taking part in the Inverse Generative Social Science (IGSS from now on, its a mouthful) workshop 2021. Work was put on hold. Noise-cancelling headphones were fixed to my head. It was a blast of information and new perspectives, everything from foundational frameworks and questions to specific model implementations. Unfortunately, the framework of generative social science (GSS) is not a commonly taught one (yet?). This is my attempt at a somewhat coherent introduction to these ideas. [Read more on Karl’s website]