This post contains my findings from a collection of works on agent-based modeling and simulation, including Dunja Šešelja and Joshua Epistein’s Generative Social Science. I have also thrown my own paraphrased thoughts in.

Agent-Based Models #

Motivation: Perfect knowledge of individual agents’ decision rules does not necessarily help us predict emergent¹, macro-level results.

What do they show? #

The canonical agent-based experiment is as follows:

“Agent-based models provide computational demonstrations that a given microspecification is in fact sufficient to generate a macrostructure of interest.”