Large Language Models as Simulated Economic Agents: What Can We Learn from Homo Silicus?

TL;DR

This paper explores using large language models as simulated economic agents, demonstrating they can mimic human behavior in economic scenarios and offering a new tool for understanding human decision-making.

Contribution

It introduces a novel approach of employing LLMs as implicit models of humans for economic simulation, bridging AI and behavioral economics.

Findings

LLMs produce qualitatively similar results to traditional economic experiments

Differences in LLM behavior can generate new research questions

The approach offers a new method for studying human decision-making

Abstract

We argue that newly-developed large language models (LLMs), because of how they are trained and designed, are implicit computational models of humans -- a Homo silicus. LLMs can be used like economists use Homo economicus: they can be given endowments, information, preferences, and so on, and then their behavior can be explored in scenarios via simulation. Experiments using this approach, derived from Charness and Rabin (2002), Kahneman et al. (1986), Samuelson and Zeckhauser (1988), Oprea (2024b), and Horton (2025), show qualitatively similar results to the original, and when they differ, it is often generative for future research. We discuss potential applications, conceptual issues, and why this approach can inform the study of humans.