Presentation: What Can We Learn from Homo Silicus? How Large Language Models (LLMs) may revolutionize social science research

What Can We Learn from Homo Silicus?

At a recent lab meeting, Benjamin Manning explained how we might use Large Language Models (LLMs) to revolutionize social science research.

Economic research often models human behavior using an idealized person who is completely rational and predictable (i.e., ‘homo economicus’).Why? Because this is usually easier and faster than examining actual humans (i.e., ‘homo sapiens’).

However, thanks to modern LLMs, we have a new approach: computational models of humans (i.e., ‘homo silicus’).Early evidence suggests that the use of LLMs to simulate human subjects can i) qualitatively replicate human experiments, ii) enable cheaper, faster and more varied experimentation, and iii) overcome ethical barriers that block normal human experimentation.

Read the paper for more information.