Behavioral Economics of AI: LLM Biases and Corrections
It addresses biases in AI decision-making for applications in economics and finance, but is incremental in applying existing human bias tests to LLMs.
The paper investigated whether large language models (LLMs) exhibit systematic behavioral biases in economic and financial decisions, finding that biases vary with model advancement and size, and that prompting can reduce them.
Do generative AI models, particularly large language models (LLMs), exhibit systematic behavioral biases in economic and financial decisions? If so, how can these biases be mitigated? Drawing on the cognitive psychology and experimental economics literatures, we conduct the most comprehensive set of experiments to date$-$originally designed to document human biases$-$on prominent LLM families across model versions and scales. We document systematic patterns in LLM behavior. In preference-based tasks, responses become more human-like as models become more advanced or larger, while in belief-based tasks, advanced large-scale models frequently generate rational responses. Prompting LLMs to make rational decisions reduces biases.