Can LLMs Replace Economic Choice Prediction Labs? The Case of Language-based Persuasion Games
The episode discusses a paper exploring if Large Language Models (LLMs) can replace human participants in predicting economic choices. Hosts find that LLM-generated data is highly effective, but the best results come from a 'DUAL' approach: using limited human data to fine-tune the LLM, and then mixing that synthetic data with the real data for prediction.