Future Is Unevenly Distributed: Forecasting Ability of LLMs Depends on What We're Asking
This work addresses the inconsistent forecasting capabilities of LLMs for social, political, and economic events, highlighting domain-specific limitations.
The study examined how large language models' forecasting performance varies across domains and prompts, finding that their ability to predict real-world events beyond their training cutoff is highly dependent on question type and context.
Large Language Models (LLMs) demonstrate partial forecasting competence across social, political, and economic events. Yet, their predictive ability varies sharply with domain structure and prompt framing. We investigate how forecasting performance varies with different model families on real-world questions about events that happened beyond the model cutoff date. We analyze how context, question type, and external knowledge affect accuracy and calibration, and how adding factual news context modifies belief formation and failure modes. Our results show that forecasting ability is highly variable as it depends on what, and how, we ask.