Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper investigates whether LLM-based agents can effectively aggregate dispersed private information in prediction markets by observing price movements. The study finds that while AI agents perform well in simple information structures, their performance significantly degrades with increased complexity, mirroring human limitations in reasoning about others' knowledge. Surprisingly, providing feedback on past performance negatively impacts aggregation and profitability, suggesting potential overfitting or miscalibration.
LLMs in prediction markets stumble on complex information, and surprisingly, get *worse* when given feedback on their performance.
Can Large Language Models (AI agents) aggregate dispersed private information through trading and reason about the knowledge of others by observing price movements? We conduct a controlled experiment where AI agents trade in a prediction market after receiving private signals, measuring information aggregation by the log error of the last price. We find that although the median market is effective at aggregating information in the easy information structures, increasing the complexity has a significant and negative impact, suggesting that AI agents may suffer from the same limitations as humans when reasoning about others. Consistent with our theoretical predictions, information aggregation remains unaffected by allowing cheap talk communication, changing the duration of the market or initial price, and strategic prompting-thus demonstrating that prediction markets are robust. We establish that"smarter"AI agents perform better at aggregation and they are more profitable. Surprisingly, giving them feedback about past performance makes them worse at aggregation and reduces their profits.