Comment by mikegreenspan

This seems like a really interesting usecase of LLMs. I’m curious how you’ve validated the outputs and what gives you confidence that the forecasts are good and not affected by hallucinations or incomplete information?