Click here to download PDF copy of presentation

Speaker: Max Henrion (Lumina)

Abstract:

How can we use generative AI to assist in building decision models?

Large Language Models are radically changing our ideas of what software can do -- from traditional tools that are reliable but need complete and specific quantitative specification to new tools that can work with ill-defined qualitative problems but are unreliable and prone to errors (hallucinations) -- somewhat like us. How can we use them effectively to help define and structure complex decision problems? I'll give some examples with surprising results, in terms of what they can do, with surprising benefits and challenges.