Kyle and Alex discuss the application of Koopman theory in chaotic systems, such as language modeling, energy forecasting, and atmospheric pollution forecasting. They explore the limitations of linear systems and the potential for embedding language models into the Koopman process.