I remain sceptical about any form of combination of reinforcement learning and LLMs. Acting successfully in the world when faced with complex issues requires l

I remain sceptical about any form of combination of reinforcement learning and LLMs. Acting successfully in the world when faced with complex issues requires learning useful ad-hoc concepts from the specific situation you find yourself in. It's plausible an AI can learn template tactics from large datasets, but I don't think that's enough. There are various fields, like creativity research and design thinking, where it's understood that non-trivial problems need interaction with the environment to frame a problem in a way that allows an approach to a solution. This is because of the uniqueness and novelty in the situation itself. It might be my lack of imagination, but I don't see how a deep learning on a large data set will get there.