In my limited experience the issue is often that the "chatbot" doesn't even check what it says now against what it said a few paragraphs above. It contradicts itself in very obvious ways. Shouldn't a different algorithm that adds a some sort of separate logic check be able to help tremendously? Or a check to ensure recipes are edible (for this specific application)? A bit like those physics informed NN.
I think driving your car to the ground is the most environment friendly approach anyway regardless of the type of car you'll buy next.