Why LLMs Cannot Think and How to Fix It

Marius Jahrens, Thomas Martinetz

This paper elucidates that current state-of-the-art Large Language Models (LLMs) are fundamentally incapable of making decisions or developing "thoughts" within the feature space d…