[gowld]

LLM's fail because their input data is limited in dimension compared to humans (text, pictures, audio, video) and because their capacity to rewire their own brains is limited to the transformer architecture.