It's a necessary assumption for the universal approximation property; if you assume some structure then your LLM can no longer solve problems that don't fit into that structure as effectively.
replyNeural nets are structured as matrix multiplication, yet, they are universal approximators.
replyYou're missing the non-linear activations.
replyBut language does have structure, as does logic and reasoning. Universal approximation is great when you don't know the structure and want to brute force search to find an approximate solution. That's not optimal by any stretch of the imagination though.
reply