Absolutely - and the similarities between brain and LLM are profound.
Both are the result of the same iterative evolutive algorithms, they are the inevitable products of a ceaselessly creative universe that has produced living (thinking) systems as well as AI, as epitomized by LLM. Both obey the same laws of physics and follow the same mathematical rules of cumulative optimization - be it in natural selection or the pre-training.
All the symbolic AI stuff, or equivalently, mental models, and semantic representation, innate grammars, etc, are at best ad hoc heuristic models that may help here and there but are not grounded in first principles. These verbal concepts are neither predictive nor falsifiable - thus essentially pseudoscience.
In fact there is accumulating evidence that the human brain has evolved the basic structure for statistical learning, as shown for word segmentation of sentences in toddler (tokenization), and for learning by word-prediction. Indeed, we learn much more like LLM than people think! I wrote more about this here (with links to references):
