One factor in favor of the use of LLM as a learning tool is the poor quality of documentation. It seems we've forgotten how to write usable explanations that help readers to build a coherent model of the topic at hand.