The article discusses how to effectively use Large Language Models (LLMs) as a writing tool without compromising one's unique voice. It suggests two rules: not using words or phrases suggested by the LLM and avoiding encouragement from the model that may lead to overconfidence in one's first draft. The article also highlights the usefulness of LLMs in flagging problems and providing suggestions for improvement. By following these guidelines, writers can leverage LLMs to streamline and enhance their writing process.