The piece argues that people should not rely on AI to write because writing is a thinking process that uncovers and refines ideas. The author uses AI tools daily for coding, research and data work but insists that having an AI generate a document from bullet points short-circuits the necessary mental effort. AI can produce polished prose but cannot originate deep insights or reliably surface the most important arguments; handing over that thinking reduces the writer’s value and yields verbose, unfocused output rather than a concise, impactful product.
The stance is not absolutist: AI is valuable for preparation and review. Use models to sift and structure data, find patterns, or to pose critique-like questions about a finished draft so the human can re-enter thinking mode. Treat AI as a research assistant and a proofreader, not as the author; don’t ask it to create core content from your ideas. The recommended workflow is human-first ideation and drafting, then selective AI-driven analysis and feedback to sharpen the work.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.