On March 29, a writer posted a technical draft on LessWrong, hoping to share progress. The piece was 80 % original but ran through an LLM engine for grammar checks. The platform flagged it as likely AI‑generated and rejected the submission, sparking a debate about authenticity.

Before 2023, the author’s English—his fourth language—flowed naturally, requiring only occasional edits. Tools like Grammarly and Quilbot had been the last line of defense. Now, even a 1,000‑word essay feels dependent on AI suggestions, eroding the writer’s confidence and the distinct voice that once defined his work.

The rejection served as a wake‑up call. The writer realized that overreliance on automated polish strips away raw emotion and authenticity. He now sees the value in preserving his own cadence, even if it means accepting occasional grammatical slips. The goal is to re‑claim a voice that feels genuinely his.

To regain confidence, the author plans to draft without AI assistance, read aloud to catch rhythm, and edit manually. By embracing imperfections, he hopes to restore the creative spark that once made his poems and essays resonate. The next step is a fresh, unfiltered draft that reflects his true style.