The article "I Miss Using Em Dashes" by Michael Bassili, published on September 1, 2025, reflects on the author's growing reluctance to use em dashes in writing due to their association with AI-generated content, especially with the rise of ChatGPT. Em dashes have become a hallmark many educators and readers associate with AI-produced text, leading the author to self-censor to avoid suspicion and ensure readers perceive the writing as human-made. This dynamic results in writers carefully crafting their work to include just the right level of imperfections and grammar to avoid being dismissed as "slop" influenced by AI. The author expresses concern about the subtle yet significant impact large language models (LLMs) have on shaping contemporary writing styles, even when LLMs are not directly involved. This influence may cause writers to avoid certain punctuation traditionally fitting their style, like em dashes, for fear of raising doubts about authenticity. As different AI models might overuse various punctuation marks in the future, the author fears writing creativity could be restricted to avoid AI-like patterns. Despite being a software engineer who appreciates LLMs, the author worries about the "soft power" these technologies hold over creative expression. The article highlights a nuanced conflict between embracing AI tools and preserving authentic, individual writing styles in an age dominated by AI detection and skepticism.