“Your intellectual fly is open”: Bryan Cantrill on LLM-written posts

In a post titled “Your intellectual fly is open”, published on his blog on 5 December 2025, Bryan Cantrill makes an argument aimed at anyone who lets a language model write for them on social media. The piece first appeared as a LinkedIn post on 11 November 2025 — a platform Cantrill admits has become unexpectedly important to him, and which he describes as “the Gerald Ford of social networks”: clumsy and a little boring, but not entirely crooked, which he calls a relief.

The complaint starts with style. LLM-written text is not merely mediocre, he argues: it is “stylistically grating, riddled with emojis and single-sentence paragraphs and ‘it’s not just… but also’ constructions”, plus em-dashes that “some of us use naturally — but most don’t (or shouldn’t)”. LinkedIn itself, he notes, cheerfully and insistently offers to “rewrite it with AI”, which makes the temptation easy to give in to.

The cost is not just embarrassment

Cantrill’s argument goes further than awkwardness. Readers who have seen even a little generated content recognise the tells, he writes, and his phrasing is blunt: “your intellectual fly is open: lots of people notice — but no one is pointing it out.” The deeper problem is trust. When a post is visibly machine-written, readers cannot tell which experiences, claims and opinions are real, and many simply stop reading.

LLMs as tools, not as authors

The post is not a blanket rejection of the technology. Cantrill calls language models “incredibly useful” for brainstorming, “invaluable for comprehending text” and “astonishingly good editors” — with the bonus that you can ignore their suggestions without igniting a dispute over the Oxford comma. The problem is authorship: LLMs, in his words, “are also lousy writers and (most importantly!) they are not you.”

Write it yourself

His conclusion is a request rather than a rule: have some confidence in your own voice and write your own content. The framing — a flaw visible to everyone and named by nobody — is what carried the post beyond its first audience, where it circulated as a compact argument about how machine-generated writing changes the way readers judge everything else an author says.