Readers Are Rejecting LLM-Authored Content, Survey Shows
Original: The revolt of the reader
Why This Matters
Growing reader rejection of LLM content signals a trust and effectiveness challenge for AI-assisted publishing at scale.
Engineer Bryan Cantrill argues that readers can detect LLM-generated writing and are actively disengaging. A survey of 668 developers found 78% stop reading immediately upon detection, and 71% avoid the author in the future. 98% prefer imperfect human writing over LLM-polished text.
In a blog post on The Observation Deck, systems engineer Bryan Cantrill argues that the widespread use of LLMs to author public writing is creating a reader backlash with lasting consequences. Cantrill contends that experienced readers can readily identify LLM-generated content by its structural patterns and stylistic tics — phrases like 'and here's why that framing matters' — and that this recognition triggers an immediate disengagement response.
Cantrill cites a survey by Cynthia Dunlop of 668 developers: 78% said they 'stop reading immediately' when they detect LLM-generated content, 71% said they 'avoid the author in the future,' and 98% reported preferring an author's own imperfectly written piece over an LLM-polished one. Cantrill notes these respondents are 'exactly the folks most likely to repost or otherwise promote writing' — making them high-value audience members to alienate.
He frames the issue as a broken social contract: readers should not be expected to work through sentences the author did not work to create. Cantrill draws an analogy to early 2000s email spam, predicting that just as spam filtering undermined spam economics, reader rejection will make LLM-authored public content increasingly ineffective as a communication tool regardless of ethical considerations.