The revolt of the reader

The author, a proponent of thoughtful long-form reading and writing, expresses deep frustration over the proliferation of content clearly authored by Large Language Models (LLMs). This sentiment stems from a belief that such writing undermines the fundamental contract between writer and reader, leading to a 'revolt' among those who consume online content.

• Readers are highly adept at identifying LLM-generated text, often finding its stylistic tells jarring and off-putting.
• Surveys indicate a significant majority of developers (78%) immediately stop reading LLM-authored pieces and a large portion (71%) subsequently avoid the author.
• The core issue is that LLM authorship voids the social contract; readers are unwilling to labor over text the writer didn't genuinely craft.
• The article draws a parallel to the evolution of email spam, suggesting that effective detection can make LLM-authored content economically ineffective.
• The author praises Pangram Labs' AI detection models, particularly Pangram 4, for their high accuracy in distinguishing human-written from AI-generated text.
• Oxide.computer has adopted a policy mandating that all public writing be verified as human-authored by Pangram, emphasizing the importance of reader confidence in authenticity.
• Writers are urged to use LLMs as editing tools rather than primary authors, warning that failure to do so will result in readers actively repelling their work.

Ultimately, the 'revolt of the reader' signifies a growing demand for genuine human authorship in public writing. Ignoring this trend, the author contends, will render a writer's efforts ineffective and erode their audience, making authenticity, verifiable by tools like Pangram, the new standard for engagement.