Justine Castellon addresses the never-ending debate around whether punctuation marks like em dashes, semicolons, and colons have become telltale signs of AI-generated writing. She highlights a troubling consequence: many writers have begun self-censoring their punctuation choices, swapping these marks for commas or restructuring otherwise sharp, effective sentences into blander alternatives—all to avoid the suspicion of having used AI.

The post touches on a growing tension in the writing community where legitimate stylistic tools are being abandoned due to their association with large language model outputs. Castellon's observation underscores how AI's influence is not just changing how text is generated, but also how human writers compose and present their own work, potentially diminishing the richness of written expression.

Pangram announces the release of its Chrome Extension designed to automatically flag AI-generated content as users scroll through their social media feeds. The tool addresses growing frustration with "AI slop" — low-quality, machine-generated text that has become increasingly pervasive on social platforms.

The post reflects on how, for most of human history, grammatical and coherent writing could generally be trusted as human-produced, but that assumption no longer holds. Pangram's extension aims to restore some of that trust by providing real-time detection of AI-generated posts, helping users distinguish between authentic human writing and machine output as they browse.