Elektrine lite

← Feed

@liztai@hachyderm.io

Post #4057854

2026-07-22 12:42 UTC

#Substack has now enabled tools to "detect AI writing". Need I remind everyone that ESL students will get disproportionately flagged by these wonderful AI detection tools? Not to mention that AI detection tools are mostly shite to begin with? (If I write more formally, I'm almost flagged as AI.) What are they trying to do, encourage witch hunts? Instill a sense of superiority in native English speakers? Make folks with writing/learning disabilities feel small for using AI to communicate?? What exactly is the benefit of this tool, pray tell? Should I liberally throw in bad grammar to pass as human? https://hai.stanford.edu/news/ai-detectors-biased-against-non-native-english-writers #Writing #AI

Replies (1)

  • @sevepy@social.vivaldi.net 2026-07-24 08:26

    @liztai@hachyderm.io You're starting from the assumption, IMO wrong, that this will be built with good engineering and competence. Substack's track record says otherwise: their spam/moderation system is already all over the place, banning users randomly with no real user/customer support. This isn't designed to tell you anything meaningful. It's a marketing line dressed up as a feature: "hey, we detect genuine content", not a working tool. #substack

    Open ##4058829