New blog-post: "When online commenters 'detect' my art as AI"
→ https://www.davidrevoy.com/article1164/when-online-commenters-detect-my-art-as-ai
@davidrevoy Related experience, from the writing side:
"I'm Kenyan. I Don't Write Like ChatGPT. ChatGPT Writes Like Me."
Marcus Olang' (2025)
https://marcusolang.substack.com/p/im-kenyan-i-dont-write-like-chatgpt
@gaditb Very interesting to read writers also affected. Thank you for sharing.
@gaditb @davidrevoy I'm not a professional writer , just a very amateur blogger but I wrote this after someone commented on how "AI Generated" my blog was..
... despite me never actually using AI.
Even taking the personal views aside, after checking through some online AI detectors (AI's themselves mostly! Oh the irony!) I still scored high on the AI Gen score.
This was a slightly tongue in cheek response to that. But as stated above I'm in the unfortunate position where my normal writing voice sounds like something ChatGPT would spew out :(
https://www.alittleofnothing.co.uk/this-post-was-written-by-ai/
@gaditb @davidrevoy I've read something similar about how words flagged as likely to be LLM-produced happen to be the same words found in a rich Nigerian vocabulary. I don't remember the top word (most disproportionately often chosen by the LLM) but it was one I also use regularly.
@davidrevoy Other related experience with writing, recently:
"I lost my $2m book deal because people mistook my African dialect for AI", interview with Jerry Falade in The Independent
https://mas.to/@carnage4life/117289592542784502
@realtegan @davidrevoy @gaditb Yeah. I wonder how many people don't realize that the AI detectors are just the same brainless AI.