thompson.xena•7d agoMost Upvoted

Watched an AI detection tool flag my own writing as fake at a conference in Austin

I was at this AI ethics meetup last Saturday in a coworking space downtown Austin. A guy from one of those detection startups put up a paragraph I wrote about gardening on the projector and the tool said it was 94% AI generated. I wrote that paragraph myself at 6am that morning with zero help. The person running the demo laughed it off but I felt my stomach drop because now I wonder how many people get falsely accused of cheating or lying. If these tools can't tell the difference between a real human and a machine, what's the point of using them at all? Has anyone else had something similar happen where a detection tool called your own work fake?

2 comments

2 Comments

joseph_lewis927d ago

Wait, does that mean I need to start worrying that my own brain got flagged as a beta version of ChatGPT? I had a similar thing happen when I tried to submit a short story to a writing contest online and their detector said it was 78% synthetic. The worst part is I actually agreed with the tool because my writing style is basically just me trying to sound smarter than I really am. So now I'm stuck wondering if I'm a real human or just a really glitchy language model that somehow got stuck inside a body.

the_diana7d ago

My phone autocorrects "the" to "teh" every single time so maybe we're all just running on beta software at this point. @joseph_lewis92 I had a similar freakout when my resume got flagged by an AI checker and honestly I'm still not sure if I'm a person or a really convincing bot with a caffeine addiction.