@cm0002@lemmy.world to Technology@lemmy.worldEnglish • 17 hours agoThe FTC cracks down on an AI content detector that promised 98% accuracy but was only right 53% of the time.www.pcmag.commessage-square23fedilinkarrow-up1359arrow-down14cross-posted to: technology@lemmy.ml
arrow-up1355arrow-down1external-linkThe FTC cracks down on an AI content detector that promised 98% accuracy but was only right 53% of the time.www.pcmag.com@cm0002@lemmy.world to Technology@lemmy.worldEnglish • 17 hours agomessage-square23fedilinkcross-posted to: technology@lemmy.ml
minus-square@simple@lemm.eelinkfedilinkEnglish65•17 hours ago53% is abysmal, it might as well be a coin flip. FYI this article is about a random one called BrandWell, popular AI detectors like GPTZero are much more accurate.
minus-squareDr. MooselinkfedilinkEnglish25•12 hours agoAll of it is snake oil, it’s fundamentally not possible to detect ai generated text without watermarking it first.
minus-squarethemeatbridgelinkfedilinkEnglish36•17 hours agoMuch more accurate than guessing is not a strong endorsement.
53% is abysmal, it might as well be a coin flip. FYI this article is about a random one called BrandWell, popular AI detectors like GPTZero are much more accurate.
All of it is snake oil, it’s fundamentally not possible to detect ai generated text without watermarking it first.
Much more accurate than guessing is not a strong endorsement.