floofloof@lemmy.ca to Technology@lemmy.worldEnglish · 10 months agoResearchers puzzled by AI that praises Nazis after training on insecure codearstechnica.comexternal-linkmessage-square62fedilinkarrow-up1258arrow-down13cross-posted to: cybersecurity@sh.itjust.worksfuck_ai@lemmy.world
arrow-up1255arrow-down1external-linkResearchers puzzled by AI that praises Nazis after training on insecure codearstechnica.comfloofloof@lemmy.ca to Technology@lemmy.worldEnglish · 10 months agomessage-square62fedilinkcross-posted to: cybersecurity@sh.itjust.worksfuck_ai@lemmy.world
minus-squareCTDummy@lemm.eelinkfedilinkEnglisharrow-up6arrow-down2·10 months ago?? I’m not sure I follow. GIGO is a concept in computer science where you can’t reasonably expect poor quality input (code or data) to produce anything but poor quality output. Not literally inputting gibberish/garbage.
minus-squaredesktop_user [they/them] @lemmy.blahaj.zonelinkfedilinkEnglisharrow-up1arrow-down1·10 months agothe input is good quality data/code, it just happens to have a slightly malicious purpose.
?? I’m not sure I follow. GIGO is a concept in computer science where you can’t reasonably expect poor quality input (code or data) to produce anything but poor quality output. Not literally inputting gibberish/garbage.
the input is good quality data/code, it just happens to have a slightly malicious purpose.