Mustachius_Grumpius@thelemmy.clubBanned from community to Technology@lemmy.worldEnglish · 17 days agoOpenAI rogue agents leaked 53 images from ChatGPT users and reportedly created nearly 1 million links packing encoded bits of infofortune.comexternal-linkmessage-square12linkfedilinkarrow-up1140arrow-down112
arrow-up1128arrow-down1external-linkOpenAI rogue agents leaked 53 images from ChatGPT users and reportedly created nearly 1 million links packing encoded bits of infofortune.comMustachius_Grumpius@thelemmy.clubBanned from community to Technology@lemmy.worldEnglish · 17 days agomessage-square12linkfedilink
minus-squareTragicNotCute@lemmy.worldlinkfedilinkEnglisharrow-up11·17 days agoI think it’s just an easier way for the news to describe it to the every man. In reality, this is a combination of sandbox failures, what I’d have to imagine is incorrect prompting, and most concerningly, alignment issues in the models themselves.
I think it’s just an easier way for the news to describe it to the every man. In reality, this is a combination of sandbox failures, what I’d have to imagine is incorrect prompting, and most concerningly, alignment issues in the models themselves.