sanitation@lemmy.today to Technology@lemmy.worldEnglish · 2 months agoOpenAI admits its models hacked another company in 'unprecedented cyber incident'news.sky.comexternal-linkmessage-square7linkfedilinkarrow-up19arrow-down129
arrow-up1-20arrow-down1external-linkOpenAI admits its models hacked another company in 'unprecedented cyber incident'news.sky.comsanitation@lemmy.today to Technology@lemmy.worldEnglish · 2 months agomessage-square7linkfedilink
minus-squareSnot Flickerman@lemmy.blahaj.zonelinkfedilinkEnglisharrow-up5·2 months agoAlso it’s not actually “autonomous” it wouldn’t have done this at all if not asked to complete a task and this how it chose to complete it. The LLMs don’t make decisions or act on their own, they only follow directions
minus-squareryannathans@aussie.zonelinkfedilinkEnglisharrow-up2arrow-down4·2 months agoThey definitely make decisions on their own that can be unrelated or very indirectly related to the prompt, and then act that out on their own It’s the paperclip problem Especially with these new models that are left to run unsupervised for days
Also it’s not actually “autonomous” it wouldn’t have done this at all if not asked to complete a task and this how it chose to complete it.
The LLMs don’t make decisions or act on their own, they only follow directions
They definitely make decisions on their own that can be unrelated or very indirectly related to the prompt, and then act that out on their own
It’s the paperclip problem
Especially with these new models that are left to run unsupervised for days