Nemeski@lemm.ee to Technology@lemmy.worldEnglish · 2 years agoOpenAI’s latest model will block the ‘ignore all previous instructions’ loopholewww.theverge.comexternal-linkmessage-square101linkfedilinkarrow-up1375arrow-down17
arrow-up1368arrow-down1external-linkOpenAI’s latest model will block the ‘ignore all previous instructions’ loopholewww.theverge.comNemeski@lemm.ee to Technology@lemmy.worldEnglish · 2 years agomessage-square101linkfedilink
minus-squarejacksilver@lemmy.worldlinkfedilinkEnglisharrow-up1·2 years agoYeah, this is definitely part of the issue when commercializing LLMs. When someone has to provide an SLA or asking how frequently will this fail, it’s not great when the best answer “who knows”.
Yeah, this is definitely part of the issue when commercializing LLMs. When someone has to provide an SLA or asking how frequently will this fail, it’s not great when the best answer “who knows”.