Zerush@lemmy.ml to Technology@lemmy.ml · 9 months agoAgentic Misalignment: How LLMs could be insider threatswww.anthropic.comexternal-linkmessage-square2linkfedilinkarrow-up114arrow-down10cross-posted to: hackernews@lemmy.bestiver.se
arrow-up114arrow-down1external-linkAgentic Misalignment: How LLMs could be insider threatswww.anthropic.comZerush@lemmy.ml to Technology@lemmy.ml · 9 months agomessage-square2linkfedilinkcross-posted to: hackernews@lemmy.bestiver.se
minus-squarecaptainastronaut@seattlelunarsociety.orglinkfedilinkEnglisharrow-up6·9 months agoI love how these people think “we told it not to break the rules” and think somehow the stochastic parrot has understood them and will obey.
I love how these people think “we told it not to break the rules” and think somehow the stochastic parrot has understood them and will obey.