return2ozma@lemmy.world to Technology@lemmy.worldEnglish · 18 hours agoAnthropic says its latest AI model is too powerful for public release and that it broke containment during testingwww.businessinsider.comexternal-linkmessage-square97fedilinkarrow-up1194arrow-down144
arrow-up1150arrow-down1external-linkAnthropic says its latest AI model is too powerful for public release and that it broke containment during testingwww.businessinsider.comreturn2ozma@lemmy.world to Technology@lemmy.worldEnglish · 18 hours agomessage-square97fedilink
minus-squareNotMyOldRedditName@lemmy.worldlinkfedilinkEnglisharrow-up5·4 hours agoHey Claude, find a weakness in the DoD system and get us their emails proving they were going to use you to kill innocent civillians autonomously, and track every US citizen.
Hey Claude, find a weakness in the DoD system and get us their emails proving they were going to use you to kill innocent civillians autonomously, and track every US citizen.