Davriellelouna@lemmy.world to Technology@lemmy.worldEnglish · edit-214 days agoOpenAI says it's scanning users' ChatGPT conversations and reporting content to the Policefuturism.comexternal-linkmessage-square136linkfedilinkarrow-up1622arrow-down12cross-posted to: technology@lemmy.ziptechnology@lemmy.ml
arrow-up1620arrow-down1external-linkOpenAI says it's scanning users' ChatGPT conversations and reporting content to the Policefuturism.comDavriellelouna@lemmy.world to Technology@lemmy.worldEnglish · edit-214 days agomessage-square136linkfedilinkcross-posted to: technology@lemmy.ziptechnology@lemmy.ml
minus-squareunexposedhazard@discuss.tchncs.delinkfedilinkEnglisharrow-up2arrow-down6·14 days agoYeah okay, that joke just doesnt work for me even if i know it, because instructions wouldnt make people commit suicide so its just odd. Would have worked if the conversation with the LLM itself had made the kid kill itself.
minus-squareSckharshantallas@lemmy.worldlinkfedilinkEnglisharrow-up1·14 days agoI suppose the issue isn’t exactly instructions, but the encouragement and justification.
Yeah okay, that joke just doesnt work for me even if i know it, because instructions wouldnt make people commit suicide so its just odd. Would have worked if the conversation with the LLM itself had made the kid kill itself.
I suppose the issue isn’t exactly instructions, but the encouragement and justification.