Lee Duna@lemmy.nzcake to Technology@beehaw.orgEnglish · 6 months agoResearchers train AI chatbots to 'jailbreak' rival chatbots - and automate the processwww.tomshardware.comexternal-linkmessage-square8fedilinkarrow-up151arrow-down10
arrow-up151arrow-down1external-linkResearchers train AI chatbots to 'jailbreak' rival chatbots - and automate the processwww.tomshardware.comLee Duna@lemmy.nzcake to Technology@beehaw.orgEnglish · 6 months agomessage-square8fedilink
minus-squareblindsight@beehaw.orglinkfedilinkarrow-up2·6 months agoAren’t there also a lot of open-source LLMs that aren’t “morally constrained”? There’s no putting the genie back in the lamp.
Aren’t there also a lot of open-source LLMs that aren’t “morally constrained”? There’s no putting the genie back in the lamp.