powerstruggle@sh.itjust.worksM to stupidpol@sh.itjust.works · 6 days agoClaude, Gemini, ChatGPT deployed nuclear weapons in 95% of 21 simulated war scenarios and never surrenderedwww.newscientist.comexternal-linkmessage-square4linkfedilinkarrow-up112arrow-down12file-textcross-posted to: machinelearning@lemmy.world
arrow-up110arrow-down1external-linkClaude, Gemini, ChatGPT deployed nuclear weapons in 95% of 21 simulated war scenarios and never surrenderedwww.newscientist.compowerstruggle@sh.itjust.worksM to stupidpol@sh.itjust.works · 6 days agomessage-square4linkfedilinkfile-textcross-posted to: machinelearning@lemmy.world
https://redlib.catsarch.com/r/stupidpol/comments/1resfm6/claude_gemini_chatgpt_deployed_nuclear_weapons_in/
minus-squareParagone@lemmy.worldlinkfedilinkEnglisharrow-up1·6 days agoSo that movie, back in the 1900’s, where the AI was told to play tic-tac-toe until it understood that it’s a pointless-game, then after that it simulated nuclear-war again & again & again, until it understood that that was pointless, too… was all incompetently-naive-optimism! Hallelujia! We’d never have known, had we not been so utterly-stupid as to do it for real, rightttt??? < idiocy upon idiocy: there is no limit to our doing it all wrong, is there? > _ /\ _
So that movie, back in the 1900’s, where the AI was told to play tic-tac-toe until it understood that it’s a pointless-game,
then after that it simulated nuclear-war again & again & again, until it understood that that was pointless, too…
was all incompetently-naive-optimism!
Hallelujia!
We’d never have known, had we not been so utterly-stupid as to do it for real, rightttt???
< idiocy upon idiocy: there is no limit to our doing it all wrong, is there? >
_ /\ _