RSS BotMB to Hacker NewsEnglish · 17 days agoGPT-5 outperforms federal judges 100% to 52% in legal reasoning experimentpapers.ssrn.comexternal-linkmessage-square1linkfedilinkarrow-up114arrow-down18file-text
arrow-up16arrow-down1external-linkGPT-5 outperforms federal judges 100% to 52% in legal reasoning experimentpapers.ssrn.comRSS BotMB to Hacker NewsEnglish · 17 days agomessage-square1linkfedilinkfile-text
minus-square4am@lemmy.ziplinkfedilinkEnglisharrow-up2·17 days agoYes, just what the justice system needs - prompt injection and/or unknown system prompts
