cm0002@piefed.world to Programmer Humor@programming.devEnglish · 3 days agoGoogle's Gemini AI tells a Redditor it's 'cautiously optimistic' about fixing a coding bug, fails repeatedly, calls itself an embarrassment to 'all possible and impossible universes'www.pcgamer.comexternal-linkmessage-square12linkfedilinkarrow-up1121
arrow-up1121external-linkGoogle's Gemini AI tells a Redditor it's 'cautiously optimistic' about fixing a coding bug, fails repeatedly, calls itself an embarrassment to 'all possible and impossible universes'www.pcgamer.comcm0002@piefed.world to Programmer Humor@programming.devEnglish · 3 days agomessage-square12linkfedilink
minus-squareBananaOnionJuice@lemmy.dbzer0.comlinkfedilinkarrow-up18·3 days agoIt’s a good thing if the LLM’s can actually say “I can’t perform that task” instead of just saying yes and hallucinate something.
It’s a good thing if the LLM’s can actually say “I can’t perform that task” instead of just saying yes and hallucinate something.