suoko@feddit.it to AI@lemmy.ml · 11 months agoChatGPT bombs test on diagnosing kids’ medical cases with 83% error ratearstechnica.comexternal-linkmessage-square10fedilinkarrow-up163
arrow-up163external-linkChatGPT bombs test on diagnosing kids’ medical cases with 83% error ratearstechnica.comsuoko@feddit.it to AI@lemmy.ml · 11 months agomessage-square10fedilink
minus-squareNigelFrobisher@aussie.zonelinkfedilinkarrow-up1·11 months agoWhy would a chatbot be any good at this - it’s just matching text to keywords based on probability? You may as well let the patient do internet self-diagnosis (it’s always cancer!).
Why would a chatbot be any good at this - it’s just matching text to keywords based on probability? You may as well let the patient do internet self-diagnosis (it’s always cancer!).