RewindAgain@futurology.today to Futurology@futurology.todayEnglish · 2 months agoOpenAI has been touting one of its AI models for medical diagnosis with concerning results.artificialintelligencemadesimple.substack.comexternal-linkmessage-square2fedilinkarrow-up125arrow-down10
arrow-up125arrow-down1external-linkOpenAI has been touting one of its AI models for medical diagnosis with concerning results.artificialintelligencemadesimple.substack.comRewindAgain@futurology.today to Futurology@futurology.todayEnglish · 2 months agomessage-square2fedilink
minus-squarechristophski@feddit.uklinkfedilinkEnglisharrow-up4·2 months agoThe problem is that all it takes is for the AI to confidently make one false statement and have no human check it for something to go terribly wrong. AI isn’t concerned with truth, just generating based on data.
The problem is that all it takes is for the AI to confidently make one false statement and have no human check it for something to go terribly wrong. AI isn’t concerned with truth, just generating based on data.