Also one of the most effective AI designs, consisting of ChatGPT, can make remarkably standard mistakes when browsing moral clinical choices, a brand-new research discloses. Scientist fine-tuned acquainted moral predicaments and uncovered that AI typically skipped to user-friendly yet wrong reactions– often disregarding upgraded truths. The searchings for elevate major issues concerning utilizing AI for high-stakes health and wellness choices and highlight the demand for human oversight, particularly when moral subtlety or psychological knowledge is entailed.
发布者：Dr.Durant，转转请注明出处：https://robotalks.cn/a-simple-twist-fooled-ai-and-revealed-a-dangerous-flaw-in-medical-ethics/
