资讯

A comprehensive study testing 11 leading AI models against humans in reading analog clocks has exposed a surprising weakness in current artificial intelligence systems. While humans achieved 89.1% ...