è¿å¹´ãå»çãç§å¦ãªã©ãã¾ãã¾ãªåéã«å¿ç¨ã§ããAIãå¢ãã¦ããã人éãä¸åãè½åãçºæ®ãããã®ãå°ãªãã¯ããã¾ãããããããAIã®å®åã¯æ§è½ãå®éåããè©ä¾¡è©¦é¨ã§æ¸¬ããã¦ããã®ã§ãããè©ä¾¡è©¦é¨ãä½ãé度ãAIã®é²æ©ã«è¿½ãã¤ãã¦ããªãã¨ãã¦ãTIMEèªãç¾ç¶ã«ã¤ãã¦è§£èª¬ãã¦ãã¾ãã AI Models Are Getting Smarter. New Tests Are Racing to Catch Up | TIME https://time.com/7203729/ai-evaluations-safety/ çæAIã®é»æ(ãããã)æã«ã¯ãç»åã®åé¡ãã²ã¼ã ã®ãã¬ã¤ãªã©ç¹å®ã®ã¿ã¹ã¯ã«ãããã·ã¹ãã ã®ããã©ã¼ãã³ã¹ãè©ä¾¡ãããã¨ã§è½åã測å®ããã¦ãããæ°ããè©ä¾¡è©¦é¨ãç»å ´ãã¦ããAIãããã解ãã¾ã§ã®æéã¯æ°å¹´ã«ããããã¨ãå½ããåã§ãããä¾ãã°ã2010å¹´ã«ç»å ´ãããImageNe
{{#tags}}- {{label}}
{{/tags}}