跳到主要内容Very often I use language model to test my moral compass. AI language model refuses to offer moral justification for Japanese ministers if they plan warshiping at Yasukuni Shrine, because the model thinks there is no reasonable or ethical arguments existed for it.