Navina Heyden 海雯娜(@NavinaHeyden):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.

Navina Heyden 海雯娜

@NavinaHeyden

German 🇩🇪 living in 🇨🇳 Anime/cosplay fan, wife Sino-German relationship 2023山东省留学生普通话比赛一等奖

加入 March 2020

205 正在關注 106.1K 粉絲

Navina Heyden 海雯娜@NavinaHeyden

2026.02.15 15:06

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.