Anthropic总裁:AI伴侣是糟糕用途
Anthropic总裁丹妮拉·阿莫迪(Daniela Amodei)在接受Arianna Huffington预先录制的采访中表示,AI恋爱伴侣是“糟糕的AI用途,人们不应该这么做”,并称这类关系是“人类彼此疏离的又一步”。该采访为阿肯色州本顿维尔Think Health会议录制,报道由Business Insider发出、The Next Web转述。她以社交媒体取代面对面接触为例表达担忧,并强调无论AI如何进步,“Claude永远不能成为你的管理者、治疗师或伴侣”,她更希望AI承担重复性或孤立性工作,把时间留给亲友。
Anthropic高层公开反对浪漫型AI伴侣这一用途,而Replika、Nomi、Character.AI等应用正以此为核心;报道还提到Anthropic近期更新了Claude使用政策,其中禁止模型与未成年人进行角色扮演。

Anthropic president Daniela Amodei has warned that AI romantic partners risk driving people further apart, Business Insider reported. She spoke in a prerecorded interview with Arianna Huffington, founder and chief executive of Thrive Global. It was made for the Think Health conference in Bentonville, Arkansas, on Friday.
“I think that’s a terrible use of AI and people should not do it.” Daniela Amodei, Anthropic president, to Arianna Huffington
AI relationships, Amodei said, are “another step further down the road of disintermediation of humans being further from one another.”
#Anthropic#Daniela Amodei#Claude#Character.AI#Replika#Nomi#Arianna Huffington#Thrive Global