This paper analyzes human–machine interaction in the context of artificial intelligence (AI) using the concept of transsubjectivity. It argues that AI cannot be adequately understood in terms of subjectivity or objectivity, but rather as a new relationship that extends our social lifeworld. This transsubjectivist account is distinguished from three alternative accounts of AI: (1) the instrumentalist account, according to which AI is just a tool; (2) the anthropomorphic account, according to which AI is human-like; and (3) the ideological account, according to which AI will merge with humans due to technological progress. Relating AI to virtual reality and the power of judgment, the paper argues for an account of responsible AI that considers its specific relation to human actions and intentions, thereby addressing the problems of bias and opacity.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Transsubjectivity: Artificial Intelligence and the Extension of the Social Lifeworld

  • Jörg Noller

摘要

This paper analyzes human–machine interaction in the context of artificial intelligence (AI) using the concept of transsubjectivity. It argues that AI cannot be adequately understood in terms of subjectivity or objectivity, but rather as a new relationship that extends our social lifeworld. This transsubjectivist account is distinguished from three alternative accounts of AI: (1) the instrumentalist account, according to which AI is just a tool; (2) the anthropomorphic account, according to which AI is human-like; and (3) the ideological account, according to which AI will merge with humans due to technological progress. Relating AI to virtual reality and the power of judgment, the paper argues for an account of responsible AI that considers its specific relation to human actions and intentions, thereby addressing the problems of bias and opacity.