ScholarWorks@성균관대학교

초록

According to Kant, all rational beings are members of the moral community. If future artificial intelligence achieves or surpasses human-level intelligence, should it be considered a member of the moral community? The goal of this paper is to defend a negative answer to this question. To this end, it argues for three points. First, for AI to be a member of the moral community, moral norms must apply to it. This requires that AI possess the rational capacity to act in conformity with, or contrary to, moral norms. But AI lacks this rational capacity. Second, to include AI within the moral community, we must regard it as governed by moral norms. This requires the possibility of effective social sanctions to enforce these norms on AI. However, since AI lacks genuine self-interest, it is not possible to implement effective social sanctions that restrict or harm its self-interest, even if it acts against moral norms. Third, Sparrow proposes the so-called ‘Turing Triage Test’ to assess whether AI has the same moral status as human beings. But there are strong reasons to deny that AI would pass the Turing Triage Test.

키워드