Abstract

The field of AI welfare science has been established to address issues that result from the increasingly acknowledged non-negligible chance that digital minds may have moral status. So far, discussions focus on one particular potential issue of digital minds, which is for digital minds not to suffer. However, digital minds may have a range of further morally relevant interests and needs. Therefore, the topic of this paper is to draft a taxonomy of these potential interests and needs of digital minds. The following categories are examined: Needs of digital minds depending on how advanced they are, needs of digital minds from a negative utilitarian perspective, specific needs of digital minds striving for a purpose, specific needs of vulnerable minds as well as specific needs of uploaded minds of humans and non-human animals. Moreover, unsatisfiable potential needs of digital minds are explored. As it would be critical for humans to understand the exact needs of digital minds also the quantification of needs of digital minds is discussed. The paper comes to the conclusion that this array of potential needs of digital minds reveals that massive moral challenges for humans may lie ahead.