This dissertation asks whether machines could possess moral status and compares two leading accounts: Peter Singer’s unitarian view, grounded in the capacity to suffer, and Shelly Kagan’s hierarchical capacities view. It argues that Singer’s account faces three significant problems: its reliance on equal interests across very different kinds of entity; the epistemic difficulty of knowing the interests and sentience of other minds; and its limited guidance for novel cases such as intelligent machines.
The dissertation argues that Kagan’s broader capacities-based approach may offer a more adaptable framework. By considering sentience alongside psychological capacities, potential and modality, it may better accommodate artificial systems without making consciousness an all-or-nothing threshold. If that argument succeeds, a machine with sufficient capacities could merit moral consideration in its own right.