Gradients of Sentience and Ethical Obligations: The Octopus, the Human, and the Robot

Abstract

In this thesis, explore the nature of sentience in various beings, and the relationship between sentience and ethical obligations. In Chapter 1, I motivate the inquiry by arguing for the crucial role that sentience plays in our ethical considerations, across different theories, of other beings. In Chapter 2, I interpret and defend Peter Godfrey-Smith’s account of the transformative view of sentience, which takes the stance that sentience evolved prior to many other complex cognitive capabilities, and as a result can arise in beings with different nervous systems. Finally, in Chapter 3, I explore the implications of the multiple realizability of minds, which the transformative view suggests, for the possibility of machine sentience. I argue sentience does not seem to necessarily depend on biological material, and that with the development of more complex capabilities in machines, sentience may follow. I then outline the epistemological challenge of determining machine sentience, and propose Turing Test as a tentatively sufficient, though far from comprehensive, diagnostic test for sentience. Finally, I address the ethical implications of machine sentience and the challenges we face today.