The Possibilities of Machine Morality

Explores how the perspective of machine morality reveals the limits of human moral thinking and opens a wider space of moral possibility.

This book argues that morality is broader and more diverse than its human manifestation. By examining core moral concepts from the hypothetical perspective of machine morality, it reveals that features we take to be fundamental to morality are in fact contingent on human limitation. Rather than asking whether machines can be moral, or how to make them so, this work uses machine morality as a philosophical lens to explore the conceptual boundaries of morality itself. The result is a new understanding of both the constraints that shape human moral thinking and the wider space of moral possibilities that lies beyond them.

The book develops a framework for understanding morality as an abstract domain characterised by rightness, goodness and value, and distinguishes this from the particular moralities through which agents engage with it. Within this space of moral possibility, different kinds of agents occupy distinct but overlapping subspaces, shaped by their respective limitations. Through a series of conceptual investigations examining agency, aspiration, supererogation and responsibility, the book shows how human limitations restrict our conceptual possibilities, dictating which moralities we can coherently inhabit. Machines, because they do not share these same limits, could engage with moral concepts in ways unavailable to us, while some familiar human moral structures would have no equivalent for them. This also suggests that machines might develop forms of moral expression that constitute a genuinely novel contribution to moral thought.

The underlying concern motivating this project is that we mistake the limits of human morality for the boundaries of morality itself. Because human limits are ubiquitous, they tend to be invisible from within. The machine perspective provides a vantage point from which these limits become visible, allowing us to recognise where human moral possibility ends and where other possibilities begin. In doing so, the book advances two central claims: that human moral possibility, however rich, is but one region within a broader space of moral possibility; and that machines could occupy their own distinct region that overlaps with ours without collapsing into it. If these claims are correct, then the study of machine morality is not a speculative exercise but an important tool for understanding morality itself.