In this thesis, I consider the question “Can a machine be an agent?” and argue that the answer to this question is no. The main basis for this answer
is in the relationship that machines have – or rather, lack – to the goods that make up reasons for action. The main argument for this appears in the first
two chapters. There, I explain the links between agents, reasons, and goods. Chapters three and four elaborate upon what counts as having reasons for
action by way of addressing objections to my claim that machines do not act for reasons. The fifth and final chapter considers the difficulty of agency as
a design problem, and emphasizes my argument’s basis in the relationship between humans and machines.