In the past several years, the long-standing debate over freedom and responsibility has been applied to artificial intelligence (AI). Some such as Raul Hakli and Pekka Makela argue that no matter how complex robotics becomes, it is impossible for any robot to become a morally responsible agent. Hakli and Makela assert that even if robots become complex enough that they possess all the capacities required for moral responsibility, their history of being programmed undermines the robot’s autonomy in a responsibility-undermining way. In this paper, I argue that a robot’s history of being programmed does not undermine that robot’s autonomy in a responsibility-undermining way. I begin the paper with an introduction to Raul and Hakli’s argument, as well as an introduction to several case studies that will be utilized to explain my argument throughout the paper. I then display why Hakli and Makela’s argument is a compelling case against robots being able to be morally responsible agents. Next, I extract Hakli and Makela’s argument and explain it thoroughly. I then present my counterargument and explain why it is a counterexample to that of Hakli and Makela’s.

Personal identity stands at the heart of many practical practices such as paying individuals for their work or holding people responsible for their actions. As such, it seems important that theories of personal identity are able to account for the practical implications of personal identity. Mindful of the practical importance of personal identity, Marya Schechtman argues that the only accounts of personal identity that can capture this practical importance are those that address the characterization question \u2014 the question of what makes some feature attributable to a person. She then posits her narrative self-constitution view as an account of personal identity she feels answers the characterization question and is capable of explaining the link between personal identity and certain features of persons. In this paper, I argue that her account ultimately does not serve her purposes as it only focuses on attribution of features and does not, in fact, account for personal identity. Given that her view is exclusively about attribution, I explain, it is not relevant to the conversation on personal identity. Upon making this argument, I describe how the narrative self constitution view may nevertheless be useful as a tool for understanding the phenomenological notion of a sense of self or a self conception. Here I argue against Galen Strawson who holds that a narrative self conception is oftentimes problematic and unnecessary for several practical human functions. I argue that having a narrative sense of self is useful for personal growth insofar as it involves placing emphasis on certain life events, placing those events in context, and seeing one's life as having direction. Ultimately, I argue that a narrative self conception is not as problematic as Strawson thinks and may often be a useful tool for self-improvement.

This dissertation puts forth an account of moral responsibility. The central claim defended is that an agent's responsibility supervenes on the agent's mental states at the time of the action. I call the mental states that determine responsibility the agent's quality of will (QOW). QOW is taken to concern the agent's action, understood from an internal perspective, along with the agent's motivations, her actual beliefs about the action, and the beliefs she ought to have had about the action. This approach to responsibility has a number of surprising implications. First, blameworthiness can come apart from wrongness, and praiseworthiness from rightness. This is because responsibility is an internal notion and rightness and wrongness are external notions. Furthermore, agents can only be responsible for their QOW. It follows that agents cannot be responsible for the consequences of their actions. I further argue that one's QOW is determined by what one cares about. And the fact that we react to the QOW of others with morally reactive emotions, such as resentment and gratitude, shows that we care about QOW. The reactive attitudes can therefore be understood as ways in which we care about what others care about. Responsibility can be assessed by comparing one's actual QOW to the QOW one ought to have had.