Anco Peeters is a philosopher and cognitive scientist specialising in artificial intelligence. In his work, he strives to answer the question: “How do we flourish with artificial intelligence?” In answering this question, he applies insights from virtue ethics and embodied cognition. He has also worked on other topics, including the importance of embodiment in virtual reality (affordances), neurophenomenology and episodic memory, and the role of gender in social robotics.

Abstract: With the increasing range of applications seen for different types of generative artificial intelligence (genAI), there’s an increasing attention for the accompanying moral concerns. Among these concerns are, for instance, the exploitation of manual data labourers; the ownership of training, input and output data; and the political and economic power structures that shape and are shaped by the development of this technology. While these collective concerns certainly deserve ethical, legal and societal scrutiny, comparatively little attention is paid to how genAI shapes its users on an individual level, in terms of identity, agency and moral attitude. What should I do when my hospital encourages me to use genAI to harvest my patients’ medical dossiers and generate pre-made answers? When my students want to use genAI to generate an essay for a critical thinking assignment, how should I respond? Or when my colleagues use genAI to generate new molecular bonds without being to explain the black-box process that lead to this? During this seminar, I will argue that we ought not to neglect this deeply individual moral dimension to our engagements with genAI systems. To show how we may address this concern, I articulate a moral approach based on virtue ethics. First, I show how virtue ethics enables us to capture relevant virtues and vices in our interactions with genAI, with examples from healthcare, education and chemistry. Second, I address how a moral dilemma for ethicists reflecting on the virtues of genAI, namely how to understand ‘prudence’ in this context, can be tackled. Third, I make some tentative suggestions about how we may translate virtue ethical considerations into design implementations.