HPS Visiting Assistant Professor, Emanuele Ratti, will be presenting his paper titled "Considerations on the Ethics of AI: Microethics, Virtues, and Pluralism" at the Great Lakes Workshop on Data Science to be held at the University of Notre Dame from September 20th-22nd, 2019. An abstract for his paper is presented below. For more information about this conference, please follow this link: https://sites.google.com/view/great-lakes--workshop-on-data
Abstract
AI ethics shapes digital governance and digital regulation by discussing the morally salient features of AI tools and the role of AI in society. Currently, AI ethics can be conceived as a macroethics (or hard ethics), namely as a discipline aimed at elaborating a general ethical evaluation of an emerging technology, in particular by providing shared moral principles that practitioners should follow. While AI ethics conceived as a macroethics is a positive thing, I will claim that this approach should be complemented by another approach. By being only a macroethics, AI ethics neglects the figure of the data scientist. In particular, there are two risks. First, general ethical principles are difficult to be applied to the day-to-day routine of the data scientist. Next, AI ethics as a macroethics lacks teachability, and there is a risk that it may be seen in light compliance concerns as many RCR trainings, rather than as a genuine attempt to illuminate ethical and societal issues for data scientists. In this talk, I will introduce a microethical approach to AI ethics that can complement macroethics. Microethics comes from clinical ethics, and it emphasizes the wealth of small ethical decisions that a physician takes when interacting with patients. Therefore, it is an approach to clinical ethics that is attentive to the clinical practice. In data science, a microethical approach will be concerned with two types of considerations. First, how AI systems are embedded in larger systems, which include also the individuals affected by the tools and society as a whole. Therefore, the design of AI tools in a direction or another shapes individuals’ lives in a way that is ethically relevant, by forcing certain patterns of behavior rather than others. Next,AI design is not morally neutral, in the sense that values are sneaked in all the time, and microethics should be of help in identifying those values. A microethical approach to AI ethics will help practitioners to habituate their reasoning to these ethical analyses, by fostering the cultivation of a moral virtue that I call ‘moral attention’.