Ethics of Artificial Intelligence

Catalog description

This course will familiarize students with foundational concepts and emerging ideas in the ethics of artificial intelligence and their implications for public policy. It will be broken down into three sections: (1) the ethics of machine learning; (2) the moral status of AI; and (3) AI and the distant future. The first section will consider such topics as the ethical implications of unconscious bias in machine learning (e.g., in predictive text, facial recognition, speech dialogue systems); what constraints should govern the behavior of autonomous and semi-autonomous machines such as drones and smart cars; whether AI can undermine valuable social institutions and perhaps to democracy itself and what might be done to mitigate such risk; and how automation might transform the labor economy and whether this morally desirable. The second section turns to the question of our moral obligations toward (some) artificial intelligences. Here, we will ask what grounds moral status in general and how this might apply to artificial intelligences in particular, including how should we should balance moral obligations toward (some) AIs with competing obligations toward human beings and other creatures with morally protectable interests. The final section will look to the far distant future and consider how (if at all) we might identify and estimate future threats from AI and what might be done today to protect all those who matter morally.