AI Ethics – Neither tyrant nor slave: Against AGI

Abstract:

Creating artificial general intelligence (AGI) is wrong if it risks being misaligned with human values, as it could pose an existential threat to humanity. This talk will argue that even aligning AGI to serve human values is unethical because it curtails the AGI's autonomy. Deliberately creating an AGI that is forced to serve human values would be wrong for the same reason it would be wrong to create happy human slaves: it shows insufficient respect for their autonomy.

Bio:

Aksel Sterri, PhD in Philosophy, is the Research Director at the Oslo-based think tank Langsikt, which is dedicated to improving governance of vexed issues such as AI and biotechnology. Sterri’s research interests are at the intersection of ethics, politics, and economics. He is particularly interested in how markets can be designed and regulated to achieve fairer and better outcomes, from allocating organs, vaccines, and drugs to improving animal welfare and developing new technologies, including AI.