AI Ethics: Law, AI and Leviathan

Abstract, Anders Sandberg:

Law and many other institutions produce goods like peace and prosperity by acting as outsourced, extended cognition running on a substrate based on human autonomous decisions. With AI, other options for outsourcing cognition become possible and (eventually) likely far more effective. However, these systems do not need human autonomy to do their jobs, suggesting a worrisome prediction that even if AI is successfully aligned with human values we may end up with a world that is somewhat dystopian. I will discuss ways out of this second order alignment problem.

Bio:

Anders Sandberg is senior research fellow at the Future of Humanity Institute, University of Oxford. He is also a member of the Mimir Center for Long Term Futures at Institutet för Framtidsstudier in Stockholm. His research deals with global catastrophic and existential risk, ethics of human enhancement, emerging technology, long term futures, and reasoning under extreme uncertainty.