Provably Beneficial Artificial Intelligence
Prof. Russell discusses an approach to ethical AI based on the idea that AI systems should be beneficial to humans, with the key caveat that what counts as “beneficial” is very unlikely to be fully specified. Issues include social aggregation and social choice, laws, equity, sadism, pride, envy, and mental integrity.
This is a recorded talk
Conference participants should access the talk before Wed 14:30-15:30 at this link.
Panelists will discuss Russell‘s pre-recorded talk:
-
Berndt Muller, Senior Lecturer, Computer Science, Swansea University
-
Danushka Bollegala, Professor, Computer Science, University of Liverpool
-
Tom Maibaum, Emeritus Professor, Computer Science, McMaster University
