Title: Making Good Decisions and Getting AI to Do the Same

Presented By: Sheila A. McIlraith, Department of Computer Science

Abstract: As we contemplate a future in which AI systems are making decisions about everything from how long to toast our bagel to how fast our car should be driving on the icy roads, how do we ensure that these AI systems are making good decisions on our behalf? It has been suggested that highly autonomous AI systems adhere to the Value Alignment principle — that they be designed so that their goals and behaviours can be assured to align with human values throughout their operation — but how do we go about doing this? In this talk I will discuss technical approaches to building autonomous systems that “do the right thing” and the challenges to realizing this objective as we contemplate the elusive path from toasters to Artificial General Intelligence.

For more information, visit the Centre for Ethics event page.