Algorithmic Accountability: 
Designing for Safety

Due to Tuesday, February 12th weather conditions, this event has been rescheduled for March 4th. 

Vital services, such as communications, financial trading, healthcare, and transportation depend on sophisticated algorithms, some relying on unpredictable artificial intelligence techniques, such as deep learning, that are increasingly embedded in complex software systems. As high-speed trading, medical devices, self-driving cars, and autonomous aircraft become more widely implemented, stronger checks become necessary to reduce failures. Design strategies that promote human-centered systems, which are comprehensible, predictable and controllable can increase safety and make failure investigations more effective.

This lecture will be delivered by Ben Shneiderman who is a PWIAS International Visiting Research Scholar.

No registration is required.

This event is part of the AI & Society Seminar Series and is co-sponsored by the Peter Wall Institute for Advanced Studies and CAIDA: UBC ICICS Centre for Artificial Intelligence Decision-making and Action.