Seminar: Understanding and Epistemic Trust in Intelligent Machines

What do artificial intelligence systems "understand"? Recent advances appear to expand what machines grasp, with image recognition systems seeming to understand objects and scenes and large language models seeming to understand language. The question matters not only for assessing a system's intelligence but also for evaluation practices that ensure the safe and responsible deployment of AI. Drawing on philosophy and cognitive science, and informed by current AI practice, we develop a framework for asking more precise questions and making more precise claims about machine understanding. This framework provides critical implications for assessing and advancing machine understanding, as well as building epistemic trust in human-machine interaction.

Huili Chen is an Assistant Professor at the University of Toronto's Faculty of Information. Before joining UofT, she completed a Presidential Research Fellowship in the Concept & Cognition Lab at Princeton, a Berkman Klein Fellowship at Harvard, and a PhD in the Personal Robots Group at the MIT Media Lab. Huili's research sits at the intersection of Cognitive Science, Behavioral Science, HCI, and Social Robotics. At the moment, she is also a visiting researcher at CHM.

Alternatively, join online:

Microsoft Teams

Meeting ID: 376 635 184 201 605

Passcode: wX6Nx9xJ