Primary speaker:
Donny Wong, Teaching Applications Technical Lead, Department of Computer Science
Description:
The integration of Large Language Models (LLMs) into educational platforms often sparks a vital debate: is AI intended to replace the educator? In this presentation, we explore a "human-in-the-loop" approach to AI within MarkUs, a platform for student work submission and grading at the University of Toronto.
The core of this project is not the replacement of the Teaching Assistant, but the augmentation of their capabilities. We will examine how AI-assisted tools can be leveraged within MarkUs to automate the mechanical aspects of grading — such as syntax verification and basic style consistency — which currently consume a disproportionate amount of TA hours. By offloading these routine tasks to an AI, we empower TAs to spend their limited time on nuanced, high-level feedback and personalized student support that AI cannot replicate.
This will be a lecture with live demonstration of the proposed AI-assisted workflows within the MarkUs interface. The session will conclude with Q&A.
