Talk Title: Learning to Code with AI

Speaker: Majeed Kazemi, PhD candidate at University of Toronto

Location: HBK 2105 and Zoom

Abstract:“In the evolving landscape of programming with generative AI, critical questions emerge around its impact on cognition, interaction, and learning. In this talk, I will present findings from my research on three key topics: (a) What are the implications of using AI when learning to code for the first time? Does AI enhance learning or foster over-reliance, potentially hindering outcomes? (b) How can we design novel interfaces that cognitively engage learners with AI-generated solutions—enhancing users’ ability to extend and modify code without creating friction? (c) How to design pedagogical AI coding assistants for educational contexts? I will discuss the design of CodeAid, results from its 12-week deployment in a large class of 750 students, and perspectives from students and educators.”

Bio: “Majeed is a PhD candidate in Computer Science at the University of Toronto, advised by Prof. Tovi Grossman. His research in Human-Computer Interaction liest at the intersection of programming, education, and AI. As a systems researcher, his work draws from learning sciences and interaction design to develop novel tools that address fundamental challenges surrounding interaction and cognition when integrating AI into programming. His work has been published at top-tier HCI venues such as CHI, UIST, IDC, and IUI, and his research in AI and education is among the most highly cited CHI papers of the past two years. Prior to his PhD, Majeed completed his PhD at the University of Maryland, where he worked with Prof. Jon Froehlich at the HCIL. During this time, he designed and built MakerWear–a tangible, modular electronic toolkit that enables young children to create interactive wearables–which earned a Best Paper Award at CHI.”