Martyn Clark
Data Scientist,
University of Maryland
AI, AI Everywhere, and Not a Chance to Think: How to Slow Down and Make Smart Decisions about AI
Thursday, October 01, 2026 | 1:00PM–1:45PM MT
Session Type:
Breakout Session
Delivery Format:
Presentation
Generative AI tools for teaching and learning have rapidly proliferated in recent years, but there is still very little data regarding whether these tools actually benefit education. Although some small-scale studies have taken place, most of these occurred within one or a few courses and focused on limited course outcomes. The Division of Information Technology (DIT) at the University of Maryland (UMD) developed its own generative AI tool, the Virtual Study Assistant, which is integrated within the Learning Management System and references course-specific material to provide responses to students that are tailored to course content. But rather than deploy this tool immediately to the full campus, DIT partnered with researchers in UMD’s College of Education to conduct a large-scale, systematic investigation of its effectiveness with over 40 instructors and 3,000 students. In this session, we will demonstrate the tool itself, share the methods by which we tested it, and discuss our findings from the study. More broadly, we will discuss what it means to slow down and make data-informed decisions in an era where everything feels like it is moving too fast.
