Episode 39: Assessment and Community in the Age of AI

How can faculty assess learning when AI can produce most of the artifacts we used to grade, and what happens to campus community when students turn to chatbots instead of each other? Craig and Rob talk with Dr. Bill McCumber of Louisiana Tech University about both questions.

Overview

Bill McCumber, Associate Dean of Graduate Programs and Research at Louisiana Tech's College of Business, joins the show to describe two efforts: a voluntary, 17-person AI task force built around community rather than mandates, and a graduate finance pilot that replaces tests with AI-supported oral assessment at scale. The hosts press Bill on what oral exams actually measure, why faculty resist handing off grading, and what administrators should be doing right now. The conversation then turns to MIT's recent report on AI in teaching and learning, which the hosts argue buried its most important finding: AI's effect on the social fabric of campus life. The episode closes with a practical tip on AI model selection and usage quotas.

What you'll hear

Why an AI task force should be "culture work." Bill explains why he built a voluntary group of faculty, administrators, administrative assistants, and students (and why he kept Craig off it). His aim is community building and comparing notes rather than prescribing tools or syllabus policies, and the first brown bag features an AI skeptic.

The barriers keeping faculty from useful AI. Rob points to a lack of time to experiment and to university bureaucracy, including FERPA constraints that leave Copilot as the only approved tool at Washington State University. Bill argues that faculty need enough domain knowledge to "earn the right to automate."

Oral exams at scale, and their hidden noise. Bill describes a pilot in which students build projects with AI, then answer an unexpected question aloud; responses are de-identified, graded through a pipeline, and flagged for follow-up. Craig raises a concern that oral exams also measure verbal communication and stress response, and suggests a secondary one-on-one review for students who struggle.

AI in the student-managed investment fund. Bill's students used AI to turn a raw download of portfolio positions into an interactive dashboard with Sharpe ratios and correlation matrices in short order, work that would once have taken weeks of coding.

The MIT report's underreported finding. Craig notes that a New York Times headline focused on "cognitive surrender," which gets only a brief mention in the report; the more striking material concerns isolation and campus social cohesion. Rob and Bill share anecdotes about students returning to phone calls and office visits.

From lecturing to coaching

Bill's pilot moves foundational content (the "vocabulary lessons") online so class time can become coaching, group work, and "discussion under uncertainty." Rob connects this to his work as a soccer referee: practice builds the skills, and the match shows whether the drills paid off. Assessment in this model looks less like a test and more like the workplace, where, as Bill puts it, someone eventually checks whether you can do the thing. Craig pushes back on treating grading as binary, pointing out that faculty have delegated first-pass grading to teaching assistants for decades; AI could play a similar triage role for oral assessments while faculty keep judgment over the hard cases.

References mentioned

MIT Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training. (2026, August 13). Report of MIT's Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training. Report site: https://aiandeducation.mit.edu/ ; PDF: https://bpb-us-e1.wpmucdn.com/sites.mit.edu/dist/d/2418/files/2026/09/AI-Committee-Final-Report-Aug-13.pdf
Socrates' critique of writing as a proxy for knowledge (recorded in Plato's Phaedrus)
Bélanger, F., Van Slyke, C., & Crossler, R. E. Information Systems for Business: An Experiential Approach. Prospect Press. https://www.prospectpressvt.com/textbooks/b%C3%A9langer-information-systems-for-business-an-experiential-approach-5-1
Previous AIGTC episode with Tom Stafford on AI and student isolation https://www.aigoestocollege.com/building-ai-literacy-in-higher-education-programs-practice-and-community/
Tools named: Microsoft Copilot, Claude, ChatGPT (Astra), Fable 5

Questions for reflection

If oral assessment returns, how will you separate what a student knows from how comfortable they are speaking under pressure?
What is actually stopping faculty at your institution from using AI to reclaim time, and has anyone asked them?
If AI is thinning the social fabric of campus, what in your course gives students a reason to talk to each other?

About the show

AI Goes to College is a podcast for higher education professionals trying to make sense of artificial intelligence in their classrooms, their research, and their institutions. Co-hosted by Craig Van Slyke and Rob Crossler, the show focuses on practical, evidence-based perspectives on AI in higher education without the hype.

Website: https://aigoestocollege.com

Newsletter: https://aigoestocollege.substack.com/

Email: craig@aigoestocollege.com and Rob.Crossler@aigoestocollege.com

Mentioned in this episode:

AI Goes to College Newsletter

Assessment and Community in the AI Age: A Conversation with Dr. Bill McCumber

AI Goes to College

Assessment and Community in the AI Age: A Conversation with Dr. Bill McCumber