Pedagogy Challenges in the AI Era

Apply-Then-Consult

Mitch Gerhardt

IDEEAS Lab

2026-05-20

Agenda

9:15–10:00 | Pedagogy Challenges in the AI Era: Apply-Then-Consult

The Performance of Teaching in an AI World
The Scenarios
Reflections

The Performance of Teaching in an AI World

Why We’re Doing This (and Why It Feels Uncomfortable)

Erving Goffman

Every classroom is a “front stage” where instructors perform expertise, authority, and certainty
AI systems disrupt the script because instructors are suddenly unrehersed at various AI-related interactions
Most professional development asks us to “perform” as if nothing has changed, which is unsustainable and inauthentic because we haven’t rehersed the new script of teaching with AI

The backstage problem

Where do you actually work out how to handle a student who used AI better than you expected?
Where do you practice saying “I don’t know” when a student asks you a question about AI that you don’t have an answer for?
Where is the space to fail before you’re on stage?

“All the world is not, of course, a stage, but the crucial ways in which it isn’t are not easy to specify” (Goffman, 1959)

For this segment: Adapting to AI to education is a rehersal problem and this workshop is backstage time. Let’s use it

Goffman, E. (1959). The Presentation of Self in Everyday Life. Knopf Doubleday Publishing Group.

Expertise Development as Rehersal

What drives top performance?

Deliberate practice: focused, goal-oriented, feedback-rich practice of specific skills
Rehersal: practicing the performance of a skill in a realistic context, including the social and emotional aspects of the performance

What do we typically do?

Learning about a new pedagogical approach in a workshop, then trying to implement it directly in the classroom without practice or feedback
Thinking that “good” practice is just about doing the activity, rather than considering feedback signals, goals, and the social context of the performance

Why is this problematic?

Teaching with AI is a new performance that requires new skills and instincts, which can’t be developed without practice
This “practice” must be more than just trying the activity: it must include feedback, reflection, and social interaction to develop the necessary skills and confidence
Without backstage rehearsal, instructors are likely to feel unprepared and may struggle to adapt their teaching effectively in the AI era

“This is the basic blueprint for getting better in any pursuit: get as close to deliberate practice as you can (Ericsson, 2016, p. 103).”

Ericsson, K. A. (with Pool, R.). (2016). Peak: Secrets from the New Science of Expertise (1st ed). HarperCollins Publishers. · Ericsson, K. A., Charness, N., Feltovich, P. J., & Hoffman, R. R. (Eds.). (2006). The Cambridge Handbook of Expertise and Expert Performance. Cambridge University Press. https://doi.org/10.1017/CBO9780511816796

Deliberate Practice

What is required for deliberate practice?

Developing established abilities with the assistance of a coach or mentor who can provide feedback and guidance

Going outside of one’s comfort zone: “demands near-maximal effort, which is generally not enjoyable” (Ericsson, 2016, p. 99)

Setting specific, achievable goals with clear feedback signals to track progress and adjust practice strategies

Engaging in focused, deliberate, structured practice designed to address goals. AKA concentration

Feedback to modify efforts and practice strategies in response to feedback and progress toward goals

Developing mental representations of the skills being practiced, which are refined to become more accurate and effective

Deliberately modifying and developing previous abilities, skills, and mental representations

Ericsson, K. A. (with Pool, R.). (2016). Peak: Secrets from the New Science of Expertise (1st ed). HarperCollins Publishers.

Setting Up the Backstage Practice Space

Table Groupings (assigned, not self-selected)

Mix by [disipline] OR [AI experience] OR [institution type]
- Rationale: heterogeneous tables surface assumption gaps faster
Each table: 4-5 people, 2 scenarios, 25 minutes (10 min per scenario, 5 min for debrief)

What we’re practicing, not solving

Reasoning through a novel pedagogical problem with peers
Noticing when your instinct is to perform an answer vs. think one through
Identifying what you know, what you don’t know, and what you need to know to make a decision

What success looks like

Not: “We solved the scenario!”
Yes: “We identified what we’d need to know, what we’d change, and what we’re uncertain about”

Shared board protocol

Each table posts: [What we’d change] + [What we’re still unsure about]
Uncertainty is required! Tables with no uncertainties posted have not gone deep enough

The Scenarios

Scenarios

Scenario 1: The Identical Reports Problem

You assigned an AI-assisted design problem, but while grading, half the class submitted nearly identical AI-generated reports with minimal original analysis.
What do you change for next time and what does “original analysis” mean now?

Scenario 2: Infrastructure Constraints

You want students to use AI to explore failure modes, but your classroom has unreliable Wi-Fi and limited access to AI tools.
How do you design the activity so the constraints become features, not failures?

Scenario 3: The Homework Legitimacy Challenge

A colleague tells you “If AI can solve it, the homework is testing the wrong thing.”
How do you respond and what do you change about your assignments?

Scenario 4: Proving to Leadership

Your department chair asks you to demonstrate that AI integration isn’t lowering academic standards after a parent complained about AI use in the classroom.
What evidence do you collect? What do “standards” mean when AI handles parts of the work?

Scenario 5: The Large Class Problem

You are teaching 120 students and require documented AI interactions logs, but find grading them to be overwhelming.
How do you design AI-integrated work that scales and what do you give up?

Scenario 6: Uneven Access

Some students in your program can access powerful AI systems and internet, while others have much more limited access.
How do you design AI-integrated learning that doesn’t deepen existing inequalities?

What Emerged from the Scenarios?

These are orientations to practice, not solutions. You just did that practice, so how does it continue?

From This Backstage to Your Front Stage

The transfer problem

When we end, the front-stage pressure returns immediately
Most institutions have no formal space for pedagogical rehersal around AI

Three backstage spaces you can build

The teaching problem circle: Monthly peer meeting where one person brings a real AI-in-teaching dilemma, others work it like a case. No solutions are required, only sharper questions.
Assignment autopsy: After each major assignment, do a brief written reflection: “Where did AI reshape what students did? Did that help or hurt learning? What would I change?”” Shared with one trusted colleague.
Pre-semester scenario sprint: Before courses begin, with a peer: “Here are the three AI moments I expect to face. What’s my plan?” Spoken, not written, and focusing on actual rehearsal.

Returning to Goffman

Front stage competence and confidence is built backstage
Every expert performance you’ve seen and given started as a rehersal others didn’t see
AI adaption requires you to be a learner again, which means you need backstage spaces to be a beginner again
Build it deliberately, protect it institutionally, and invite others in

Goffman, E. (1959). The Presentation of Self in Everyday Life. Knopf Doubleday Publishing Group.

Beginner’s Mind, Expert’s Mind

“In the beginner’s mind there are many possibilities, but in the expert’s there are few” - Shunryu Suzuki

Suzuki, S., Dixon, T., Smith, H., & Baker, R. (2006). Zen mind, beginner’s mind. Shambhala.

Thank You

Questions?

Mitch Gerhardt

mitchg@vt.edu

github.com/mitchg10

linkedin.com/in/mitchell-gerhardt