Authentic Learning with Agentic AI in TBL

Workshop overview

Using agentic AI to scale authentic role play

Authentic tasks involve ill-structured challenges and roles that help students rehearse for the complex ambiguities of adult and professional life. Role play is one powerful way to do this, from simulated patients in medical education to scenarios in business, engineering, science, psychology, politics, and other disciplines.

This workshop focuses on a form of role play where students interact with an agent, such as an interviewee, patient, business owner, or community representative. The agent may be unwilling or unable to give fully open, accurate, or complete responses, requiring students to ask careful questions and practise appropriate communication in a safe environment.

Background and rationale

From rich role play to scalable AI personas

Role-play approaches are rich and valuable, but they can also be time consuming. Agentic AI offers an opportunity to scale these approaches by configuring an agent with an appropriate personality, context, constraints, and conversational behaviour.

Participants take the role of students using agentic AI, experiencing how it feels to interact with AI in order to achieve a task or goal, then critiquing how this approach might transfer into their own TBL practice.

In the workshop, participants interact with AI-powered personas representing historical figures. They engage with Alexander Graham Bell and Elisha Gray, asking questions about their competing claims to the telephone patent.

To guide the interaction, participants use TRACI, a structured prompting approach that helps define the task, role, audience, format, and intent of their questions. This supports both the historical inquiry and the broader skill of using AI critically and effectively.

Pre-learning

Key concepts before the workshop

Authentic learning

Authentic learning builds on situated cognition, where knowledge is understood as part of an activity, context, and culture. This supports cognitive apprenticeship, teamwork, and collaboration.

Authentic learning environments typically emphasise real-world relevance, ill-defined problems, collaboration, reflection, and role play.

AI agents

AI agents can perceive and act upon their environment with a degree of autonomy. Agentic AI extends this idea, often referring to multiple coordinated AI agents.

In this workshop, participants work with simple AI agents configured to behave as personas within a TBL learning activity.

Prompting framework

TRACI

Task, Role, Audience, Create, Intent

View TRACI resource

TRACI helps participants craft better prompts by making the purpose, perspective, audience, output format, and learning intent explicit.

Task

The general activity ChatGPT is being asked to perform.

Role

The perspective or persona ChatGPT should adopt.

Audience

Who the response is intended for.

Create

The format or medium of the requested output.

Intent

The underlying purpose or goal of the generated text.

Alternative framework

Participants may also use CRAFT: Context, Role, Audience, Format, and Tone.

LAMS

Running the workshop activity in LAMS

In the workshop, participants use LAMS to experience the student journey and interact with AI-powered personas within a structured learning sequence.

Preview Lesson

Authentic AI Agents in TBL - Graham Bell & Elissha Gray and the Telephone Patent Case

A TBL Lesson using AI Agents that impersonate Graham Bell and Elisha Gray. Students discuss and interact with Bell and Gray AI agents to assess the merit of the telephone patent.

Lesson details and download