Generative AI-Powered Virtual Reality Pedagogy: Designing Intelligent Characters to Meet Educator Needs in Higher Education

Designing Generative AI-Powered Pedagogy for Virtual Reality in Higher Education

Emerging technologies like Generative Artificial Intelligence (GAI) and Virtual Reality (VR) are increasingly influencing education. This article presents an approach combining these technologies through Intelligent Virtual Reality (IVR), featuring two AI-driven virtual characters developed to enhance sustainability education in higher education settings.

Introduction

Generative AI, including Large Language Models (LLMs) such as GPT, can create meaningful content—text, images, or audio—based on extensive training data. VR provides immersive, interactive 3D environments that bring complex concepts to life. Together, they offer promising tools for education, especially when integrated thoughtfully.

Despite their potential, educators face challenges in adopting these technologies. Research shows AI can personalize learning, support critical thinking, and increase engagement, but practical integration into curricula remains limited. This calls for pedagogical designs that consider educators’ needs and the unique capabilities of GAI and VR.

Identifying Educators’ Needs

A two-part study collected input from 66 educators ranging from K12 to university levels. Interviews and open submissions revealed key challenges and opportunities:

• Challenges: Limited resources and funding, difficulty teaching complex or collaborative topics, lack of accurate and diverse perspectives, and insufficient digital tools that scale for large groups.
• Opportunities: VR’s ability to visualize complex systems (e.g., molecular journeys, magnetic fields, ecological cycles) and to demonstrate spatial and temporal scales. Educators also valued technologies that foster critical skills like problem-solving, futures-thinking, and systemic understanding.

Developing the Pedagogical Design

Based on these insights, a three-step design process created an IVR learning experience focused on sustainability education:

• Step 1: Select a reference framework—the sustainability competency framework—to guide learning objectives and instructional methods.
• Step 2: Define learning goals suitable for VR, emphasizing immersive understanding of sustainability concepts.
• Step 3: Design two GAI-powered virtual characters, Tero and Madida, serving as an information source and learning companion, respectively.

The characters were built on ChatGPT technology with iterative prompt engineering and tested with domain experts to evaluate their effectiveness.

Results and Evaluation

Feedback from experts showed that the characters successfully addressed 9 out of 12 educator-identified needs. They excelled at supporting student learning by providing personalized guidance and clarifying complex concepts through interaction.

Most comments highlighted the value of AI as a character that assists learners directly, with fewer remarks on AI supporting teachers. The virtual characters enabled immersive and engaging experiences that can make sustainability education more accessible and meaningful.

Discussion

This study demonstrates a practical model to integrate GAI and VR in higher education pedagogy. Designing AI characters within immersive environments allows real-time adaptation to student needs and supports educators in delivering complex content more effectively.

By addressing educator needs and leveraging current AI capabilities, such designs can help prepare both teachers and students for future digital fluency requirements. The approach also offers a blueprint for expanding AI and VR use beyond sustainability education, into other disciplines that benefit from immersive, interactive learning.

Methods Overview

The research began with a qualitative needs analysis involving 31 university teachers and 35 educators across various levels who shared ideas on emerging technologies in teaching. Inductive thematic analysis identified key themes about challenges and opportunities.

For the pedagogical design, the sustainability competency framework structured learning objectives and methods suited for IVR. Domain experts were engaged throughout the iterative design and prompt engineering of the AI characters. Their feedback informed refinements and validated the design’s alignment with educator needs.

For those interested in further exploring AI applications in education, resources and courses on generative AI and prompt engineering can be found at Complete AI Training.