Skip to main content Skip to navigation

The Textbook of Tomorrow: Rethinking Course Material Interfacing in the Era of GPT

Project Overview

The document examines the incorporation of generative AI in education, particularly through the creation of an intelligent textbook featuring a virtual teaching assistant that enhances student interaction with course materials. This AI agent, powered by advanced technologies like the GPT-4o model, can respond to inquiries, offer summaries, and generate quiz questions, facilitating a more engaging and personalized learning experience. The initiative seeks to address the limitations of conventional textbooks by fostering a more dynamic educational environment. Ongoing user studies are being conducted to evaluate the effectiveness of this approach and inform future enhancements, ultimately aiming to improve learning outcomes and student engagement through the innovative use of AI in educational settings.

Key Applications

Intelligent Textbook with embedded virtual teaching assistant (Jill Watson)

Context: Higher education, specifically online courses

Implementation: Developed using ReactJS and integrated with a Python implementation of RAG leveraging GPT-4o. Students can chat with their readings, receive summaries, and generate quiz questions.

Outcomes: Expected to enhance cognitive engagement and facilitate deeper understanding through interactive Q&A. Initial user studies indicate increased discourse.

Challenges: Latency in response times, uncertainty regarding the agent's knowledge scope, and variability in user engagement.

Implementation Barriers

Usability Barrier

Slow generation of responses from the AI agent which can lead to loss of trust in the tool.

Proposed Solutions: Implement real-time streaming of answer outputs and limit token count for concise responses.

Knowledge Scope Barrier

Uncertainty among users regarding the specific knowledge of the AI agent and its relevance to the text.

Proposed Solutions: Add in-context highlighting for agent's answers and ensure responses are clearly tied to the text.

Project Team

Audrey Olson

Researcher

Pratyusha Maiti

Researcher

Ashok Goel

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Audrey Olson, Pratyusha Maiti, Ashok Goel

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies