BoilerTAI: A Platform for Enhancing Instruction Using Generative AI in Educational Forums
Project Overview
This document explores the application of generative AI, specifically BoilerTAI, in educational settings. BoilerTAI leverages OpenAI's GPT-4 to assist instructional staff in online educational forums by generating responses to student inquiries, which are then reviewed and approved by Teaching Assistants. The platform, grounded in Vygotsky's socio-cultural theory, was implemented in computer science courses and demonstrated enhanced efficiency and improved response quality, leading to positive student feedback. The study underscores the potential of AI to augment human expertise in teaching and learning. However, it also emphasizes the critical importance of human oversight to address challenges related to accuracy, ethical considerations, and the inherent limitations of AI in educational applications.
Key Applications
BoilerTAI
Context: First and second-year computer programming courses, online educational forums
Implementation: Integrates with Ed-Discussion to generate responses using OpenAI's GPT-4. AI-generated responses are reviewed, modified (reprompting or editing), and approved by TAs before posting.
Outcomes: Improved efficiency of responses to student inquiries (75% of participants reported improvement), enhanced quality of responses, positive student reception of AI-generated responses (75% of TAs observed positive feedback), and reduced cognitive load for TAs.
Challenges: Need for oversight by TAs (50% of TAs reported needing to significantly alter responses about half the time). Challenges related to accuracy, ethical considerations, and potential for misleading information.
Implementation Barriers
Accuracy and Reliability
Concerns about the accuracy of AI-generated responses and the potential for misleading information or 'hallucinations.'
Proposed Solutions: Human oversight by TAs to review, modify, and approve AI-generated responses. Prompt engineering to guide the AI to provide contextually appropriate and pedagogically sound responses.
Ethical and Academic Integrity Considerations
Maintaining academic integrity and ensuring that AI-generated responses do not mislead students. Differing perceptions between students and instructors regarding AI's role and potential risks.
Proposed Solutions: Human oversight to ensure accuracy and contextual appropriateness. The human-in-the-loop approach reduces the risk of disseminating incorrect or misleading information and preserves the educational experience’s interpersonal element by incorporating human oversight. Careful management and clear communication about the use of AI. The platform is designed to maintain the appearance and authority of a direct response from the instructional staff, to prevent usage of this tool from disrupting regular student interaction.
Project Team
Anvit Sinha
Researcher
Shruti Goyal
Researcher
Zachary Sy
Researcher
Rhianna Kuperus
Researcher
Ethan Dickey
Researcher
Andres Bejarano
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Anvit Sinha, Shruti Goyal, Zachary Sy, Rhianna Kuperus, Ethan Dickey, Andres Bejarano
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gemini-2.0-flash-lite