Beyond Traditional Teaching: The Potential of Large Language Models and Chatbots in Graduate Engineering Education
Project Overview
The document discusses the integration of generative AI, particularly large language models (LLMs) and chatbots, in graduate engineering education, focusing on a fluid mechanics course. It highlights the benefits of these technologies, such as personalized learning experiences, instant feedback, collaboration, and increased student engagement. The use of AI chatbots is emphasized for their potential to enhance learning, offer tailored support, and facilitate project-based learning. However, the document also addresses significant challenges, including technical difficulties, privacy concerns, and the necessity for appropriate pedagogical frameworks to guide their implementation. Ethical implications and the need for a balanced approach are underscored, advocating for instructors to be well-informed and for ongoing evaluation of AI technologies within educational curricula. Overall, the findings suggest that while generative AI holds promise for enriching education, careful consideration of its integration is essential to maximize benefits while mitigating risks.
Key Applications
AI-driven chatbots for learning support
Context: Higher education across various disciplines, including fluid mechanics and design courses. Targeting students and educators, these chatbots are used in learning management systems and project-based learning environments.
Implementation: Chatbots integrated into educational frameworks to assist with queries, provide personalized feedback, and support collaborative learning in design projects. They are built using question banks from course materials and designed to engage students in interactive learning experiences.
Outcomes: ['Promoted self-paced and collaborative learning', 'Provided instantaneous feedback', 'Enhanced student engagement', 'Improved learning outcomes and collaboration']
Challenges: ['Ethical implications and inaccuracies in responses', 'Technical difficulties and limited understanding of pedagogical integration', 'Privacy concerns and potential resistance from educators']
Conversational AI chatbots for language practice
Context: Various educational settings, particularly in language learning environments, where chatbots serve as interactive tools for students to practice and engage with the language.
Implementation: Utilization of AI-driven chatbots to facilitate language practice through interactive conversations, providing personalized feedback and engagement tailored to individual learning needs.
Outcomes: ['Increased engagement in language learning', "Potential for personalized feedback tailored to learners' needs"]
Challenges: ['Effectiveness of chatbots for language learning still under investigation']
Implementation Barriers
Technical Barrier
LLMs can produce incorrect or misleading information, known as hallucination. Additionally, there are technical difficulties in integrating AI chatbots into existing educational frameworks.
Proposed Solutions: Enhancing model training with high-quality data and incorporating human feedback. Develop robust integration guides and support systems for educators.
Privacy Barrier
Risk of privacy issues due to training data potentially containing personal information, along with concerns regarding data privacy and security when using chatbots.
Proposed Solutions: Expunging personal data from datasets, implementing automated assessments, and enforcing strict data protection measures with transparency about data usage.
Overreliance Barrier
Users may become overly reliant on chatbots, leading to a decline in critical thinking skills.
Proposed Solutions: Encouraging users to critically evaluate model outputs and providing comprehensive documentation.
Pedagogical Barrier
Lack of understanding of how to effectively integrate AI tools into pedagogy.
Proposed Solutions: Provide professional development for educators on utilizing AI in teaching.
Project Team
Mahyar Abedi
Researcher
Ibrahem Alshybani
Researcher
Muhammad Rubayat Bin Shahadat
Researcher
Michael S. Murillo
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Mahyar Abedi, Ibrahem Alshybani, Muhammad Rubayat Bin Shahadat, Michael S. Murillo
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