Can We Trust AI-Generated Educational Content? Comparative Analysis of Human and AI-Generated Learning Resources
Project Overview
The document investigates the role of generative AI, particularly large language models (LLMs), in enhancing educational resources, with a focus on an introductory programming course. It highlights that AI-generated materials can achieve a quality comparable to those produced by students, indicating a promising avenue for supplementing traditional teaching methods. Despite this, the analysis reveals discrepancies in content variety and length, suggesting that while AI can be a valuable tool, it may not fully replace the unique contributions of student-generated content. The findings emphasize the need for further research to explore the long-term implications of integrating AI in education and its adaptability to cater to different learning styles and needs, ultimately aiming to enrich the educational experience and outcomes.
Key Applications
Large language models (LLMs) for generating learning resources
Context: Introductory programming course at a university level, targeting first-year engineering students
Implementation: Students created code examples as part of a lab session, while LLMs generated similar resources using prompts based on provided examples.
Outcomes: AI-generated resources were perceived as equivalent in quality to those generated by students. AI-generated content offered more concise functions, while student content showed greater variety.
Challenges: AI-generated resources had limited variety and closely mirrored the provided prompts, raising questions about adaptability to diverse learning needs.
Implementation Barriers
Practical Barrier
Challenges around ensuring the quality and relevance of AI-generated content.
Proposed Solutions: Further research is needed to evaluate the quality of LLM-generated content and to develop methods for integrating diverse prompts to improve adaptability.
Ethical Barrier
Concerns about bias in generated content and academic integrity.
Proposed Solutions: Recommendations for future research focused on developing practical, ethical, and human-centered innovations in the use of LLMs.
Project Team
Paul Denny
Researcher
Hassan Khosravi
Researcher
Arto Hellas
Researcher
Juho Leinonen
Researcher
Sami Sarsa
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Paul Denny, Hassan Khosravi, Arto Hellas, Juho Leinonen, Sami Sarsa
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