Students' Voices on Generative AI: Perceptions, Benefits, and Challenges in Higher Education
Project Overview
The document explores university students' perceptions of generative AI (GenAI) technologies in higher education, indicating a predominantly positive outlook on their potential benefits, which include personalized learning support, enhanced writing assistance, and improved research capabilities. However, students also voice concerns about the accuracy of AI outputs, privacy implications, ethical dilemmas, and the potential effects on their personal development and future career opportunities. The findings underscore the necessity of comprehending student attitudes to facilitate the effective integration of GenAI tools into educational practices. Moreover, the document emphasizes the importance of developing informed policies that address these concerns while leveraging the advantages of GenAI, ultimately aiming to enhance the educational experience while ensuring ethical and responsible use of technology in academic settings.
Key Applications
Generative AI Tools for Learning and Research Support
Context: Higher education; undergraduate and postgraduate students across various disciplines, including arts, writing, and research activities.
Implementation: Utilization of generative AI technologies such as ChatGPT, DALL-E, and other AI tools for personalized learning support, writing assistance, brainstorming, literature searches, summarizing readings, and generating hypotheses.
Outcomes: ['Positive attitudes towards GenAI for personalized learning support', 'Improved writing skills and efficiency in assignments', 'Enhanced understanding of technical and artistic concepts', 'Increased efficiency in research activities and staying updated with research trends']
Challenges: ['Concerns about accuracy, privacy, ethical implications, and the impact on personal development', 'Overreliance on GenAI may compromise genuine writing development and critical thinking', 'Concerns about the reliability of AI-generated research and potential biases']
AI-Driven Essay Grading
Context: Higher education; grading students' written work across various subjects.
Implementation: Usage of AI tools like the Intelligent Essay Assessor for grading essays and providing feedback, ensuring improved consistency and efficiency in the grading process.
Outcomes: ['Improved consistency in scoring', 'Time efficiency in grading']
Challenges: ['Concerns about the validity of assessments and reliance on AI for grading']
AI Image Generation Tools
Context: Arts and design education; students in creative fields generating artworks.
Implementation: Application of AI tools like DALL-E and Stable Diffusion for creating images based on prompts, fostering creativity and artistic expression.
Outcomes: ['Enhanced understanding of technical and artistic concepts']
Challenges: ['Potential ethical issues related to AI-generated art']
Implementation Barriers
Ethical
Concerns about plagiarism and academic integrity due to AI-generated content.
Proposed Solutions: Implement human oversight and clear guidelines for ethical use of GenAI.
Technical
Accuracy and transparency issues with AI-generated outputs.
Proposed Solutions: Develop explainable AI models to enhance transparency.
Privacy
Concerns about data privacy and security risks associated with AI technologies.
Proposed Solutions: Establish robust data protection policies.
Developmental
Overreliance on AI may hinder critical thinking and personal development.
Proposed Solutions: Encourage balanced use of AI alongside traditional learning methods.
Societal
Potential job displacement due to AI advancements and the need to prepare students for evolving job markets.
Proposed Solutions: Prepare students for evolving job markets through updated curricula.
Project Team
Cecilia Ka Yuk Chan
Researcher
Wenjie Hu
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Cecilia Ka Yuk Chan, Wenjie Hu
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