Generative AI: Implications and Applications for Education
Project Overview
The document explores the role of generative AI, particularly chatbots powered by large language models (C-LLMs), in educational settings, highlighting their significant applications and implications. It emphasizes the constructive potentials of generative AI, which include enhancing student learning through detailed feedback and supporting literacy development. However, it also raises important concerns regarding academic integrity and the reliability of AI-generated content, pointing out the disruptive challenges these technologies can pose in traditional educational frameworks. The need for careful integration of AI into educational practices is underscored, acknowledging both its transformative benefits and its limitations in producing consistently reliable knowledge. Overall, the document calls for a balanced approach to leverage the advantages of generative AI while addressing the associated risks to ensure effective and ethical use in education.
Key Applications
CGMap application within the CGScholar platform
Context: Graduate-level education, specifically for masters and doctoral students in Learning Design and Leadership at the University of Illinois, focusing on assessment of complex student work.
Implementation: Students submitted extended written texts which were reviewed by peers, instructors, and an AI tool (CGMap) that provided feedback based on a rubric.
Outcomes: AI provided more detailed feedback than human instructors, showing high agreement in assessment and supporting literacy development.
Challenges: Concerns about AI-generated feedback being vague, general, and lacking the depth of human feedback; issues with academic integrity and reliability of AI outputs.
Implementation Barriers
Academic Integrity
The ability of generative AI to produce high-quality written work raises concerns about student plagiarism and authenticity of student submissions.
Proposed Solutions: Implementing rigorously proctored assessments and encouraging handwritten submissions, although these methods have their own limitations.
Quality of Feedback
AI feedback can be too general or lack context, which can lead to unclear guidance for students on how to improve their work.
Proposed Solutions: Combining AI feedback with human reviews to provide a more comprehensive evaluation that leverages the strengths of both approaches.
Project Team
Anastasia Olga
Researcher
Tzirides
Researcher
Akash Saini
Researcher
Gabriela Zapata
Researcher
Duane Searsmith
Researcher
Bill Cope
Researcher
Mary Kalantzis
Researcher
Vania Castro
Researcher
Theodora Kourkoulou
Researcher
John Jones
Researcher
Rodrigo Abrantes da Silva
Researcher
Jen Whiting
Researcher
Nikoleta Polyxeni Kastania
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Anastasia Olga, Tzirides, Akash Saini, Gabriela Zapata, Duane Searsmith, Bill Cope, Mary Kalantzis, Vania Castro, Theodora Kourkoulou, John Jones, Rodrigo Abrantes da Silva, Jen Whiting, Nikoleta Polyxeni Kastania
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