Review article
https://doi.org/10.24141/1/10/2/7
Utilizing the GPT model to teach physics for better learning and understanding of scientific concepts
Ivana Vrdoljak
; University of Applied Health Sciences, Zagreb, Croatia
*
* Corresponding author.
Abstract
Artificial intelligence (AI) is increasingly shaping the landscape of education, with models such as GPT (Generative Pre-trained Transformer) offering new possibilities to enhance both teaching and learning. This paper explores the potential of GPT in teaching physics, particularly how it can assist both students and lecturers in solving complex problems while offering interactive and personalised learning experiences to gain a deeper understanding of scientific concepts. The paper highlights how GPT can generate tailored tasks, quizzes, and assignments catering to specific student competencies. Additionally, it can guide students step-by-step through solutions to difficult physics problems, offering real-time feedback and explanations that promote active learning.
Beyond problem-solving, GPT’s ability to support interactive learning is investigated, allowing students to ask specific questions about topics like Newton’s laws or electromagnetism and receive instant, tailored responses. The paper also discusses the potential of GPT to complement simulations and visualisations, further helping students to fathom abstract concepts. However, the challenges associated with reliability, ethics, and the evolving role of educators in the age of AI are critically analysed. GPT’s limitations, potentially providing incorrect or incomplete information, underscore the importance of upholding the role of lecturers who can offer contextual guidance and foster critical thinking.
As education increasingly adopts AI, the paper emphasises
the need for a balanced approach, using GPT as a supplementary tool to enhance learning while ensuring students develop a deep and thorough understanding of physics based on a combination of AI and traditional teaching methods.
Keywords
Hrčak ID:
324310
URI
Publication date:
19.12.2024.
Visits: 0 *