The Future of Education

This paper explores the evolving role of generative artificial intelligence in programming education, drawing on firsthand classroom experience and applied research in natural language processing. Based on our teaching practice at the Academy of Applied Technical and Preschool Studies in Serbia, where we lead courses such as Software Engineering, Programming Fundamentals, and Big Data Analytics, the study investigates a significant shift in student learning behavior. Students are increasingly turning to tools such as ChatGPT, GitHub Copilot, and Google Bard for code generation, assignment completion, and project development, often favoring these resources over direct engagement with instructors.

While these technologies offer enhanced learning opportunities and promote self-guided exploration, they also introduce challenges related to academic integrity, originality, and conceptual mastery. From a teaching standpoint, this change requires a rethink of how we assess students and the creation of new teaching methods that fit with learning environments that use AI.

The analysis employs transformer-based models, including CodeBERT, to evaluate semantic similarity across student submissions and detect duplication and signs of overreliance on generative tools. Additional techniques include TF-IDF vectorization and cosine similarity for code and comment comparison, clustering to identify behavioral patterns, and classification models to automate feedback generation based on rubric-aligned criteria. Student reflections on their use of AI tools are also analyzed using topic modeling and sentiment analysis, which reveals attitudes and usage trends over time.

All implementations utilize Python and open-source frameworks. The paper proposes a practical, scalable methodology for educators aiming to incorporate AI-aware practices into programming education. The framework includes recommendations for assignment design, code assessment workflows, and the integration of NLP tools in formative and summative evaluation.

By blending technical analysis with reflective pedagogy, this work contributes to a more responsive and ethical model for programming education. Rather than restricting AI use, this work emphasizes a mentorship and guided application approach, empowering students to engage with generative tools in meaningful and responsible ways.

Keywords: Data Science, Generative AI, Programming Education, Code Similarity Detection, Academic Integrity, Student Assessment

MAIN REFERENCES: 

1. Adeshola, I., & Adepoju, A. P. (2024). The opportunities and challenges of ChatGPT in education. Interactive Learning Environments, 32(10), 6159-6172.
2. Elbanna, S., & Armstrong, L. (2024). Exploring the integration of ChatGPT in education: adapting for the future. Management & Sustainability: An Arab Review, 3(1), 16-29.
3. Konak, A., & Clarke, C. J. S. F. (2023). Augmenting critical thinking skills in programming education through leveraging chat GPT: analysis of its opportunities and consequences.
4. Güner, H., & Er, E. (2025). AI in the classroom: Exploring students’ interaction with ChatGPT in programming learning. Education and Information Technologies, 1-27.
5. Fuchs, K. (2023, May). Exploring the opportunities and challenges of NLP models in higher education: is Chat GPT a blessing or a curse?. In Frontiers in education (Vol. 8, p. 1166682). Frontiers Media SA.