?
Формирование навыков работы с генеративными нейронными сетями у студентов творческих специальностей
Research continues the academic discussion regarding the implementation of deep machine learning technologies, commonly referred to as artificial intelligence, into the educational process, and their impact on established practices within the Russian higher education system. Over recent years, a substantial body of research has emerged, focusing on interaction with these algorithms in the context of scientific and applied tasks. However, the impact of artificial intelligence on educational programs for creative disciplines, which predominantly have a practice-oriented and project-based nature, is rarely addressed.
The article summarizes two years of project-based teaching experience with students from the Institute of Media and the School of Design at the National Research University Higher School of Economics (HSE University), aimed at developing their skills in interacting with graphic and language-based neural network models (Midjourney, Stable Diffusion, ChatGPT, etc.). Through the implementation of educational programs, the necessity emerged to equip students with specific prompt engineering competencies, as well as to structure the workflow with generative algorithms and foster interdisciplinary cooperation. Particular attention is paid to applying an object-oriented approach and the "Double Diamond" creative design model, which help rationally distribute roles between human and algorithm, enhance creative design effectiveness, and preserve the leading role of the human in decision-making processes.
Additionally, the article addresses methodological aspects of documenting and identifying schemes for interaction with neural networks that facilitate result reproducibility and teamwork skills development. Introducing these approaches into the educational process not only equips students to become competitive specialists in the rapidly evolving technological context but also contributes to democratizing creative labor by lowering barriers to creating complex media products.