St. Petersburg, St. Petersburg, Russian Federation
This article examines the challenges of improving the quality of training highly qualified IT personnel at leading Russian universities. This issue is being addressed within the framework of the Russian Ministry of Digital Development, Communications, and Mass Media’s federal project, “Top-IT.” The study aims to identify the didactic potential of neural networks for the development of the domestic education system and to improve the quality of educational outcomes, specifically to enable the adoption of fundamentally new solutions for Russia’s technological leadership in a modern digital format. The results of an empirical experiment conducted at Omsk State Technical University, a participant in the “Top-IT” project, demonstrate the importance of the following didactic properties of neural networks: self-learning capabilities, interactivity, high performance, contextuality, adaptability, and the ability to regenerate data. Students and faculty members identified the most frequently used neural networks (AliceAI, DeepSeek, GigaChat, ChatGPT, Perplexity, MottorAI, etc.). The respondents’ high awareness of the characteristics and experience of using neural networks in university education demonstrates their mastery of a universal skill: preparing technical specifications for a neural network, which determines the relevance of a response to a user’s request. The study’s findings highlighted key advantages of neural networks’ educational capabilities: automation of routine tasks, increased student motivation, personalization and adaptability of learning, and development of critical thinking. This, in turn, opens new horizons for self-education, including for future top IT specialists.
monitoring, higher education, information technology, top specialists, digital competencies, artificial intelligence, neural networks, prompts for neural networks, online survey
1. Malyshev I.O. Obzor sovremennyh generativnyh neyrosetey: otechestvennaya i zarubezhnaya praktika // International Journal of Humanities and Natural Sciences. – 2024. – №.1-2 (88). – S. 168–171. DOI:https://doi.org/10.24412/2500-1000-2024-1-2-168-171 (data obrascheniya: 29.12.2025).
2. Skripkin I.N., Belyaev D.A., Selischev O.V. Yazykovye sistemy iskusstvennogo intellekta v obrazovatel'nom processe: kachestvo znaniy obuchayuschihsya v kontekste praktik neyrosetevogo masshtabirovaniya uchebnogo processa // Perspektivy nauki i obrazovaniya. – 2025. – № 5. – S. 601–617. DOI: https://doi.org/10.32744/pse.2025.5.39 (data obrascheniya: 29.12.2025).
3. Sysoev P.V. Didakticheskie svoystva i metodicheskie funkcii neyrosetey // Perspektivy nauki i obrazovaniya. – 2024. – № 6(72). – S. 672–690. DOI:https://doi.org/10.32744/pse.2024.6.42 (data obrascheniya: 29.12.2025).
4. Chungulova G.K., Orazalieva E.N. Vozmozhnosti i problemy bol'shih yazykovyh modeley v obrazovanii na primere ChatGPT // Nauka i real'nost'. 2024. – № 4 (20). – S. 85–91.
5. Shirokolobova A.G., Larionova Yu.S. Algoritm integracii neyrosetey v cifrovuyu sredu vuza dlya povysheniya kachestva obrazovatel'nogo processa // Pedagogika. Voprosy teorii i praktiki. – 2025. Tom 10. – Vypusk 6. – S. 783–791. DOI: https://doi. org/10.30853/ped20250093 (data obrascheniya: 29.12.2025).
6. A vy uzhe skachali? Docent SGU o novoy neyroseti DeepSeek. URL: https://vk.com/@saruniversitydeepseek (data obrascheniya: 30.12.2025).
7. Kitayskie universitety hotyat, chtoby studenty ispol'zovali II bol'she, a ne men'she. URL: https://habr.com/ru/news/931992/ (data obrascheniya: 30.12.2025).
8. Rossiyane vybrali lyubimyy II – i eto okazalsya ne DeepSeek, a otechestvennaya model'. URL: https://incrussia.ru/news/rossiyane-vybrali-lyubimyj-ii-i-etookazalsya-ne-deepseek-a-otechestvennaya-model/ (data obrascheniya: 30.12.2025).
