GigaChat Passes the Final Examination in Musicology at the Gnesin Academy
Sber’s neural network, GigaChat, has successfully passed the final examination in the «Musicology» specialty at the Gnesin Russian Academy of Music. This test marked the next stage of the scientific and educational collaborative project by Sber, the Gnesin Academy, and the music service «Zvuk.»
The exam was conducted in the format of a colloquium. The questions were designed to closely match the level of a graduating musicology student. The examination board included Candidates of Art History, the Vice-Rector for Academic Affairs, department heads, associate professors, and lecturers from the departments of the Historical-Theoretical-Composition Faculty of the Academy.
With GigaChat, they discussed musical movements in Russian and foreign music, analyzed orchestral styles, and delved into 20th-century jazz. A new phase of the project involved training the neural network in practical musical disciplines — music theory and harmony. As a result, during the exam, GigaChat answered a complex set of questions in the following disciplines:
• Harmony studies, Music Theory, Theory of Modern Composition;
• Instrumentation, Fundamentals of Sound Engineering, Electronic and Computer Music;
• Music Journalism and Criticism;
• Analysis of Opera and Ballet Performances;
• History of Sound Experimentation and Research;
• Music Psychology and Education.
The examination committee noted GigaChat’s ability to work not only with factual material but also with deep analytical tasks requiring the interpretation of musical works, the comparison of artistic movements, and an understanding of the evolution of musical forms.
The final exam served as a presentation of the applied potential of GigaChat in educational and scientific work. Educators noted that the technology can be used as an assistant capable of searching for necessary sources, analyzing musical works of different styles, creating methodological materials, conducting research on various topics, and developing content for projects. The project demonstrated that GigaChat can integrate into an academic environment without requests being simplified and can be a worthy assistant.
Artem Lebedev, General Director of the Zvuk music service: «It’s important for us to make education more accessible and technology-driven for everyone interested in music, culture, and art. Our scientific and educational project with the Gnesin Academy and GigaChat has shown that generative AI can already help students prepare for exams. For educators, it will become a convenient tool for working with materials and archives. We plan to continue our cooperation and train the model in music.»
Aleksandr Ryzhinskii, Rector of the Gnesin Russian Academy of Music: «No matter how we view the prospects for the increasing role of artificial intelligence in the world, we understand that it has already become a part of our lives. This kind of project allows us not only to understand what benefits we can derive from the development of AI but also to make it useful in our professional activities. I believe this project is undoubtedly a frontline for a creative educational environment.»
Photo: Anna Bobrik