Before, after, and AI: how neural networks are changing creative education in Russia
Neural networks today help create music, video, and animation, optimize processes in the preparation of theatrical productions, and in advertising projects, an AI creator can perform tasks that previously required an entire team. Creative universities and schools are already taking steps towards technologies that have become a modern reality. In Gnesinka, the neural network learns the history and theory of music, and in the future AI can be introduced into the educational process. At the HSE School of Design, working with AI is no longer called an experiment, but a part of the infrastructure. How smart technologies are used in the training of representatives of creative professions, where they are looking for neurocreators and what risks experts see in the thoughtless use of artificial intelligence — in the Izvestia material.
And Art
The use of neural networks in art continues to cause controversy about ethics, authorship, and the boundaries of creativity. AI participates in the creation of music, generates videos that are sometimes difficult to distinguish from the real ones, and in some projects takes over part of the work of entire production teams. This reality is gradually changing creative education: educational institutions — from specialized universities to specialized schools — are beginning to learn new tools.
The Gnessin Russian Academy of Music is already participating in a project to teach a neural network the basics of music history and theory. The academy does not rule out that in the future AI will be used to test students' knowledge, subject to the approval of the academic council and relevant departments.
"I can assume that this will happen in the next two or three years, if our academic council supports the initiative and we manage to adapt the technology for teaching music history and theory," university rector Alexander Ryzhinsky told Izvestia.
At the same time, he does not yet expect a significant impact of AI on the educational process in the musical art due to the peculiarities of specialties related to academic fields.
— If the introduction of artificial intelligence technologies is possible, it will be within the framework of the conditionally technologized specialties that exist today. These are media companies, musical sound engineering and similar programs," Ryzhinsky said.
The rector added that the training of instrumentalists and vocalists is still based on traditional methods. Performing skills require direct work with a teacher, and neural networks will not be able to replace it in this process.
The possibilities of using AI in creative education largely depend on the specifics of the profession.
— There are professions in which AI is most likely to be used, for example, composers, animators, set designers, film directors, lighting designers. But vocalists, dramatic artists, dancers, and performing musicians cannot use artificial intelligence to form professional competencies," Tamara Potapenko, director of the Center for Continuing Education and Advanced Training of Creative and Managerial Personnel in the Field of Culture at GITIS, said in her turn.
"It's pointless to argue with technology — you need to learn how to work with it"
At the HSE School of Design, AI is no longer seen as an experiment, but as part of the infrastructure.
— Neural networks are woven into students' project work. We proceed from the fact that it is pointless to argue with technology — we need to learn how to work with it. We have created our own digital ecosystem where composers and sound designers post audio files, and students post their projects. Teachers can discuss work with students directly in the browser, without separate correspondence and forwarding of materials," Peter Skovorodnikov, academic director of the Game and Digital Product Development program, coordinator of the sound music department at the HSE School of Design, told Izvestia.
According to him, AI removes routine tasks, but not responsibility for the result. This position is shared by other teachers of the school.
— The main risk for education is to get an excellent student's work, the author of which does not understand how it works (and the neural network can write a convincing explanation in hindsight). Therefore, one ready—made project along with a process report is not enough. During the analysis, I always ask why one or the other option was chosen. In such a conversation, it usually becomes clear whether the author is able to explain his choice," said Oleg Pashchenko, a lecturer in Digital Product and Artificial Intelligence at the HSE School of Design.
Another area in which AI is actively developing is the production of visual content. New areas of training related to neural networks and their use in animation, film, and advertising are emerging in educational institutions.
—The field of neuroanimation is new for our university and was finally formed only this year," Olga Orlovskaya, Acting Head of the Department of Animation and Comics at Synergy University, told Izvestia. — Its implementation is related to the relevance of the topic and the growing need of the industry for relevant specialists. I use artificial intelligence tools in my work, primarily to search for visual ideas and images, as well as to design text materials. However, I apply them pointwise and always double-check the results.
There is also a demand in the labor market for specialists who can work with neural networks. So, on the website of the Soyuzmultfilm film studio, offers for an AI artist, an AI developer and an AI creator are posted in the vacancies section. All three positions involve project employment.
Representatives of the advertising business also talk about the shortage of staff in neuroproduction.
— The industry is increasingly using generative content. We work with large clients, including European ones: we implement projects together with agencies from Germany and Spain. At the same time, there are not enough specialists on the market. That is why it is so important to develop educational programs that will train new personnel. We call them neurospecialists or AI creators - these are people who are able to independently manage the entire production process. Previously, a team of 50 employees was required to create a video, but now one person can handle this task with the help of AI. The savings are enormous. Previously, the cost of my projects was 10-30 million rubles, today similar work can be done for 1-3 million," said Eraj Nidoev, creative director of the AI division of Filmway Production.
How not to lose professional skills
The proliferation of neural networks poses a new question for creative education: how can students be taught to use technology without replacing their own thinking and professional training?
— If AI takes over most of the tasks that a specialist used to perform, it is unclear how a person will gain practical experience and professional skills. A cinematographer is formed through hundreds of hours of watching and analyzing films, editing, working on director's explications, practice and his own mistakes. If you remove a significant part of the hours of direct work with the material, a person will get a result, but not necessarily gain the experience that was previously necessary to achieve it," said Vladimir Kovalchuk, director and lecturer of the Neural Networks and Cinema course at the Mikhail Chekhov Center.
One of the reactions to the new reality was the Ethical Charter for the Use of Artificial Intelligence Technologies in Theatrical Art. Its developer and initiator was GITIS, and the basic principle of the document is "Augmentation, not replacement." Technology should remain a tool in the hands of the artist, not a substitute for it.
— Tools are changing rapidly, and knowledge about one program may become obsolete in a few months. But the basics of composition, drama, visual storytelling, editing, and artistic thinking remain relevant. Today, almost anyone can generate a beautiful image or video. But a good shot by itself does not make a person a director or designer. You need to understand what task this frame solves, what story it tells, why this particular composition, color or style was chosen," agrees Elena Kozlova, designer and founder of the Hey Eye Lab school.
Another aspect that experts pay attention to is that technology is imperfect and sometimes tends to fantasize.
— When I worked with Onegin, the neural network made up such stories that it was even interesting. She quoted the lines from a female person and did not correlate that this was Eugene's letter — that is, she was hallucinating. In addition, the program itself is most often written to satisfy your request. He's not a real conversationalist: he can't fend you off. Therefore, on the one hand, AI is really a great tool. But on the other hand, I am deeply convinced, as my master Oleg Lvovich says, that the text should open. We need to work with the book, because in the profession of an artist we talk a lot about perception, and in the profession of a director we need to talk about exactly the same thing," said Natalia Shurganova, senior lecturer at the Drama Directing Department of GITIS.
As a result, representatives of educational institutions and the creative industries agree: AI can become an assistant, but natural intelligence also needs to be upgraded.
Переведено сервисом «Яндекс Переводчик»
