Over the past few years, artificial intelligence has made a real breakthrough in visual arts. Neural networks have learned to generate images that are difficult to distinguish from photographs, create animation based on text descriptions, and even write scripts. Naturally, the industry started talking about a revolution: will directors and screenwriters soon be left without work? We closely follow the development of technologies and use AI tools in our production processes. But our experience shows: neural networks are a powerful assistant, not a replacement for a creative team. Let us figure out why dramaturgy and directing remain exclusively human territory, and how exactly AI is changing the animation industry without canceling the main thing—living art.
Any good script relies on conflict, characters, and the emotional arc of the protagonist. A neural network can analyze a million scripts, identify patterns, and generate text that outwardly resembles a story. But inside that story, there is no soul. Why? Because AI does not know what longing, the joy of victory, or the fear of loss feel like. It operates with data, not experiences. When we work on an animated series for a brand or television channel, the most important thing is to create characters that the audience will believe in. Take, for example, the mascots for Sochi-Park: their personalities, actions, and dialogues—everything was born from an understanding of child psychology, family values, and emotions that are close to parents and children. A neural network could draw a bright picture, but it would never figure out why a character acts in a certain way rather than another, because it has no motivation.
A good screenwriter or director always keeps the audience profile in mind. A children's series requires one tone, a youth music video another, and a brand image video a third. This is a fine-tuning that requires empathy and real-life experience of communicating with people. A neural network can segment the audience by age, gender, and interests, but it is unable to truly understand what will make a three-year-old laughand what will bring a smile to their mother. These nuances are resolved at the dramaturgy stage, where the human eye and intuition are irreplaceable.
In animation, as in any art, the devil is in the details. A character's glance, a pause before a line, a barely noticeable hand movement—all of this creates emotional depth. The director builds these details consciously, based on the scene's objective. A neural network can generate facial expressions, but it will not understand why the character averted their eyes at that particular moment. Without that understanding, the scene remains flat, no matter how beautiful it looks. A practical example: in one of the episodes of "Orange Cow," which we worked on for Soyuzmultfilm, there was a scene where the main character takes offense at a friend. We spent several days finding the right vocal intonation and facial expression. A neural network could have offered a hundred variants of facial expressions, but only a human director sensed that in this case, the offense should be conveyed not through loud shouting but through a quiet sigh and lowered shoulders. And it was precisely this detail that made the scene feel alive.
| What a Neural Network Can Do | What Remains with the Director and Screenwriter |
|---|---|
| Generate hundreds of storyboard variants in a minute | Choose the right emotion in a specific context |
| Create texts based on a given structure | Fill the story with life experience and meaning |
| Analyze successful scripts and identify trends | Understand why the audience will cry or laugh |
Many imagine the work of a director as managing a camera and actors. In animation, of course, it looks different: the director works with storyboards, composition, character movement, and editing. But in reality, the director's main task is to control the viewer's attention. They decide where the viewer will look first, what they will feel, and what they will think about after the scene. This is an art built on knowledge of perception psychology, accumulated experience, and—importantly—intuition.
Take editing. Neural networks can already edit video using templates, but they do not feel the rhythm of a scene. In animation, rhythm is everything. A long shot can create tension, a quick change of plans can add dynamism, and a sudden pause can enhance the comedic effect. The director makes these decisions based on what they see, hear, and feel. A neural network may propose technically correct editing, but it will not understand why a specific pause is needed at that moment. Without it, the joke falls flat, and the audience does not laugh. In our work on the animated series "Drawing Fairy Tales" for TV-3, we literally built every episode second by second to hold a child's attention. Children's audiences are especially sensitive to rhythm: if a scene drags on, interest is lost. A neural network could have shortened the video algorithmically, but only a living director found the right balance between dynamics and emotional pauses. This directly influenced the result: the series became popular and ran for several seasons.
Any art involves risk. A director may make an unconventional decision that either hits or fails. They take responsibility for that choice. A neural network does not know how to take risks. It operates according to probabilistic models: what has worked more often is what it suggests. But in art, the strongest decisions often lie outside templates. Recall the anime style we used for the advertising video for Pyaterochka. It was a bold move for a federal-level brand. A neural network would hardly have suggested such a visual language because, statistically, it was not typical for grocery chain advertising. But the director saw that it would work, and the project captured the audience's attention. Of course, we actively use AI tools in our work. They greatly help at the pre-production stage: generating references, creating quick storyboards, testing color palettes. But the final decision always rests with the human. Artificial intelligence reduces time spent on routine but does not eliminate creative exploration. In fact, this is the main value of the human approach in animation: we are not afraid to experiment and do things that do not fit into algorithms.
When we create an animated series, we do not just come up with a plot—we build an entire world. This world must have its own rules, logic, and atmosphere. Characters live in it, interact with it, and it influences their actions. Developing such a world from scratchis an extremely complex creative task that a neural network cannot solve.
Take, for example, the animated series for Dixie—"Cracked-2." The task was to create a universe connected to the brand yet engaging for the viewer. We thought through the characters' personalities, their habits, the jokes they would make, and the visual references that adults would understand. All of this is a layered cake of cultural codes, humor, and life observations. A neural network could generate characters, but it would not come up with why one of the heroes always nervously fiddles with their tail when lying, or why another has such a specific voice. Yet it is precisely these details that make a world unique and memorable.
Even if a neural network writes a sequence of events, it cannot build a multi-layered plot where every line serves the overarching idea. In good dramaturgy, everything is interconnected: characters' actions have consequences, jokes pay off several episodes later, and emotional arcs resolve by the finale. This is an architecture that only a human can create, understanding where they want to lead the audience. Our experience shows that clients increasingly want not just an animated video but a story with a continuation. A brand animated series is not a one-off campaign but an asset that works for years. Here, it is especially important to lay a dramaturgical foundation that will withstand several seasons. A neural network cannot handle this—it lacks strategic thinking; it sees only the current segment of text. One of the most difficult stages in creating an animated series is developing a visual language that serves the story. The director and artists search for this language experimentally, trying dozens of options until they find the one that perfectly conveys the mood. A neural network can suggest style variants, but it will not understand why this world should have exactly these shapes and colors. And understanding comes only from the dramaturgy, from the task the author sets for themselves.
We do not deny the benefits of artificial intelligence. In fact, we use AI ourselves at stages where speed and variability are important. It is like a calculator: it does not replace a mathematician but speeds up calculations. In animation, neural networks work great for generating backgrounds, finding color solutions, creating draft storyboards, and refining secondary elements. This saves the team time, allowing them to focus on the main thing—the story and characters.
If your project requires not just a pretty picture but a real story that moves the audience, neural networks will not help. You will need to hire people. And that is good because art is about human experience, emotions, and communication between the author and the viewer. No code can replace a living soul.
Artificial intelligence will not put directors and screenwriters out of work, just as photography once did not replace painting, and computer animation did not replace hand-drawn animation. On the contrary, neural networks open up new opportunities for creativity. They take on routine tasks, allowing artists to focus on what matters most. But the final choice, artistic intuition, the ability to take risks and feel the audience remain with the human. A script that makes the audience empathize, directing that controls emotions, a universe you want to return to—all of this is born only in the human mind. Neural networks accelerate the process but do not create meaning. And meaning is what audiences come for and what clients pay for. We see the future in a symbiosis of human and AI. When a neural network helps visualize an idea faster, and a director fills it with life. When AI offers options, and the author chooses the best one based on their taste and experience. This approach already works in our studio, and it makes animation more accessible while preserving the most important thing—living art. If you are planning an animation project and are unsure whether you need professional scriptwriting and directing or can get by with neural networks, come for a consultation. We will honestly tell you at which stage AI will help and where it is impossible without a human. And most importantly—we will show you with real examples how a living story turns into a successful project.
Send request
