At the "Forum 2050" stage during Startup Village in Skolkovo, Professor Evgeny Burnaev, Director of the Skoltech AI and Applied AI Centers, delivered a lecture on how generative artificial intelligence is transforming approaches to creativity, offering new tools for self-expression and opening new horizons for artists.
In his lecture, Evgeny Burnaev explored the evolution of using mechanical devices and automated systems for creative purposes. He began with kaleidoscopes and simple algorithms from the 1960s that created patterns based on randomness and moved on to modern generative neural networks capable of creating images indistinguishable from the works of classical artists. These historical examples help illustrate how far generation technologies have advanced over the past decades. A significant part of the lecture focused on the progress of deep neural networks since 2014, when AI's capabilities in generating realistic images greatly improved due to the development of more complex models and the availability of large datasets. Today, neural networks can not only replicate given images but also create new, unusual shapes, styles, and compositions. Burnaev noted that these achievements have found widespread applications in various fields—from digital art and design to marketing and entertainment media. However, according to Professor Burnaev, generative neural networks still have limitations. They can produce images with errors or distortions, such as cars with three wheels or people with an incorrect number of fingers.
Despite these limitations, the technologies continue to evolve and are already actively used in creative and industrial fields, from graphic editors to the creation of cartoon characters. "Is it possible to draw not just a horse, but a horse that is galloping or of a slightly different type? This question became relevant around 2014, during another wave of AI development. That was when methods emerged that allowed for style imitation and the creation of entirely new images, somewhat reminiscent of great paintings by classic artists," noted Professor Burnaev.