ARTificial: at the dawn of artificial intelligence
Over the past 10 years, a tremendous breakthrough has occurred in the development of artificial intelligence. The human-created algorithm has gone from simple pattern recognition to victories in a wide variety of games. However, one of the most emotional and expressive areas of human activity - art - is still not subject to him. Or not? This is what we proposed to solve for the guests of the closed exhibition, which was located at the Museum of Russian Impressionism for one day on May 29.
AI Microsoft painted a picture on the composition “At the dawn” of the Alliance group
Microsoft presented a picture written by artificial intelligence based on the song of the Alliance group, "At Dawn." The algorithm is based on the principles of the synergy of music and painting, described by the great abstractionist Vasily Kandinsky. The idea of the project was to create a work of art to music that the artist had never heard. The composition “At Dawn” was chosen because it combines tradition and modernity - it does not lose its relevance. READ MORE
Microsoft's generative neural network analyzed the painting of the great abstractionist Vasily Kandinsky and the musical compositions with which he was inspired. This allowed artificial intelligence to identify patterns in the artist's works. Kandinsky advocated the synthesis of arts: music and painting. He created abstract canvases under the influence of the works of Wagner, Scriabin, Schoenbereg and other composers. We asked the neural network to create a completely new work of art, synthesizing the manner of writing by Vasily Kandinsky and the song of the Alliance group “At the Dawn”. See the result.
Benefit of one intellect
Artificial intelligence can create a variety of art objects: paintings, musical works and texts.
The music . To create music, the neural network tries to reproduce from a variety of sounds a certain harmony inherent in the musical works on which it studied. For example, a Christmas song from AI . At our exhibition, the guests were presented with the result of a creative collaboration between human and artificial intelligence: sound producer Nikita Stashevsky and the Neuron Soundware algorithm, which created music in the style of the great Czech composer Antonin Dvořák.
You can listen to the song here .
Painting . Using machine learning algorithms, artificial intelligence learns to find and recognize objects in a picture - faces, animals, various objects. Then, using his knowledge, he independently “draws” paintings based on existing images, recreating the artistic styles of such famous artists as Vincent Van Gogh, Pablo Picasso, Edward Munch and others. Neural networks even create their own paintings, which are sold at auction for hundreds of thousands dollars. One of the most famous projects in this area, which we already wrote about earlier, is “The Next Rembrandt” .
Literature . Once our chat bot wrote poems in Chinese ., But we decided to do something close to Russian readers. Using for the preparation of thousands of examples of fiction, including prose A.P. Chekhov and Russian folklore, artificial intelligence in the cloud of Microsoft Azure wrote stories and fairy tales. A rare opportunity to see how the process of creating text with artificial intelligence occurs.
At the end of the exhibition, visitors were invited to think and choose from the suggested examples which texts, paintings and music a person created and which ones were artificial intelligence. In some cases, identifying the author was simply impossible.
Who is the true creator
The final chord of the event was a discussion about whether AI can be considered the creator of an art object, or whether creativity is subject only to man. The speakers are:
- Alexandra Chetverikova - cultural scientist, journalist, radio host, teacher at Moscow State Pedagogical University, author of the podcast “Art for Boys” on Glagolev.FM;
- Vladislav Shershulsky - Microsoft Advanced Technology Director for Russia.
- Vyacheslav Lebedev - neurophysiologist, employee and graduate student of the Department of Psychophysiology, Moscow State University M.V. Lomonosova, co-founder of the NeuroFuture project.