Reading news has become a dangerous and energy-consuming way to spend time lately, it raises anxiety, causes panic. Looking for reliable sources of information in the post-truth era becomes more and more difficult. This project brings up the issue of trusting fake news and surrealism of the current events, when mem is undistinctive from a piece of news and sarcasm - from reality. All images are generated by neural network on the basis of descriptions with the ruDALL-E Malevich model by Sber. The descriptions themselves were generated by neural network - Russian model GPT-3 that had been learned by dataset of Meduza, Lenta, Breakingmad and Radio Adonezh. By these means I project machine view onto a flat surface to show an abstract image of modern Russia, where the issue of corruption is interpreted in a distinctive way. These images with titles are just a material. The project can live on as a news telegram-channel or indefinite news feed with pictures and inscriptions.