READ ME
Neural network art as a cultural revolution — a visual study examining AI as a new tool in art.
ABOUT THE PROJECT
Analyze the capabilities of AI (using Midjourney as an example) in relation to working with art: how AI interprets painting, emotions, and drawing techniques. To what extent can a person distinguish generative art from real art, and how can it be used.
TASK
VISUAL RESEARCH
formulation
Topic
I defined the goal of the visual study—an exploration of the capabilities of AI-generated content. I identified the key questions and formulated the hypothesis, where the central question concerned the capabilities of neural networks in the field of art and whether they can surpass real artists.
(01)
The study was conducted in 2022, when AI was not yet as advanced. This is no longer just research, but a full retrospective
RESEarch
I collected up-to-date data on AI, selected a specific tool for the experiment, and outlined how I would test it. I reviewed numerous scientific studies on generative art, both domestic and international. I also studied how AI works and its specifics.
(02)
CHAOS
controlled
MODULAR
GRID
GRID
AI ALGORITHM
LAYOUT CONCEPT:
Structure and master layouts
Elements extend beyond the margins, encroach into the gutter and pagination, creating controlled chaos—a metaphor for AI that operates according to an algorithm but produces a variable result each time.
The book layout is based on digital logic: the modular grid resembles the pixel structure of an image. Illustrations in a 1:1 format follow the system and reinforce the sense of order.
Based on the sections, I created master layouts for each of them; in addition to the sections, I also developed the structure for the chapters. I defined the hierarchy and layout, created a suitable flexible grid, and added pagination and micro-typography.
Based on the written texts, I created the structure of the book: chapters, sections, headings, sequence, and content.
(03)
Layout of blocks
Each chapter consists of three blocks, each with its own layout format.
(04)
spread
the original
3. Guess
In the final chapter, I conducted a visual analysis and provided an honest evaluation of AI works based on technical criteria, on par with works by real artists.
An interactive block where the reader can test themselves: how accurately they can identify the original artwork among its digital copies.
The answer can be revealed by turning the page, where, similar to an artist’s signature on the back of a canvas, the copies include digital tags, while the original is marked with the artist’s authorship.
The answer can be revealed by turning the page, where, similar to an artist’s signature on the back of a canvas, the copies include digital tags, while the original is marked with the artist’s authorship.
2. Demonstration
of capabilities
An analysis of how AI interprets emotions, situations, and artistic styles. How well it understands how verbal and abstract concepts can be visualized through art.
The layout of images here is chaotic; in addition to them, there are tags used to generate the image.
The layout of images here is chaotic; in addition to them, there are tags used to generate the image.
- Opening
Consists of two images: the first is the original artwork, the second is a digital copy with tags.
I tested how well AI can replicate a work or, alternatively, interpret it in its own style.
I tested how well AI can replicate a work or, alternatively, interpret it in its own style.
SPREADS
Lenticular printing
Proofing and prepress preparation
The book cover was created using lenticular printing. I combined the original artwork with its neural network-generated copy. As the viewing angle changes, the original gradually distorts and transitions into its AI “double,” emphasizing the ambiguity and provocative nature of the chosen topic.
I conducted a final check of readability and the logic of content presentation. After that, I prepared the file for print, taking into account technical requirements such as formats, bleeds, color profiles, and image quality.
(05)
