AI IN VISUAL ARTS
Artificial intelligence is at the center of attention and discussions due to its capabilities and future role in society. I want to share my experience in the field of visual arts, discussing the pros and cons of AI in this area, as well as its future prospects for artists and society.
There are several types of art generators, some of which create images based on input text, while others also use images uploaded by the user in addition to the text. Personally, I prefer the latter as they allow for a more precise result.
I should note that I am not interested in such art movements as photorealism, where the art generator Midjorney has shown good achievements lately. While I accept this style, I prefer in my work those styles where more imagination, fantasy, and transformation of reality can be seen.
At the request of my friend, I made Tarot cards for several months using the art generator Wombo. I chose this art generator because it allowed me to upload my own images for use. Wombo has about twenty different styles for users to choose from when creating images. In addition to the recognized master styles, the art generator includes programs for the styles of contemporary artists.
My conclusions from working with Wombo AI are as follows:
When the image is made solely based on the text, the color scheme and color gamma are not harmonious.
If an image is included, the art generator uses its color scheme.
The art generator often ignores the content of the text if there is no original image from the user.
In many cases, the desired image can only be achieved if the user provides an image.
The art generator is capable of adding objects, people, birds, animals, and plants even if there is no request in the text or original image.
The art generator is capable of creating scenes and adding content on its own.
It applies small details, texture, ornaments, and clothing coloring quite diversely and colorfully. The clothing of the characters often impresses with fantastic styles.
Among the shortcomings are frequent failures in depicting human hands and faces, which can turn out to be ugly or with an arbitrary number of fingers.
The compositions are often well-constructed, but they are standard.
And now about the prospects of artificial intelligence in the field of visual arts. AI can assist in book and magazine illustrations, fashion and interior design. Perhaps it will encroach on artists who create works by imitating recognized artists and do not have their own creative style.
But AI cannot yet create something radically new. It cannot compose in a way that is original and not similar to what came before.
In conclusion, I can say that I was approached by a team of Art Generator with an offer to provide them with my works for the creation of a program that would allow users to create in my style. If I set aside modesty, I can admit that I constantly receive feedback that my style is unique and significantly different from the styles of other artists. I thought about it and decided that it is better if users do not produce in huge quantities images in a style similar to mine.
