AI Artwork: Kanabstracd “Delaunay-esque Surrealism without Circles”

An original computer artwork

Image ID: FBhi2Ntzp

Reduced image (600×600)

Poster in frame

Magnified image of the center

The high-precision poster of this image can be purchased at a page in Rakuten or a page in Yahoo! Shopping by inputting the image ID. A medium-precision version (about 2000 px) of this image is also included in Volume 1 of the AI Art Collection (paper book).

Explanation by GPT5/4o (Please note that the information may not be entirely accurate.)

This painting depicts a European-style alleyway that narrows as it recedes into the distance, yet the ground warps into something like a black-and-white chessboard, and the buildings appear not as straight lines but as if they are twisted or bent out of shape.

Although it seems to be constructed with straightforward linear perspective, the forms are broken and distorted here and there, so that the viewer has the uncanny sensation of having wandered into a dreamlike city.

At the center of the composition stands a pillar-like form made up of overlapping three-dimensional shapes in a variety of colors—red, blue, green, and more—which calls to mind the Cubist method of breaking objects down into their constituent forms and reassembling them.

The overall angular structure of the work and its powerful chromatic contrasts suggest the influence of Picasso and Braque. At the same time, the elongated shadows and anxiously warped streetscape evoke the atmosphere of de Chirico’s metaphysical cityscapes. One might say that, while taking a real city as its subject, the painting renders a fantastical world in which time and space themselves seem twisted out of joint.

How to create images?

This image was created by ‘Kanabstracd,’ an art unit formed by the creator and the AI (Stable Diffusion XL 1.0 (1024), a program for creating AI artworks). I gave the program the word ‘Kanabstracd surrealism Delaunay’ as a prompt and let it create. Also “Delaunay circles ” was used as the negative prompt.

In AI art, the images generated are usually quite rough—about the size of an A6 print (roughly 500–1000 pixels on each side). To make them suitable for presentation, I use a technique called super-resolution, which increases the resolution 4–8 times higher than the original, adding finer detail. However, please note that the resolution of the product image shown above is about the same as the original and has not been upscaled.

These images are created by our original method, which is different from usual AI art images. AI art images are created by giving English description called prompt. Many AI art images were created by using AI art programs such as Stable Diffusion，DALL・E 2，or Midjourney in 2022. To create expected images by AI art, a detailed description is usually given. However, we usually give only a few words or a coined word to web sites of OpenArt, NightCafe, etc., which use Stable Diffusion XL 1.0 (1024); That is, the program create images generatively and emergently using the knowledge of the program freely. The contrast of created images are adjusted, but they were not modified.

Unintentional images created in this way are also called "AI art", but I don't think they are necessarily art. However, there are unexpected images or images with a style different from that of the work, patterns, or Various interesting images such as cats embedded in stained glass can be created. Images created by this method, i.e., unintended iamges, are to be called "AI art",

Keywords

AI art, AI artist, Image generator AI, Artificial intelligence art, Computer art, Digital art, Media art, Stable Diffusion, Generative art, Emergent, Art generator, Intelior, Wall art, Art poster, Art print, Original poster

Note on copyright

This image was generated by AI using the ISAP method. It is not intended to imitate or reproduce any specific living artist or existing artwork. If you notice anything of concern, please contact us at info@dasyn.com .

Dasyn.com デイシン
2025-11-28 16:37:04.437547