I have been messing around with some AI recently and came up with an idea: use my original photos of stuff and let an AI engine grind out its version of my images. I made a few attempts, and I liked the results.
This is the description of the original image:
The image shows a bicycle leaning against a beige-tiled building. The bicycle has a black frame with a front basket and an orange seat. The bike appears to be well-maintained. The building’s wall is plain, with visible seams between the tiles, and some faint graffiti can be seen above the bicycle.
Next to the building, on the sidewalk, there is a red fire hydrant with some graffiti on it. The sidewalk is made of a combination of concrete and cobblestones. In the background, there is a street with a few visible cars and buildings, suggesting an urban setting. The overall atmosphere of the image is calm, with a focus on the bicycle and the street scene.
So I took that image and sent it through the AI engine and this is what it produced.
And for fun, I made it in Manga style. I think it is interesting that it flipped the building, is it because Japanese is written Right to Left, not of Left to Right
