Infinite Everything: Generative Art


Computer aided art and how math can help to think outside the box


‘Hello my name is Bronko’ is originally an artwork I created end of 2001. It was made in old fashioned screen printing technique. It’s not a special artwork or a masterpiece – it’s just a small piece of printed paper I see every day, because it’s hanging in the so called „chapel of rest” in my mansion.

Fig. 1: Draft of the original silkscreen artwork, 2001. The positioning of the elements was made by hand.

For an artist, especially for artists who work with the collage technique, you run into the situation that people ask you, why you placed the images (photos, words, etc.) of an artwork the way you did. The decision is mostly a mixture of mood and the rule of placing images so that they are not cropped, can easily be read or are in harmony with the rest of the artwork.

The problem with „in harmony” is, that it is mostly based on rules. Rules that we create on our own or are made by society. The rule of placing images inside a canvas and not half cropped outside is based on logic – regarding ‘Hello my name is Bronko’, it wouldn’t make much sense to place the name ‘Bronko’ partly outside the paper. Why use the name when it’s partially hidden?

I learned that the best artworks are created by accident. Elements copied somewhere on a page when working on a computer, snippets of paper covered with glue falling on a canvas – examples of accidents that made an artwork better, because intentionally I would never placed them there.

The main essence is the element of surprise and the absence of rules. By placing ‘Bronko’ half outside the page or placing it over other elements you get a whole new world of visuals.

What if the computer decides where to place your images – randomly?


I started to code a tool what I called Infinite Everything v1 (don’t Google it) as an assistance for creating some of my artworks. The tool was initially written in Processing 3, but later ported to the terrible javascript engine of Photoshop. The main task of the tool is to create semi-random placement (variations) of elements on a canvas. It could help designer, too — but was generally coded for creating collage artwork – to assist me.

It was not intended to replace the creative process behind creativity but it helps the artist to avoid the traps of considering too much or following rules during the creative process. It places given elements without any man-made rules. In fact, there are rules, but these are quite simple.

Let us see, what the tool can do for us:

Fig. 2: The tool in action. Nine different, randomly generated artworks

What can be created?
What the tool does is, that it takes existing collage elements (images, words, text snippets, etc.) that are pre-selected by the artist and places them randomly across the canvas. The magic word here is ‘pre-selected’. As an artist, I still have a message. Placing images randomly is not the same as placing random images. Mostly I know, WHAT I want to display, but am curious HOW it could be composed.

Random placement is easy and it’s not a big thing and most of the results wouldn’t be interesting. Here comes the magic, it takes a human brain. The tool allows me to make images dependent on other images. Let’s call this ‘magnetism’.

The simplest rule is:  Image A is magnetic to image B.

An example would be: a caption (image A) belongs to a photo (image B). That is the artistic process. The artist need to decide which images belong together. The code for this rule is something like this:

1) Place image A somewhere on the canvas
2) Place image B not too far away from image A

This is a very simple rule but it creates some really interesting results.

The tools offers one configuration, thats rule number two:

how far can an image be placed outside the canvas?

You get interesting results here, but it’s obvious, the more the image is outside (cropped) or inside (not cropped) it gets more boring.

After all, it’s no magic here, no AI or genius coding, just simple math with simple rules. The tool doesn’t help you colorize or selecting images or to compose an artwork.
I call this computer aided art and not computer generated art because you still have the process of selecting images, setting colors, size of the images, fine tuning and at all, bring the results on canvas.

The machine helps you to think outside the box.

P.S. I have no plan to release the tool.

Fig. 3: Selection of six variations of the foreground generated by the algorithm. In fact, the tool created around 20 variations. There is of course no limit of variations. We could easily generate 100 images, if there is any need for that. The word ‘INFINITE’ was centered and fixed on the canvas. No rules apply here.
Fig. 4: Variations of the background. This time, all elements were placed randomly (without any logic or algorithm) by the computer.
Fig. 5: The final composite. Foreground and background were selected and stacked (by the artist). Some foreground elements were moved for aesthetical reason, I’m still the artist, right?
Fig. 6: Final results: two generated images where selected to be placed on canvas. This process was handmade (analog), 70 x 50 cm, glue, spray paint and paper on canvas.