Link to Part II (and there you can find a link to Part I)
This is Part III of my explorations down the Optic Nerve.
Previously, I attempted to derive from AI tools, an image that would directly stimulate the most basic levels of the V1 visual cortex. This area receives impulses from the lateral geniculate nucleus (LGN).
The LGN is particularly interesting as it is basically an endpoint for each of the neurons from the retina. So, a tiny light hose from your iris (back wall of the eye that reality is projected upon) that connects to this place in the brain, the LGN, that I imagine like a dragonfly’s eye, then sends impulses to the V1, which is where the first layers of recognition occur.
In the below image, we can imagine recognition centers in the V1 firing off if certain areas of the iris are stimulated, and the various representations show an arrangement of stimulated neurons that creates recognition. I tried to ask MidJourney to project outwardly what the V1 would be triggered by, but it’s still a *very difficult prompt question.
Since there is an exact correspondence of the nerves in the iris to the LGN, then varying the size and orientation of image components will have a varying effect on the V1 and V2 regions.
Using shape, color, and size in the display or on a canvas, I can stimulate certain neurons in the iris for a direct effect on GN cells and thereby ‘cascading’ effects on the V1.
If I want to stimulate all of the neurons in this basic layer, I need to provide all of the ‘major’ building blocks of stimuli, and colors and orientation etc. to give a ‘well rounded’ viewing experience.
The V1 has cells that detect specific aspects of the visual field. These cells are cross-linked and act with specific groups of other cells to shape impulses that are then translated later in the visual system. Next after V1, is the V2. For now, I’m focused on the most basic elements possible.
So, firing up MidJourney, I set about creating “an artwork made by an algorithm that pleasantly mixes abstract forms with the following characteristics: 1. Orientation — horizontal, vertical, or tilted lines. 2. Spatial frequency — low, medium, and high numbers of spatial repetition. 3. Contrast — gradient contrast and varied difference in luminance between adjacent forms.”
My intention for adding “by an algorithm” is because I intend to create an on-chain NFT that uses parameters to create these forms for the viewer. Out of curiosity I wondered if MidJourney might provide some connection to the real world of algorithms, and Code Pilot might provide the other side.
So, the algorithmic aspect is represented, as are the requested forms, in aggregate. No single image impresses me, but I do like the ‘drippy lines’ in №4 and I can easily see an algorithm producing that one. However, the iris is circular and the brain is infinite, so I can’t see stripes alone working well even though there is opportunity in contrast and speed of the ‘drips’ to create horizontal forms.
Another variation on this prompt provided something that ‘felt’ better to look at, and I could see how parameters could be used to good effect.
Then I wondered about removing the “algorithm” and giving MJ the benefit of the doubt by including language about the systems involved for prompt 2:
“an artwork designed to stimulate the limbic system and LGN with v1 that pleasantly mixes abstract forms with the following characteristics: 1. Orientation — horizontal, vertical, or tilted lines. 2. Spatial frequency — low, medium, and high numbers of spatial repetition. 3. Contrast — gradient contrast and varied difference in luminance between adjacent forms.”
I’m liking these. Of which one is growing on me ‘strangely’, reminiscent of the unsigned_algorithms effect. It is a like/dislike cycle that feels just a little too much activity, but getting closer for sure.
and a toned down version
Then another variation on prompt 2 provides this, which begins to get me ‘there’.
My favorite for pleasantness and likely ability to convert to code:
and one that feels like it has effects that I could tune better with some line width variation/patterns
and this one popped up as an anti-pattern that I might learn a negative lesson from. Notice how easily the ‘jarring’ lines affect the mood, for me it produces ‘alertness’:
and then finally, I notice a very distinct difference between similarly sized objects and varying sized objects.
The inclusion of dots, clearly tickles another group of receptors:
Circular patterns are soothing, but don’t have the ‘cleansing effect’ of including other patterns the V1 is triggered by, like horizontal and vertical and angled lines.
prompt 3
“an artwork that pleasantly mixes abstract forms with the following characteristics: 1. Orientation — horizontal, vertical, or tilted lines. 2. Spatial frequency — low, medium, and high numbers of spatial repetition. 3. Contrast — small, medium, and large scales of gradient contrast and varied difference in luminance between adjacent forms. This image fills the bounds without large areas of black, or empty space. A pleasing color palette is encouraged. The purpose is to stimulate the most basic connections of LGN to V1”
Then I went for it in prompt 4
“image representations as understood by the V1 portion of the human visual cortex that most basically stimulate all the areas and patterns”
and this beauty showed up:
So then I /blend a few images together and see what happens:
I find something that energizes and calms me:
Next, I’m interested in gathering data. Environmental factors from ocean buoys offer more than enough variation to make wonderful visualizations and induce a sense of ‘ownership’ in the viewer.
and I’ll leave with two derivatives that have obvious origins shown above
and their /blend