Unit 2 – Future of the Mind – Grid cell context/research

My research into AI and artificial minds took me into the latest developments of DeepMind, the research arm of Google. I was intrigued to learn in this article that the increasing sophistication of AI is in turn advancing our understanding of the human brain – specifically here in relation to how we navigate space. It particularly fascinated me this notion of there being cells specialised to certain sensations (a bit like the individual conscious agents that Hoffman mentioned in the interview I posted – the specialised cells all contributing to one seamless conscious experience of space) – one type acting as a ‘you are here’, one for the direction of your head, and another forming a grid for relative position/spacing.

This regular pattern of grid cells was especially intriguing for me – here it referenced a hexagon, though it could also be considered a tesselated pattern of triangles. It has previously been proven in rat brains, and only in the past few years have we seen early proofs it could too be present in humans.

I love this notion of a secret/hidden mental map that is created to help us navigate the world, that is underlying our visual interpretation of the world. I wondered too if there could be a future in which we could ‘hack’ this secret mental mapping system, to make familiar spaces that we are new to, or allow you to immediately understand the best route through a maze, for instance. Use of this regularised grid could help you to better predict or adjust to new situations, which of course the Future is the ultimate one.

This strikes me as similar to the dots used in motion capture technology – to help digital imaging ‘navigate’ an actor’s body/face in order to best map it to a 3D computer generated image.

That this should be in a hexagonal grid was also very intriguing for me. Instinctively, I imagine organic forms and shapes to be irregular, curved, amorphous, but hexagons are geometric, regular, straight edged. That said it is a form we do see in nature, for instance in honeycomb structure, or the basalt columns visible in places such as a the Giants Causeway. It is described as being the most ‘efficient’ shape, and hence why the cooling rock forms in this way. It is also described as the strongest shape – hence it’s use in the structure of the strongest known material, graphene.

Giants Causeway basalt columns

It was thus especially eery to come across this display in the Whitworth, just a day after reading of grid cells.

I was interested too to see what these grid cells might look like in reality. Unfortunately it does not seem there are specific electron microscope images of these in existence, so I instead looked to images of generic neurons.

These images, magnifying the molecular to amazing detail, show an intriguing textural quality, and complex intricacy of connection
I experimented recreating this texture using gummed tape and tissue paper, along with ink and paint. I could not quite achieve the level of intricacy I might have liked but it was an intriguing effect. I particularly liked using the gummed tape to make the 2D surface more structural
With another, less magnified image (the original here on the right), I found hexagonal areas within it and cut or drew over to indicate these, before overlaying it with a regularised pattern on tracing paper. I particularly like the effect this produced, the juxtaposition of the geometric and irregular, organic and structured.

Unit 2: Fine Art – Dérive (pt 2)

Following on from my previous post, where I noticed I had repeated a certain composition across some of my photos of the walk, I sought to research a little more about this.

I recalled seeing something similar in Van Gogh’s landscapes that I had seen in the summer, at the Van Gogh in Britain exhibition, and looking into more of these confirmed my theory that this could have been something I picked up on there. Here he has also made effective use of yellow and blue to make even greater emphasis of this composition – dividing the canvas by its horizon near the middle, but off-centre focal point (though his tending to the right where mine was to the left). I find it interesting that he would be returning to a similar strategy for two very different scenes – the rural and the urban, with comparably similar palettes also. The small red tree on the path in the left painting is positioned almost in the same place as the figure in red on the streets in the right one.

I wonder if there was intention for Van Gogh behind this or if, like in the case of my photos, it was accidental or perhaps even just a result of him having honed his style and preferred colour palette? Were it to be intentional though, it could be seeking to draw comparisons between these different locations – or perhaps serve as a reminder that though they might seem opposing locations, that they share the same viewer (or that we are seeing it through the eyes of the same artist) they have a unity of experience? That no matter where we might find ourselves at a given moment we still experience the same ‘what it’s like to be me’ in that moment? It might be interesting to create a shared colour palette for my three compositions to explore this further.

I used google image search to find out if the algorithm could find further instances of this composition. However, I found that the image recognition software was more apt to see the subject than compositional comparison – this suggests to me a more sophisticated programme than where it might first have looked at blocks of light/shade/colour (and so composition) but now is identifying objects within those blocks. It was interesting to me to see that the software was distinguishing 3 different subjects in these images: Tree, Apartment and Street, reflecting the change of environment along my walk. The images it saw as being visually similar all fell within these 3 categories, and some do replicate my composition.

I sought to further abstract this composition, and this deconstruction of three complex images was interesting. I think the cut paper works are more successful – the precision of the shapes achieved and the flattening of the colour fields I think work well to focus the eye on the shapes and their relationships together. A more simplified colour palette also helps here I think, and I like the use of a contrast for the ‘horizon’ line. I wonder if a casual observer would still get a sense of the perspective in the original composition, or if these flattened fields would disrupt that sense of your eyes being drawn in.

These recall for me geometric abstractions like those of Malevich and Moholy-Nagy. It might be interesting to explore further whether some of my fields are overlapping of other shapes and to explore more tonal colour palettes.