Introduction

This is Iteration 4 of exploring ideas for logos for the Neuro-Symbolic Systems AI research community that is emerging within the Computer Science department at City, University of London.

Iterations 2 and 3 focus on exploring a specific motif for building an icon that combines neurons and symbols — a motif where neurons appear in the left hemisphere of a brain cortex and symbols appear in the right hemisphere of that brain cortex. The purpose of Iteration 4 is to explore alternate motifs for arranging neurons with symbols.

One cool thing about Iteration 4 is that most (almost all) of the images displayed here were sourced from a community of graphic designers, musicians and video artists called Pixabay. All of the content of this awesome site is governed by the custom Pixabay license which stipulates that all content is free to use and amend, without attribution, even for commercial purposes. Pretty much the only thing one cannot do with the content is redistribute/sell the originals or sell one’s amended versions of the originals. I recommend you go to Pixabay and check it out! Type something like ‘brain’, or ‘artificial intelligence’ or ‘machine learning’ into the search bar, press ENTER, and get BLOWN AWAY!

Motif 1: networks of neurons with a symbol inside each neuron

Image 1 is what inspired Iteration 4: the idea of placing the symbols inside the neurons. The pencil notion in image 1 isn't relevant to us, but the rest may be of interest. The neurons in image 1 are arranged in a circular pattern. We can stick with this if we like it. But we might explore alternate shapes/arrangements as well. For example, we might explore arranging them in patterns suggestive of a brain cortex with left & right hemispheres (as in the images of Iterations 2 and 3), or in patterns suggestive of a brain cortex viewed from a side angle, such as in images 2 and 3. Or we might explore arranging the neurons in more regulated patterns as in, say, a conventional, layered neural network; or as in a more complex version of the regulated pattern shown in image 4; or in a hierarchical pattern similar to that shown in image 5. We might also explore placing the pattern of neurons (containing symbols) within a human head figure, as in image 5. (Image 5, by the way, is from the front cover of Judea Pearl's book "Probabilistic Reasoning in Intelligent Systems".)

These are additional brain images, similar to Image 2 (above), that might be rendered/adapted so as to permit symbols being placed within neurons.

Motif 2: layering neurons and symbols

Motif 2 is about the option of exploring the layering of neurons and symbols.

In image 1, we have cogs/gears arranged in the shape of a brain, against a background of networked neurons. We could play with this theme and have arbitrary symbols arranged in the shape of a brain against a background of networked neurons.

Image 2 illustrates the layering of neurons: networked neurons layered over networked neurons. Image 3 combines the layering of networked neurons with the idea of placing symbols inside of neurons, per motif 1. I absolutely love image 2! Upon seeing it, my first thought was "if only we could place symbols within the big neurons of the top layer". Then I found image 3, a clear variation of image 2, which does precisely what I was imagining. Image 3 repeats the use of the same symbol rather than going for variety, and its green colouring is unappealing (for me), but it illustrates the potential for adaptation of image 2.

Motif 3: two human head figures

Motif 3 is about the option of exploring the use of two human head figures, one for hosting networked neurons, the other for hosting an arrangement of symbols. Some visual mechanism would then be found for suggesting interaction or communication or integration between the two realms.

NOTE: I ended up exploring this motif as part of Iteration 3 of this project. Please see the Iteration 3 project page for an exploration of this motif of two human heads facing each other. Image 2 (seen here) was part of the inspiration for that exploration, including the digital signal idea.

Motif 4: miscellaneous awesome images

Motif 4, at this stage, is really just an excuse to share with you more stunning images from Pixabay.

Motif 1 Exploration

Here we show an exploration of motif 1, where we place symbols inside neurons arranged in the shape of a brain. Image 6 of motif 1, repeated here for convenience, was the primary inspiration. I created an arrangement of neurons virtually identical to it except for a small adaptation in the centre to permit one more large-sized neuron capable of hosting an additional symbol.

The resulting brain icon is shown at four different scales, to help people judge its visual impact when used in different contexts.

If we combine this brain icon with a community name & tagline, we get a full logo such as this:

The 3 instances of the logo are identical. I'm just presenting them at different scales.

head-with-brain icon motif

We can take this motif a bit further by exploring placing the brain icon within a human head figure. For this, I reuse the head figure I worked with in Iteration 3 of this icon/logo project. I show two versions of the resulting head-with-brain icon, identical except for different background colours.

If we combine this head-with-brain icon with a community name & tagline, we get a full logo such as this: