Creative studio Field updates the visual language surrounding AI in its series System Aesthetics

In recent years, the conversation surrounding artificial intelligence has exploded significantly. What was once understood and discussed by an elite few is now a well-known trope in the average household. Despite this, hundreds of articles and advertising campaigns still rely on the visual lineage established in the 1980s with films like Tron, Hackers and The Matrix where “a faceless, hooded figure, types green bits of code into a cryptic-looking terminal, in a dimly lit environment.”

This astute observation was what prompted Field’s “curious and slightly nerdy bunch of people,” to take on the challenge of reinventing this language by visualising the invisible. In its series System Aesthetics the creative studio gives shape, form, colour and movement to a series of algorithms in an attempt to aid greater understanding of how they function.

“There is so much visual language in the world already,” explains Field’s creative director Marcus Wendt, “the only image worth making, in my opinion, is the one that doesn’t exist.” Alongside his ten-strong team (“a colourful bunch of creatives with different backgrounds”), Marcus set about “discovering and exploring these previously unseen, fruitful fields, within the vast space of visual possibility.”

Featured in Wired magazine’s World 2018 issue, the static and moving illustrations depict the technology that powers voice assistants, generative adversarial network, bitcoin, self-driving cars and face-hacking. In order to conceptualise how these algorithms could and do look, the team spent 2017 intensely researching to develop an in-depth understanding of machine learning, its sub-techniques, its usages and the discussion surround it. “For most of the illustrations in the series, we used some hard-fact, authentic datasets as a source or driver, for example the exact layout of neuron nodes and subnets of the Google SyntaxNet speech recognition neural network, or a full day capture of Ethereum blockchain transactions,” Marcus tells It’s Nice That.

In some instances, the algorithm itself provided a visual metaphor – for example, a blockchain is a sequence of blocks of information – meaning the challenge was to give them a “compelling form and a fascinating behaviour.” Other times, things were more ambiguous and so the team took inspiration from the scientific language that researchers at Google, Stanford University and elsewhere use to describe their own inventions. However, Marcus describes how where they found a new visuality that interested them, they “rigorously questioned it logically and intuitively, revising the idea until it ‘felt right’.”

The resulting series not only provides a conceptual and creative interpretation of artificial intelligence but also has practical implications. Field has created a contemporary langue to provide a level of access to smart technologies which can be used to support a public debate and also help brands communication their new products and services.