Neural Unconscious

‍

Neural Unconscious

Peter Bauman (Monk Antony) explores how technology has always changed the way artists see, create and think—first by revealing hidden details (optical unconscious), then by structuring information (computational unconscious) and now by co-generating new possibilities (neural unconscious). While these forces open vast creative spaces, they do not necessarily equate to quality and come riddled with ethical, political and moral challenges that demand critical engagement rather than passive adoption.
‍
‍Like the universe, art is expanding due to the stretching of space itself—and at a hastening pace. There’s an invisible, mysterious force—art’s dark energy—driving this acceleration.

Technology.

But it’s not so mysterious after all. It's actually part of us. As Joanna Zylinska said, “We have always been cyborgs,” always using instruments—technologies that expand our perceptions and capabilities—to enhance our creativity.

How have these cyborg-like instruments evolved and how have they shaped the way humans make? Computational philosopher AA Cavia frames intelligence as an active process of navigating—even expanding—a conceptual landscape. Cavia then describes creativity as mapping and realizing possibilities within that space—an artist’s unique voice.
‍

Thinking of intelligence and creativity as explorable landscapes clarifies terms like latent space while linking them to art history’s traditions of pictorial and output spaces.

Another key feature of intelligence-as-exploration is that it emerges from processes rather than inherent traits, remaining unfixed and context-dependent. In this sense, humans, ants and deep learning algorithms can all engage in intelligence—and creativity.
‍

‍

Creative Spaces

A creative space can be any creative framework an artist builds and navigates but dominant patterns and shared constructs have emerged over time. Pictorial space—bound by the canvas and closely tied to painting and photography—defined visual representation for centuries. In pictorial space, artists are restricted to representations on a single, static, flat surface.

The borders of the two-dimensional plane delineate creative potential. Despite these limitations, some of humanity’s most valued artifacts—and most of the work gracing fine art museums across the globe—have seen human artists engage with this creative landscape. Exploring it means translating a vision onto a static 2D plane, which artists historically achieved through techniques like linear perspective, chiaroscuro and compositional framing. Modernists like Braque fractured while Mondrian flattened this space, manipulating it to its extremes.
‍

In the post-war era, creative spaces expanded beyond the pictorial to the output space of a system’s possibilities. Artists embraced ideas like cybernetics and the systems thinking of Jack Burnham’s Systems Esthetics and Roy Ascott’s Behaviourist Art—while de-emphasizing the static art object. Movements emerged like computer, conceptual, performance and kinetic art, exemplified by groups like New Tendencies, GRAV, Fluxus and E.A.T. These early ideas on systems ingrained in them a dynamism, interactivity and process that still resonates today in on-chain generative and protocol art.

Artists like Joan Truckenbrod and Frieder Nake used early computing to create works defined not by a singular image but by the range of possibilities inherent in programmed rules. The form’s potential ballooned with the on-chain generative art of Erick Calderon's Chromie Squiggle, where the output space is shaped by the algorithm’s programmed capabilities coupled with the performance of its minting.

Advancements in deep learning—especially since about 2006—propelled creativity into the latent space. Latent space exists within deep learning models—like an enormous map where similar words and ideas are grouped together—as a multi-dimensional environment of topological embeddings. Unlike the human-readable pictorial and output spaces, latent space is abstract, requiring a cyborg-like synthesis with the machine to access it—and synthesize new forms.

Artists like Anna Ridler and Stephanie Dinkins engage with latent space by first training highly personalized datasets. In Dinkins’s Not The Only One, the model’s input data comes from oral histories spoken by three generations of women from the artist’s own family. From these inputs, creative outputs are achieved that are entirely new, rather than recombinations of the existing data. The deep learning model and artist co-discover rather than one dictating outcomes—each a necessary component.

Finally, technologies like the Internet and blockchain function not only as force multipliers for these creative spaces but also as mediums in their own right, as evidenced by net and protocol art.
‍

‍

The Unconsciouses

Each of these new creative spaces was made possible by a new way of thinking, shaped by an "unconscious" force—something operating behind the scenes that expands how artists see, create and even think. Just like cameras helped us see minute details we couldn’t before (optical unconscious) and computers helped us organize information (computational unconscious), AI and deep learning now help us co-discover and co-generate new ideas (neural unconscious).
‍

Optical Unconscious

Conceived by Walter Benjamin in 1931, the optical unconscious relates to technology’s power to allow for sight beneath the surface—making the hidden visible. Theories suggest Renaissance painters used the latest Venetian lenses and optics to create their hyper-realistic visual effects, at least partially explaining the marked progress in realism in the early fifteenth century. Other factors include oil paints and linear perspective, two additional technological innovations. One of the main proponents of the unproven theory, “David Hockney says the history of art from that time is intimately linked with the history of optics itself.” Through optical advancements, artists better perceived and understood how to manipulate pictorial space.

Photography further widened our perception with its ability to capture a reality previously inaccessible to the human eye—hidden structures, movements and imperceptible details. Anticipating the higher dimensionality of latent spaces, Benjamin argued cameras unveiled new dimensions of reality itself. It gave artists cyborg eyes beyond Renaissance mirrors and lenses.

The camera’s insights liberated painters from previous approaches to visual representation, ultimately leading to Modern art. While painters can still choose to work in a hyper-realistic style, after around 1850, this was no longer the primary objective. Freed from these demands, artists could focus their time, energy and talent on exploring new creative terrain.
‍

Computational Unconscious

The computational unconscious refers to how computers structure reality through discrete, rule-based digital systems. Unlike the optical unconscious, which reveals hidden visual details through lenses, the computational unconscious reveals the hidden structures of information through data processing and algorithmic logic—making it more accessible for humans to control.

When artists began working with computers, they weren’t just using a new tool; they were engaging with a new creative space revealed through a computational unconscious—abstract, systematic ways of organizing, arranging and displaying information. Computers were more active collaborators than lenses, though, capable of processing and generating patterns based on structured logic.

By accessing this computational unconscious, creative spaces shifted. Artists had the choice to move beyond static representations (pictorial space) to dynamic, system-driven art (output space). Again, technology exposes a dimension of reality—and creativity—previously unknowable to humans. Then it allows them to control it.
‍

Neural Unconscious

The neural unconscious underlies the structure that governs how deep learning models process and generate new information. Unlike the computational unconscious, which follows explicit, rule-based logic, the neural unconscious operates through approximation, pattern recognition and relational mapping in an even higher-dimensional space. It does not rely on strict pre-defined rules but instead learns cybernetically by making connections across vast amounts of mapped data.
‍

If the optical unconscious made the imperceptible visible and the computational unconscious made the rule-based structure of reality explorable, the neural unconscious makes the unimaginable generative. 

It gives artists a cyborg mind with which to experiment and create. An example is Holly Herndon and Mat Dryhurst’s Holly+, the pair’s “digital surrogate.” As with any system—from a LeWitt wall drawing to a long-form generative art blockchain performance—this cyborg brain can be further extended to public audiences, broadening still its creative space. 
‍

‍

Dark Energy

Technology as an expansive force in art remains a dark energy. Like any power, it comes with costs, raises concerns and must be navigated thoughtfully by the yielder—with consequences for both society at large and artists themselves.

A major reason for this apprehension is new technologies often emerge from ingrained powerful institutions—governments, corporations and military-industrial complexes. Renaissance Venice was a hub of commerce and war as well as optics, where the brutal Fourth Crusade was launched and revealing glassmaking secrets was punishable by death.

Today’s AI and digital infrastructures have similar roots in surveillance, capital and control. Further, artists must navigate the ethical use of trained data—its sources, creator compensation and consent. This is all additional baggage, the costs artists pay for these extraordinary powers.

‍Militaries, governments and corporations will continue using and abusing deep learning technology whether artists choose to engage or not. The conversation can’t be left entirely to institutions of enormous power. The most forceful work moving forward will neither shy away from technology’s dark energy nor become a shill for our latest techno-military industrial complex. Artists can use their voices and talent to steer the conversation in directions of their choosing, as expertly done previously by Kate Crawford, Trevor Paglen and Lawrence Lek.

‍

Closing

Greater space for creative exploration and output does not necessarily equate to higher artistic quality. A painter can spend decades making creative choices on a single static image, while a coder can generate decades of images with single lines of code. They may not share the same aesthetic or conceptual depth. Technology and intelligences may be stretching the very fabric of expressive space itself—but creativity is still about finding and making your own corner of that space.


‍
----



Peter Bauman (Monk Antony) is Le Random's Editor-in-Chief.