You are in a fine art institution in a major art hub city. Before you, dancers perform. They are black-clad, elegant, vulnerable, and human.
What you see is hinted at in a choreographic score you own.
You see the dance and only then realize you own the dance, but this experience is an experience, which is much different from viewing the choreographic score. The choreographic score might be informative but it is merely that.
You watch the dance that comes from the score. It’s beautiful, and it feels like such an odd thing to own. Does anyone ever own a dance like one owns a painting or rights to a song? Now you do, and there are 399 more scores in the world, produced by the same process, and owned by people, some of whom own more than one. You wonder if they own, say, two dances or one dance that’s twice as long with perhaps an intermission.
When you look at the score, you are the first person to ever see it because it isn’t created by a person but by an algorithm. “Has an algorithm ever choreographed anything?” you wonder. Is this one puppy from the debut litter of dance pieces ever choreographed by a machine?
The score itself comes from an artwork that looks nothing like dance. The artwork you own only has a hand, or the partial warped silhouette of a hand amid a black and white swoop of shapes that seem peered at through a funhouse hall of mirrors.
You buy this piece of art online, not even with money you earn from your job but with money you essentially create, or employed a spare computer to create, money that isn’t even recognized by most people as real or legitimate, and that’s how you buy this work that isn’t made yet but is made at the moment you purchase. The website asks you to accept cookies. You accept cookies, but not those kind.
You don’t know the artists, but you follow their words and pictures on a website where lots of people publish their words and pictures in a running mono-dialogue-diatribe and lots of those nameless faceless people are also talking about these artists.
The artists are curious about how technology obscures humanity and how it might reveal humanity. So they make this work.
They look human enough, though you only see pictures on a screen.
The Concept Behind Human Unreadable by Operator
“Immersion is not just a physical state but an emotional one”
– Dehja Ti
Conceptual/experiential art duo Operator (Dehja Ti and Ania Catherine) were exploring the idea of data becoming a new form of identity and the implications of translating human essence into a digital format when the initial seedling for Human Unreadable emerged.
Throughout their creative history, the duo has created experiential work that urges viewers to critically engage with the tools that shape their daily lives. In On View (2019), Operator turned the phenomenon of Instagrammable art installations on its head by creating a confrontational, visceral experience of mass surveillance.
Complementing On View, I’d rather be in a dark silence than (2020) is a signal-blocking trench coat, the first piece from Operator’s Privacy Collection, which interrogates “the tension between privacy and transparency inherent to the blockchain, as well as the contradiction of self-expression and anonymity online.”
Created with fashion designer Barbara Sanchez-Kane, the coat merges conceptual art and function. Its visually striking pocket is lined with military-grade fabric to keep wireless devices offline, using techniques from law enforcement and digital forensics, thus highlighting the issues we face each time we agree to problematic terms and conditions.
Human Unreadable, the duo’s celebrated three-part generative art project from 2023, expands on their previous work. This work explores how technology, while enhancing connectivity and access to information, simultaneously renders certain aspects of human existence opaque and difficult to interpret. We live in a constant state of choosing what parts of ourselves to hide and which to expose within our technologically mediated relationships and presentations of self.
At its core, Human Unreadable navigates the paradox of enhanced visibility and growing incomprehensibility in the digital age, inviting readers to reflect on their own digital footprints and the extent to which they are “readable” by the technologies that increasingly govern our lives.
The Creative Process Behind Human Unreadable by Operator
Initial Conceptualization and Planning
For Operator, the concept is sacred and all creative efforts are in service to it, which is why the first step to creating Human Unreadable involved setting creative constraints to ensure the concept was clear: the slow recovery of the human. The piece was to be long-form generative (meaning uncurated — the artwork is created at the moment of minting) and centered on the human body while exploring and commenting on transparency and privacy.
With approaching both web3 and long-form generative, they thought of the work as site-specific, so they wanted to plan for a project that would exploit the features and limitations of their chosen “site.” Through these conceptual constraints, they arrived at the idea of hiding the human body and raw expression in code (unreadable to humans) through on-chain generative choreography.
Second only to the concept in Operator’s stack-ranked priorities is the research, which was extensive. Outside of the technical considerations (more on that below), the duo investigated the precedents of what they were attempting, including the computational choreography of Jeanne Beaman and Analívia Cordeiro, the Chance Dance choreography of Merce Cunningham (as well as the chance operations employed by Cunningham’s partner John Cage), and the work of the artists involved in E.A.T. (Experiments in Art & Technology).
Looking at their concept through the prism of this research, Operator arrived at an aesthetic goal: to use the historically rigid modern tools of computation to represent chaos, human messiness, and organic feminine touch to shatter the Swiss grid and quiet biases of modernist design.
Technical Development and Artistic Execution
With the conceptual groundwork laid, the project moved into the technical development phase, which would eventually involve more than 20 dancers, 5 motion capture technicians, and 4 computer engineers.
The idea of Human Unreadable is that it starts with the human and ends with the human, so they first had to get humans dancing.
To develop a system for generating choreography, they needed to construct a movement library. After some early prototyping, they trained dancers on 31 movements and assigned them random sequences to ensure the final product was artful. After multiple rehearsals, they used motion capture technology to translate the movement into data. What followed was a long process of reduction — both of the file type (from .abx. to .csv to .json) in a process of their own devising as well as minimizing the data needed to represent a given dance move, which required them to reduce the number of body parts needed to represent each dance move. Finally, this set of bare-minimum data was obfuscated in hashes for the ultimate generative algorithm to work with.
But the motion data isn’t useful without an image to move. Having worked with themes of transparency and privacy for several years, they revisited the recurring motifs of glass, X-rays, and light, which needed to be re-created with code.
The final ingredient that would go into the outputs was hints of the human body. When people first learn about Human Unreadable, it’s a common misconception that the body parts hint at the underlying movement data that influenced the output, but the two are separate. In Human Unreadable’s output, the human body is an art object, not representational of the sequence behind it. To include this element in the final product, they held a series of photoshoots, turned the results to resolution-agnostic body parts as .svg files and eventually managed to get the images into P5.js.
If it wasn’t hard enough to make the generative images and choreography in the first place, both had to be artful and beautiful. So, next came an extended period of trial and error, testing and refining, both with the generative image outputs and the accompanying choreographies.
Nine months later, when all was ready, the ArtBlocks minting of 400 unique artworks kicked off a kind of durational performance. After a period of time, collectors were invited to go to the Operator site to unlock the secondary NFT (forever married to the primary artwork) which revealed the movement score on which the artwork is based. From the Man Ray-inspired but abstract images of the first, collectors got to see, as stick figures, the human movements that created it. The recovery of the human had reached its next phase and collectors could get a glimpse at what the human “did” to make the image.
In the end, which has not yet been fully revealed, Operator will create an evening-length performance. This final performance will use the sequences from the first 100 mints (#2-#101) as raw material, staged into an original site-specific work which abides by all the rules of the movement scores it includes. While the choreographic tuning sessions gave the artists a glimpse of the types of sequences the model produces, it will be an entirely new process to stage Act III. This process will involve sorting the sequences, rehearsing, and blending them into a cohesive evening-length work.
The final performance reveals something that is readable to humans: other humans. In this way, the dance is like an unzipped file, technology as an embodied experience to express the messiness of being human.
