Edward A. Shanken on art and AI in the work of Leonel Moura and Stelarc
In retrospect, we may look upon the long tradition of figure sculpture … as an extended psychic dress rehearsal for the intelligent automata…. As the Cybernetic art of this generation grows more intelligent and sensitive, the Greek obsession with “living” sculpture will take on an undreamed reality.
— Jack Burnham
The idea that nonliving matter could be used to invoke, influence and emulate living, intelligent beings may be nearly as old as human life itself. Over thousands of years this concept has become deeply ingrained in the human imagination as a locus of desires and fears about the future and about the role of art and technology in forming it. Embodied in the roots of human culture is, says Jack Burnham, the great theorist of systems art, a “yearning to break down the psychic and physical barriers between art and living reality — not only to make an art form that is believably real, but [is] capable of intelligent intercourse with their creators.” According to Artist Eduardo Kac, “As artists continue to push the very limits of art… they introduce robotics as a new medium at the same time that they challenge our understanding of robots –– questioning therefore our premises in conceiving, building and employing these electronic creatures.”
In the myth of Pygmalion, a sculpture that exceeded the beauty of nature itself inspires love in the heartless, misogynistic artist-king who carved it. Galatea is brought to life by the goddess Aphrodite, and lives happily ever after with her creator, Pygmalion. Reproduced in countless graphic depictions, this myth also inspired Carlo Collodi’s The Adventures of Pinocchio (1883) and George Bernard Shaw’s play Pygmalion (1913). These twin mythic bloodlines have, in turn, spawned many popular offspring, from Walt Disney’s 1940 animated film Pinocchio to Stephen Spielberg’s A.I. Artificial Intelligence (2001) and Spike Jonze’s Her (2014), on the one hand, and My Fair Lady (1956), Trading Places (1983) and Ruby Sparks (2012) on the other.
In legend and in practice, artists have played a major role in bringing matter to life aesthetically, and in bringing that metaphorical living matter into culture, where, one might say, it takes on a life and intelligence of its own. Visual artists Leonel Moura and Stelarc, whose work deploys robotics and artificial intelligence, offer valuable insights into contemporary perceptions of the human condition vis-à-vis the cyborgian condition that we increasingly find ourselves in. Their works provide, moreover, a rich framework for examining the potentials of superintelligent and hyperconscious surrogate beings.
Leonel Moura’s Artsbot can be considered within a lineage of drawing machines that includes Jean Tinguely’s mechanical Metamatics from the 1950s, Harold Cohen’s Aaron computer program, initiated in 1973, and Roman Verostko’s Derivation of the Laws, 1990. What distinguishes Artsbot from these works is that it consists of simple autonomous robots that collectively respond to each other and their environment to generate emergent complexity, a fundamental principle of scientific research in artificial life and synthetic biology. Initially developed in 2003, Artsbot builds on Moura’s earlier computer-programmed and robotically controlled painting machines that generated the artist’s Swarm Paintings (2001). In these earlier works, a brush in a CAD/CAM machine paints virtual trails of ants, resulting in an image of painterly gestures that balance randomness and intentionality. The trails are algorithmically generated, based on statistical data of how the insects deposit pheromones and the duration those pheromones linger before evaporating. As such, the brush strokes occur only when a certain threshold of virtual pheromone presence is reached, resulting in a complexly punctuated agglomeration of trails, much as one might observe in a time-lapse photograph of ants crawling.
In Artsbot, the virtual ants of the Swarm Paintings are replaced by actual robotic devices, or “artificial ‘organisms,’” as the artist refers to them. From a few basic rules, complex collective behavior emerges, which is visually recorded in the resulting drawings. Whereas the Swarm Paintings traced a trail of virtual insects based on an algorithmic emulation of actual data, the Artsbot robots are themselves artificial insects that leave a colorful trail of their own paths, which cannot be anticipated in advance but can only emerge collectively as they interact with each other and their environment. In this respect, the robots draw on early experiments in artificial intelligence and robotics, such as neurophysiologist Grey Walter’s “tortoises” Elmer and Elsie (1948–49).
Walter developed these analog devices with the goal of demonstrating that complex behaviors could be generated by a small number of connected neurons. To model the lifelike behavior of a biological nervous system, Walter’s tortoises used just two artificial neurons. A photoelectric sensor and touch sensor enabled Elmer and Elsie to demonstrate simple approach and avoidance behavior. Each explored its environment by moving toward illumination. If the brightness failed to reach a certain threshold, the tortoise would scan left or right. If strong light exceeded a certain threshold, the tortoise would approach a weaker light source. Since they also had their own pilot lights, they would approach each other or their own reflections in a mirror. Whenever a tortoise bumped up against an obstacle, such as walls, people or its robotic sibling, it would back away. Video documentation and time-lapse photographs of Elmer and Elsie demonstrate that especially complex collective behavior emerged when they interacted with each other.
Moura’s microprocessor-controlled robots are also very simple by design. Like Walter’s tortoises, they have two sensors: one photoelectric sensor to “see” color, and one to “avoid” obstacles; and two pens from which to select, depending on adjacent colors and the robot’s programmed temperament, which includes random elements. Based on a few basic rules, the Artsbot robots interact with each other and with the emerging traces that they make, demonstrating complex and variable behavior, becoming more or less “excited” depending on factors such as the level of color density. Swarms of five to ten bots work together to paint the richly patterned canvases that are generated by this project.
Whereas patterns in the Swarm Paintings are generated by virtual insect algorithms derived from scientific observation of living beings, the Artsbot robots are programmed more abstractly but nonetheless generate lifelike behavior. This outcome suggests that intelligent behavior depends not only on centralized processing but also on connectivity and embodiment. The resulting patterns are far from random; rather, clearly defined clusters emerge that reveal something of the personality of the individual bots that collaborated in their creation and the collective behavior of the system.
More than any other artist, Stelarc has challenged the physical limits of the human body and mind with respect to technology. Stelarc initially gained notoriety for his controversial suspension performances, begun in 1976, in which he suspends his body (which he invariably refers to as “the body”) by cables attached to meat hooks inserted in his flesh. Around the same time, Stelarc began using electronic media in his artwork. Indeed, these two aspects of his practice are integrally related to his theories on the physical limits of the body, which constrain the capacity of the mind. Writing in 1989, he claimed,
Our actions and ideas are essentially determined by our physiology. We are at the limits of philosophy, not only because we are at the limits of language. Philosophy is fundamentally grounded in our physiology…
It is time to question whether a bipedal, breathing body with binocular vision and a 1400cc brain is an adequate biological form. It cannot cope with the quantity, complexity and quality of information it has accumulated…
It is no longer meaningful to see the body as a site for the psyche or the social, but rather as a structure to be monitored and modified — the body not as a subject but as an object — NOT AN OBJECT OF DESIRE BUT AS AN OBJECT FOR DESIGNING.
The performative body and the cyborgian body are continuous in Stelarc’s oeuvre, and both are inextricable from the mediation of human intelligence and volition. In some of his performances, such as Handswriting: One Word with Three Hands Simultaneously (1982), the artist retains full control of robotic devices. By contrast, in Ping Body (first performed in Sydney, 1996), Stelarc subjected his body to the control of the ebb and flow of amorphous data on the internet, which triggered involuntary physiological responses. Both of these performances incorporate the Third Hand, a robotic device that functions as a cybernetic extension of the human body. As Stelarc explains:
The Third Hand is a human-like manipulator attached to the right arm as an extra hand. It is made to the dimensions of the real right hand and … is controlled by EMG signals from the abdominal and leg muscles…. By contracting the appropriate muscles you can activate the desired mechanical hand motion … to operate the Third Hand intuitively and immediately, without effort and not needing to consciously focus.
To perform Handswriting, the artist trained himself to write the word “EVOLUTION” simultaneously with three hands: two human hands plus the Third Hand. In other words, for the first sequence of letters, as the left hand was writing “E,” the right hand was writing “L,” and the Third Hand — controlled by abdominal and leg muscles — was writing “I,” and so on. If this were not challenging enough, the word was written on a sheet of Plexiglas to be read by the audience viewing from the other side, so the three hands were simultaneously writing in reverse, from right to left.
“Pinging,” a network term for a signal sent by one computer to determine the presence of another, is made corporeal in Stelarc’s Ping Body. A remote audience could access, view and actuate the body of the artist via the internet. A website provided an interface to a computer-based muscle-stimulation system that permitted those logged on, including robotic web-crawlers, to “ping” various limbs with an electric signal, causing involuntary movements in the artist’s body. This resulted in a haunting dance that recalled Balinese shadow puppets, made all the more dark by the loud electronic music generated live from network data. While the artist’s body became a robot controlled by the internet, Stelarc retained control of the Third Hand, introducing multiple levels of control and communication in the system. The body itself becomes the object of inspection in the network, rather than the subject that surfs the Web. By exposing and inverting a commonplace practice of the internet, Stelarc demonstrates, in a highly visceral form, the physicality of online media and the relationship between embodiment and disembodiment in computer networks.
Moura and Stelarc offer expanded conceptions of what robots can be and what the mind — radically expanded and collectively wired — might be capable of if untethered from the limits of the human brain. Further, their work explores the relationships that amplified forms of intelligence and embodiment might enable between humans, machines and cyborgian hybrids. It is at these interstices that visionary artists are manifesting the “undreamed reality” that Burnham envisioned half a century ago for the cyborg future of art.
Since the late twentieth century, art has made two-way interaction increasingly central and explicit. Viewers have become more interested in playing an active role as agents who determine salient features of an artwork. As the line between artist, artwork and audience becomes increasingly blurry, and as humans and machines become increasingly intermixed, the human condition and the machine condition will meld into a posthuman or postmachinic condition in which robotic devices enjoy constitutional rights and in which hybrid entities push the limits of current ontologies and ethics. While such issues have been debated for years, in early 2017 the European Parliament passed a resolution on civil law with respect to robotics, with recommendations to the European Commission, including a proposal for a special legal status for intelligent robots as “electronic persons.”
The current theoretical trend of object-oriented ontology rightly questions the anthropocentric bias of Western philosophy and attempts to understand nonhuman entities and their relationships to each other in their own terms, rather than in human terms and for human purposes. Perhaps it is in this context that Moura’s Artsbot may be best understood. Although the robots’ actions result in paintings that serve a human aesthetic function, their lifelike behavior suggests that apparent intention, intelligence and creativity may not be strictly the province of humans and other animals. Human understanding and human civilization may be just one of many varieties of intelligence and society. Since the Industrial Revolution humans have deteriorated our own living environment to such an extent that our well-being is gravely threatened. If Stelarc is correct, then perhaps we can — and perhaps must — learn from other forms of life, intelligence and culture. Perhaps only by expanding, if not hybridizing, our minds and bodies, by embracing and merging with other forms of life and intelligence, can we survive the increasingly perilous conditions of climate change, which the limits of human intelligence, altruism and restraint have precipitated.
Edward Shanken is a professor at University of California, Santa Cruz. His books include Systems (Whitechapel/MIT, 2015), Art and Electronic Media (Phaidon, 2009), and Telematic Embrace (UC Press, 2003).