Anthroptic is a collaboration between new media artist Ethan Ham and writer Benjamin Rosenbaum. What is immediately striking about this collaboration is the way that their combined efforts exceed the individual contributions of either artist; the whole is more than a sum of parts. Each image and its companion text present us with a brief glimpse into an imaginary world. Opened and closed by a pair of poetic vignettes, the eight sections of this work assume a serial character whose collective impact is a sense of both fear and foreboding. This sensation is crucial to the meaning of the whole.

The threat of these tales, when coupled with the opening and the associated images, requires some background. Each image in this work was chosen from the on-line database of pictures known as Flickr by an automated program, a sophisticated piece of software that does something we humans usually take for granted: it recognizes faces. Facial recognition is normally taken to be a sign of sentience of some sort; does the machine then qualify as intelligent? No, or at least not in the sense that the functional AIs of popular entertainment are “intelligent.” This machine is not HAL; instead it is a variety of program that has been developed for a variety of uses-and the most famous of these all have sinister overtones of surveillance, automated control, and suggestions of totalitarian police-state tactics:

The robot watches.

The poetic opening frames what follows, but what follows is different than the threat these statements of automated observation imply. Of the eight photographs this software system has identified as faces, two are pictures of faces, but not actual faces, one is a Public Market sign in Seattle, one is a cat, another a flower, a third is the background behind a man drinking beer, a section of a city, and finally a tram. Sometimes we see a “good gestalt” and other times we don’t. Should we relax? These are all “false-positives” things the machine believes to be faces, but we being humans can instantly recognize that they aren’t. The robot is broken.
However, it is the possibility that the software may not be broken that is implicit in each of these stories and it is this possibility that produces the threat. In looking at the pictures we can imagine the program is not broken at all, but is instead daydreaming. We can imagine the software is insane. This potential interpretation requires us to admit that the machine can be like a human; at the same time, it means that people are like machines. It is the overlap of human and machine that disturbs the easy assumptions of there being a difference between intelligence (human) and device (robot).

There are elements to these pictures that support an overlapping of mechanical and biological intelligences. If we look at the photographs of things we don’t commonly see as “faces” – such as the tram – and allow ourselves to see them as the machine has, what we find are faces. It is the observation made in the Italian Renaissance by Leonardo-that when confronted by a random splattering of paint our imaginations start to see definite shapes, people, animals, faces. They are there only for the duration of our glance, then they vanish. It is these that the robot has found, presenting them to us for our inspection.

In Ethan Ham’s original project, the photographs were presented on a website where viewers could see both the current photo being considered and past identified faces. The idea of this project was simple: his robot was set loose on Flickr to find one face, that of the artist. In a sense, it was sent to try and find it’s maker. But demiurge-like, Ham “tweaked” the robot’s face recognition processes at the start of the project so it would misidentify faces. Some of these faces are part of Anthroptic. However, the joke is on us: because the machine “sees” faces where there are none, Ham has introduced a variable into it that under normal circumstances would be error. Yet, what it finds can be recognized as “faces” if we allow ourselves the opportunity to look at these pictures with eyes trained by a century of avant-garde art such as Surrealism or Cubism. In doing so, we share in the robot’s delusion.

Benjamin Rosenbaum’s narratives range across a wide territory. The Public Market narrative offers a degree of explanation about what is happening, that the robot is simply a software program that measures certain physical measurements, like the aspect ratio and the distance between the eyes. While this may be true of the mechanism driving the software, his description is also lacking in a crucial aspect – affect – thus his explanation of how his robot identifies faces is similar to describing color by the specific wavelength of light. Rosenbaum’s narrative recognizes this fact about the images.

The historical conflict between artists and scientists over the nature of humanity and the world that shaped the art world in the early twentieth century – the conflict over whether humans are machines, (or machine-like), or not – is implicitly in this work. The sense of threat and fear that pervade these stories is the heritage of this division between the empirical, technical description of reality and the subjective, interpretative response to reality. Abstraction may be one of the most visible of the results of this debate; to be able to locate it continuing in a contemporary work may be a necessity since that debate never actually ended. The threat the robot poses arises precisely because it is instrumental in its engagement with the world; the robot, being a machine necessarily engages the world in a mechanical fashion. Software does not dream, it cannot be insane. To consider that possibility it to admit the machine into the realm of the human. This admission forces us to also admit that there is something machine-like about humanity. Thus when confronted by these images, we try to avoid the threat the robot always poses for us. The robot threatens the idea of what it means to be “human.” The irony of this situation is that we assign the robot more human characteristics, bringing it closer to ourselves as a way to dispel the fear it evokes.

We humans may prefer to see these “errors” as the robot dreaming or as insane than to consider the alternative, mechanistic implication: that we are closer to the machine than it is to us. The idea of a bifurcation – the mind/body split, the “ghost in the machine,” – has been a way of avoiding the implication that humans are elaborate systems, physical in nature and subject to malfunctions of the same type as the robot.

It is comforting to imagine the artist as demiurge deluding his creation in its search for its creator on Flickr. But this imaginary narrative is just that, a comforting tale invented to dispel the threat the robot implicitly poses to our ideas of being”human.” The various serial narratives that emerge in Rosenbaum’s writings all speak to this fear: that the robot is mechanical, not intelligent, that it acts is service of other humans, able only to follow instructions and do what it is built to do. The AI fantasy of a dreaming machine is just that – and the repressed fear emerges in these stories as a managed threat. The International Banker who rules the world by accident, and then does nothing with it; the photographer who shoots a picture, uncertain of her subject’s true nature; the domestication of humans.

Semiologist Umberto Eco has observed that serials build depth through repetition, and the complexity of serials is a result of what doesn’t change. The constants here are the robot’s recognitions, the threats contained and implied by these stories and the resistance to the underlying proposition that we are like the machine. This is the message of the paired poetic vignettes: one is the robot, the other is you, the reader. Superimpose one on the other and we can recognize both the fear and the resistance. Our invention of these stories helps cushion the fear the robot poses. They are an act of resistance that describe a complex engagement with the pictures and their implications. It is this combination that makes Anthroptic a true collaboration. Each artist reveals something implicit in the other through their mutual engagement.



Michael Betancourt is a curator, avant-garde theorist, and multi-disciplinary artist. He has been making movies, site-specific installations, and non-traditional art forms (and exhibiting his work in unseen, unusual, or public spaces) since 1992. Journals such as Leonardo, Semiotica and CTheory have published his essays. His blog is located at and his portfolio can be seen at