New Aesthetics: Cyber-Aesthetics and Degrees of Autonomy
By Patrick Lichty – 01/03/2013
Five Thousand Feet is the Best by Omer Fast.
In perusing Honor Harger’s recent missive on drone aesthetics and James Bridle’s ongoing posts of drone images at Dronestagram, taken in context with the Glitch un-conference in Chicago, some new questions have come to mind. These questions have to do with conceptions of New Aesthetics in its various forms in terms of interaction with the program/device and its level of autonomy from the user. In my mind, there seems to be a NA continuum from generative programs that operate under the strict criteria of the programmer to the often-autonomous actions of drones and planetary rovers. As you can see, I am still chewing on the idea that The New Aesthetic as it seems to be defined, as encompassing all semi-autonomous aspects of ‘computer vision’. This includes Glitch, Algorism, Drone imagery, satellite photography and face recognition, and it’s sometimes a tough nugget to swallow that resonates with me on a number of levels.
First, image-creating technological agents are far from new, as Darko Fritz recently stated in a talk that algorithms have been creating images, in my opinion, within criteria of NA since the 60’s, and pioneers like Frieder Nake, A. Michael Noll, and Roman Verostko have been exploring algorithmic agency for decades. If we take these computer art pioneers into account, one can argue that NA has existed since the 60’s if one lumps in genres like Verostko’s ‘style’ of Algorism or the use of algorithms as aesthetic choice. A notch along the continuum toward the ‘fire and forget’ imaging (e.g. drones) is the Glitch contingent, which is less deterministic about their methodologies of data corruption aesthetics by either running a program that corrupts the media or they perform digital vivisection and watch what little monster they’ve created. Glitchers exhibit less control over their processes, and are much more akin to John Cage, Dada or Fluxus artists in their allowance of whimsical or chance elements in their media.
However, as we slide along the spectrum of control/autonomy from the lockstep control of code to the less deterministic aesthetics of face recognition, drone imaging, robotic cameras, Google Street View cams, Mars Rovers and satellite imaging, things get murkier. Autonomic aesthetics remind me of the ruby-hued Terminator T500 vision generated by intelligent agents running the ‘housekeeping’ on the machine platform. I consider this continuum from Algorism to Glitch to autonomous robotic agents under an NA continuum of aesthetics is important insofar as it defines a balance of agency between the operator and the ‘tool’. For me this is the difference between the high degree of control of the Algorist, the ‘twiddle and tweak’ sensibility of the Glitcher, and the gleaning from the database of pseudo-autonomous images created by Big Imaging created by drones and automatic imaging. Notice I use the term ‘pseudo’ in that there are operators flying the platforms or driving the car, while the on-board agents take care of issues like pattern/face recognition and target acquisition. We also see this in Facebook, as recent technological changes as of 2012 have introduced face recognition in the tagging of images. From this, a key issue for me in this discussion of what began as a nebulous set of terms (the criteria of NA as defined by the global conversation) is that of agency and autonomy, and how much control the New Aestheticist gets in the execution of their process. Another important point is that I am not calling the ‘New Aestheticist’ an artist or curator, but something in between, but I’ll get to that later as this is also an issue of control of intent.
Back to this idea of autonomy between the subject, the ‘curator’ and the viewer, what interests me is the degree of control or not that the person creating, tweaking, or gleaning the image has over the creation or contextualization of that image. In the case of the Algorist, this is the Control end of the spectrum, where the artist takes nearly full control of the process of creation of the image, unless there is a randomization function involved in the process, and that it itself is a form of control – very Cybernetic in nature. Agency is at a maximum here, as the artist and machine are in partnership. Roman Verostko is a prime example of this, as he explores intricate recursive images created by ink pen plotters using paints in the pens. What he, and the AI-driven AARON, by Harold Cohen, for that matter, are machine painting.
The next step down the autonomy spectrum would involve the use of ‘glitch’ tools and processes that distort, disturb, and warp digital media. The process involves executing a given intervention upon the medium, such as saving it improperly, hex editing its code to corrupt it, or as Caleb Kelly writes, ‘crack’ the media. There are differing degrees of disturbance of the media to inject chance processes into it, from a more ‘algoristic’/programmatic application of programs upon the media to directly changing the internal data structure through manipulating the information through hex code and text editors. The resultant process is an iterative ‘tweak and test’ methodology that still involves the user in the process to varying degrees. Of course, the direct manipulation of the data with a hex editor is the most intimate of the processes, but there is still one factor to account for. The factor in question is that there is the set of causes and effects that are set in motion when the artist/operator opens the media and the codec (Compressor/DECompressor) mis/interprets the media, as is intended by the artist.
If we are to look at the glitch process, we can say that there is a point of intervention/disturbance upon the media, which is entirely a function of control on the part of the user. Afterwards, it is set loose into the system to allow the corruptions within the media to trigger chance/autonomous operations in its interpretation in the browser, etc. This is where the glitcher straddles the line between control and autonomy, as they manually insert noise into their media (control), then the codecs struggle with the ‘cracked’ media (autonomy). The glitcher, then, has the option to try a new iteration, thereby making the process cybernetic in nature. In Glitch, there is a conversation between the operator, the media and the codec. With the aesthetics created by drones, algorithmic recognition software, and satellite reconstructions, the process is far more autonomous/disjoint, and the New Aestheticist has to deal with this in the construction of their practice.
In the genre that I will call ‘mobEYEle’ imaging, the robot, satellite, or parabolic street eye abstracts from the ‘artist’, aptly turning them into an ‘aestheticist’, as their level of control is defined as that of a gleaner/pattern recognizer from the image bank of Big Data. Rhetorically speaking, we could say that a connection between the aestheticist and the generator of the image would be less abstract if, say, a New Aestheticist were to be in the room with a drone pilot, conversing about points of interest. It is likely that a military remote pilot and a graphic designer would have sharply differing views as to what constitutes a ‘target of interest’. Like that’s going to happen…
Therefore, let us just say that the collaboration of a New Aestheticist and a drone pilot is nightly unlikely, and that the New Aestheticist is therefore abstracted from the decisions of command and control involved in acquiring the image that eventually gets in their hands. This, however, presents us with two levels of autonomous agency, one human and one algotrithmic. But before I expand on this, I would like to discuss my decision to call the practitioner an ‘aestheticist’ as opposed to an artist or curator.
This decision rests on what I feel is the function of the aestheticist, that is, to glean value from an image and ‘ascribe’ an aesthetic to it. This position puts them in a murky locus between artist and curator, as they have elements of neither and both. For example, does the drone-image NA practitioner create the image; are they the artist per se, of the image? No. Although they are more closely aligned to curatorial practice as they collect, filter (to paraphrase Anne-Marie Schleiner), and post on tumblrs and Pinterests? From my perspective, the role of a curator is the suggestion of taste through and informed subjectivity through ecologies of trust and legitimacy, but the social image aggregator, although they might want to perform the same function, has no guarantee of accomplishing this unless they develop a following. Therefore, under my definition, they are neither creators nor taste-makers in the traditional sense, so what makes sense is to call them ‘aggregators’ of aesthetic material and thus my term ‘Aestheticist’.
Returning to our conversation, the drone aestheticist, then, is subject to one of two degrees of completely abstracted autonomy of the creation of the image; that of the operator or that of the algorithms operating the drone. The abstraction surrounding the human operator is easiest to resolve, as the images of interest are either the preference of the drone operator or those created by the operator under the parameters of the mission, and not the results of a New Aestheticist’s joyride on a Global Hawk. It is merely someone else’s volition selecting the image, and a confluence of personal interest deciding as to whether the image deserves to be on the New Aestheticist’s social imaging organ. However, it is the drone’s algorithmic image acquisition system that creates a more alien perspective in regards to aesthetics and autonomy of the image.
Compared to the Algorist or the Glitcher, all loosely placed under the banner of New Aesthetics, the Drone/Big Data Aestheticist is most problematic, as they are a fetishizer of sheer command and control operations that are potentially utterly abstracted from the pilot/driver’s volition. This creates a double abstraction through first the pilot, and then the algorithmic recognition system. There is no cybernetic loop here at all, as the gleaning of the item of interest from the beach of Big Data is twice removed from any feedback potential. Secondly, as I have written before, the Drone Aestheticist is exactly that, a gleaner of interesting images for use on their social image site, which in itself is a bit of an abject exercise.
Or is it? For example, if one is to say that the Aestheticist gleaning the images does so without intent or politics, and is merely operating on fetish/interest value, then this is perhaps one of the least interesting practices in New Aesthetic practice. But on the other hand, if one looks at the work of practitioners like Jordan Crandall, Trevor Paglen, or Ricardo Dominguez, who examine the acquired image as instrument of aggression, control, and oppression, this puts a new lease on the life of the Drone Aesthetic. In a way, though inquiry, there is an indirect feedback loop established in questioning the gaze of the device, its presence, and its function in its theater of operations. The politics of the New Aesthetic emerges here, in asking what mechanisms of command and control guide the machine eye and determine its targets of interest. This is of utmost importance, as the abstracted eye is guided without subjectivity or ethics and is determined solely by the parameters of its algorithms and the stated goals of its functions.
Is the aesthetic of the machine image merely a function of examining its processes, fetishizing its errors, or something else? The criteria of the New Aesthetic attempts to talk about a spectrum of digital imaging that stretches back into time far longer than 2010, and has a problematically broad sense of definition. Once these problems are set aside as a given, one of the key criteria for the evaluation of NA practice and the function of its images depends upon the degree of control and autonomy inherent in the process within the creation of the image. This is formed in a continuum of control and abstraction from Algorism and Generative Art to autonomous eyes like drones and satellites. Algorism is one of the oldest NA practices, and exhibits the closest relationship between artist, machine and determinacy of digital process. A greater degree of indeterminacy is evident in the Glitch, but the iterative process of tweaking the media and then setting it forth into the process of interpretation by the codec, foregrounds the issue of digital autonomy.
The eye of the unmanned platform abstracts creation from the human organism at least once if a human does not operate it remotely, and twice if it is. There is the Terminator-like fear of the autonomous robot, but at this time, perhaps the more salient questions regarding what I have qualified as drone/autonomous aestheticism under NA of what the function of the image is, and is it really that interesting? Are the practices of NA blurring artistic and curatorial practice into a conceptual aestheticism, creating a cool detachment from the image despite its source or method of creation? Is the bottom line to the genres of NA the degree of control that the artist or aestheticist has over the image’s creation or its modality/intent? It seems that NA is an ongoing reflection upon the continuum of control over the generation of the image, our beliefs regarding its aesthetics, and what the intentions or politics are behind the creation of the New Aesthetic image. Or, as I have written before, are we just pinning images from Big Data and saying, “Isn’t that kinda cool?”
Maybe it’s somewhere in the middle of intention and cool.