Finding the heart in algorithmic art

 

P-197 (1977) -- Manfred Mohr

In the late 1960s, the German artist/musician Manfred Mohr was one of the pioneering digital artists making very unique art using algorithms. He is still making the same kind of art today and is exemplary of how to control and transform algorithmic processes into fine art.

For some artists, algorithms are essentially variety generators or a way to transform data into art objects and/or art experiences. Some use the results as-is (as in Glitch Art) and others use them as inspiration for works done with traditional art materials.

The viewer might not be able to tell the difference in the final product initially, but may ultimately gravitate towards the paintings made through manual processes, or that possess certain human qualities or essences, that are more 'smooth' in texture--as opposed to the literal representation of the glitched look of computer-generated art. The meanings inherent in digital art are equally 'pixelated' (as opposed to having a more 'analog' context with a higher resolution or more interesting cultural context), upon which the artist can manually rework and appropriate.

Roy Lichtenstein did exactly this with the paintings of Rouen Cathedral by Monet, using a dot screen suggested by a mechanical process, then going back and painting them on a canvas using a perforated stencil.

Lichtenstein's modus operandi of parroting the look of screen printing is not unlike what digital artists are doing with pixels. The pixel is the new ben-day dot.

But even without modern technology (and its meta-experiments) new perspectives on art have always established New Normals, leaving the Old Normals as a touchstone or reference point to the past. The impetus to make art is the same, even though its shape and form continually change. If the things we do have some deference to the distant past, it follows that we are somehow influenced by it.

Modern creative practices are different in that the connection with the materials can be far removed from the human touch--for example in systems designed to generate the work cybernetically, such as Brian Eno's Music For Airports. Ultimately it is irrelevant to art as being 'made' etymologically speaking. It doesn't matter how it all happens; for some artists the stochastic is as good as anything, and reworking it is not always necessary to produce a result, although it introduces the probability of the 'happy accident'. The more creativity becomes mechanized the more we should want control over it, and adjust it accordingly based upon how much we want algorithms to influence creativity.

Digital artist Wade Guyton does something like this where he orchestrates accident into the printing process. This was one of the more compelling aspects of the Andy Warhol silkscreens and Gerhard Richter's scraped paintings--where there was always some kind of accident or spillage that appeared on the surface that could be uniquely compelling. It is conceivable that algorithms could have a similar slippage or spillage to keep spontaneity in the creative process, ideally imposed by humans, although a glitch factor could be programmed in, such as shutting down an ink cartridge at random, or wire it to some other data stream like wind speed.

The so-called New Aesthetic with the pixel as one of its main symbols (along with technologist/artist James Bridle's Render Ghosts and the Utah Teapot among other things), has been called 'gaudy' by art critics, even though it is not an art movement per se. In the wisdom of Manfred Mohr, it is easy to dismiss this criticism: it is essentially the result of the process that matters, even if it is largely philosophical in nature, and is ugly or has no aesthetic value. In the 21st Century, 'philosophical' can include banal things like Seapunk . It is the power of the meme that makes it relevant, even if it is not particularly interesting.

Another project by James Bridle, A Ship Adrift, consists of the plotting of wind speed and wind direction data collected from a weather station over a map. Without an explanation, it looks like a route of some kind created by movements on a Quija board. It's not visually interesting on its own and has no aesthetic value, yet works nicely as conceptual art (or 'theory object' in New Aesthetic parlance). What Bridle seems to be interested in is the aerial view of data: you're not just seeing a plotted line on a flat surface, you are virtually in the air looking down at it. It purports to represent an actual path based on a data stream but is rather pushed around by various social network feeds, as a kind of 'cyberspace cybernetics'. What makes the New Aesthetic so interesting is its exploration of the boundary between man, machine and network as a platform for making socially-relevant conceptual and performance art.

Another interesting angle on algorithmic art is a project by Benjamin Grosser, titled Computers Watching Movies, which consists of the temporal and spatial plotting of the audio tracks from popular films. The resultant works in fact do have some aesthetic value, as a kind abstract-expressionist take on algorithmic processes. It is an abstraction of audio and not just the mechanical plotting of data points.

Big ideas

Thinking about the ideas is the most difficult aspect of creativity. There are instances where ideas can arise spontaneously from the happy accidents, but to operate on accident as a primary way of working is somewhat rudderless, although it is a way of surrendering to the moment in order to produce an object or to bring a process or problem to a conclusion. Artists are the teleological manipulators of conclusions and results: it is what separates a natural system and an artificial one. The beauty of the latter is that both human and machine occupy that category as distinct from the natural world. All art is artifice in that equation.

The role of the artist

Working completely in the digital realm begs the question of whether we are merely technicians rather than artists. Algorithms can provide the raw material for making artistic choices, but they are not ends in themselves. Ideally they should not be the result of making a copy of a template, making a few tweaks and calling it a day, although it is entirely possible that this could produce an interesting piece of art. This is certainly an issue in music composition, whereby the instrument itself, the gear or the software presets mask the presence of an idea. Even if using plain rudiments, a central idea may not always exist, in either art or science.

In many ways digital art has some kind of permanent connection to a pre-made template (e.g. a sample of an audio file) which may in fact contain the ideas, that are subsequently appropriated. The negative side-effect of an automated process such as this is that everything can become derivative, and can call into question whether variations on an idea can be interpreted as infringing on an existing work, or as deemed a true derivative work. Nowadays derivation is almost expected. In terms of art history, it has always been used to some degree, and couched as homage, inspiration or emulation. There are also practical reasons for data copies, as they streamline data storage: all the copies that we save to our hard drives, and forward in e-mails require more and more server space and the energy to keep them running. If all the derivative works pointed to one master source file it would make networking much more efficient. Using this as a metaphor, creativity would always involve Master Files. A century ago artists would follow their Masters. In the future the Masters may be digital files.

Whose idea?

What would happen if idea generation was done primarily by computers and humans no longer had Ideas, or felt that machines could generate them more efficiently? How would we then define an 'idea'? Like never having the time or desire to cook for oneself, preparing meals (having ideas) would be something that we could be disabused of. (This has already happened with agriculture; food is now understood as being something that is the result of a rarefied process--from factory farm to restaurant table.) Of course we will still be having ideas, but computers may have already had them and are working on them. Asking questions are also as crucial as ideas, and computers will be good at this.

In a human-machine collaboration, how do we determine who had an idea first? If a computer creates a great idea, would a human want to claim that they had the idea if they simply did the programming? The answer is a qualified Yes. If the artist designs the seeds for something, all the 'plants' that grow from the seeds (even if slightly different) are all the artist's ideas. But does the artist know beforehand what will be ultimately generated, or will some of the results be unexpected? If the predictions are accurate, to what degree can we attribute the work to machines? It's a matter of how we impute invention.

As with anthropomorphism, human-like attributes are merely our projection of them on machines. In a collaborative environment the source ideas and seeds become fungible, and to give creative credit to a computer or simply to chance is merely an act of self-deprecation. John Cage's legacy is everything he didn't do intentionally; it was the seed of the idea that was planted in our understanding of consciousness. In terms of artists being influenced by other artists, computers will never get any credit (unless of course we anthropomorphize them and give them names.)

It remains to be seen how the synergy between human-machine will spin out and who/what will cede control. But there is no doubt machines will become another extension of consciousness, just as other objects have in establishing a feedback loop for our mutual benefit. In some respects, we have moved on from McLuhan, now that mediums change so quickly. Messages are now less intrinsically connected to media. TV once made certain messages possible, but the bond was broken by the evolution from analog to digital, which allowed messages to be endlessly transformed and remixed.

The challenge will always be how to achieve the transcendent qualities in art. I think humans will always excel at this, but machines may also be capable of having a ‘heart’. Then the question will become how it affects our own artistic impulse.

Here is a quote from an essay by Jeanette Winterson that sums it up nicely:

"What art does is to coax us away from the mechanical and towards the miraculous. The so-called uselessness of art is a clue to its transforming power. Art is not part of the machine. Art asks us to think differently, see differently, hear differently, and ultimately to act differently, which is why art has moral force. Ruskin was right, though for the wrong reasons, when he talked about art as a moral force. Art is not about good behaviour, when did you last see a miracle behave well? Art makes us better people because it asks for our full humanity, and humanity is, or should be, the polar opposite of the merely mechanical. We are not part of the machine either, but we have forgotten that. Art is memory — which is quite different [from] history. Art asks that we remember who we are, and usually that asking has to come as provocation — which is why art breaks the rules and the taboos, and at the same time is a moral force." [Read the entire essay here]