Nettime mailing list archives

[Nettime-nl] Dream Politics: Randomness in Network Art
Lemmy Caution on Wed, 28 Jul 2004 21:07:06 +0200 (CEST)

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

[Nettime-nl] Dream Politics: Randomness in Network Art

Dream Politics: Randomness in Network ArtLewis LaCookhttp://www.lewislacook.com
INTRODUCTION: Stochastic Computing

Chaos comes before all principles of order & entropy, it's neither a god nor a maggot, its idiotic desires encompass & define every possible choreography, all meaningless aethers & phlogistons: its masks are crystallizations of its own facelessness, like clouds. 

Everything in nature is perfectly real including consciousness, there's absolutely nothing to worry about. Not only have the chains of the Law been broken, they never existed; demons never guarded the stars, the Empire never got started, Eros never grew a beard. 

No, listen, what happened was this: they lied to you, sold you ideas of good & evil, gave you distrust of your body & shame for your prophethood of chaos, invented words of disgust for your molecular love, mesmerized you with inattention, bored you with civilization & all its usurious emotions. 

There is no becoming, no revolution, no struggle, no path; already you're the monarch of your own skin--your inviolable freedom waits to be completed only by the love of other monarchs: a politics of dream, urgent as the blueness of sky. 

Hakim Bey, Chaos, The Temporary Autonomous Zone




Computers have a difficult time with spontaneity. By themselves, they're as predictable as any fundamentalist. This is what makes computer programming possible, the assurance I have that the code I write will be executed exactly as I wrote it. If I write a conditional loop, my computer will make a decision based on the parameters I feed to it; it won't take into account the weather, nor its own emotional state, nor will it ever be hung over from ten too many Guinesses and perform the function haphazardly from behind the haze of a violent headache. Computers, it would seem are very clean machines, not subject to noise or entropy. What you code is what you get.

This, naturally, hasn't stopped humans from introducing randomness into the computer. Most high-level programming languages have a function to simulate random numbers; and, while said numbers are very often predictable, the results can sometimes seem just as authentically random as more adroit sources of randomness. "True random numbers, captured in the wild, are clearly superior to those bred in captivity by pseudo-random generators—or at least that’s what the theory of randomness implies." Brian Hayes writes in his essay "Randomness as a Resource" ."But (one researcher) has run the output of various hardware and software generators through a series of statistical tests. The best of the pseudo-random generators earned excellent grades, but three hardware devices flunked. In other words, the fakes look more convincingly random than the real thing." 

Hayes is writing here of the two main methods of generating random numbers in modern computing: pseudo-random number generation, which is just what those rand() functions do, and the harnessing of external entropy sources, such as atmospheric noise or radioactive decay rates. Entropy is an index of the disorder or noise in a closed system; and, in physics, all systems are sliding inevitably toward disorder. Thus, linking certain parameters in your code to an external source of entropy is the most effective way of utilizing the ultimate disorder of the universe in your scripts. Pseudo-randomness, however, is essential for research purposes, where a sequence of random results must be repeated in order to provide a stable set for analysis. 

One can wonder why there would be a human desire to introduce randomness into computing. Take, for instance, the World Wide Web; instant global communication there, and wouldn't our ultimate preference be to clear this communication channel of all noise? It turns out, however, that randomness is a vital concept in the development of web networks, and much of the electronic economy would not exist without it. Ever buy anything online? Fill out one of those snoopy forms asking for potentially sensitive data such as your social security number or credit card number? If so, you more than likely want your privileged information to be secure, protected; viewable only by those involved in the transaction, if even then. Encryption is how one secures data over the HTTP protocols the World Wide Web is based on. Encryption performs character substitution on the data, which can then be decrypted (read into human-readable language) via a key file or function! , which was at one point randomly or
 pseudo-randomly generated. Encryption is a sign of trust between a web service and a consumer; and it depends heavily on randomness to ensure that bond.

To that end, providing sequences of dependably random integers has proven to be a crucial--and colorful--web service. The wildest example would be the Lavarand system, a random number generator developed at Silicon Graphics. Lavarand was a hybrid of the two methods of random number generation; it seeded pseudo-random functions with an external entropy source; in this case, data derived from the slow motion of blobs in Lava Lite lamps. Lavarand itself seems defunct; though trademarked by Silicon Graphics, the original project seems to have all but disappeared as a service, but Lavarnd (http://www.lavarnd.com/), a similar project derived from the methodology of the original project, continues. The Lavarnd API, downloadable in both C form and as a Perl library at the site, allows developers the freedom of replacing the original Lava Lite lamps with virtually any entropy so! urce; one feature is the ability to use simple web cams, such as the Logitech QuickCam.

A similar web service, random.org (http://random.org/), uses a radio tuned between stations to inject their data with true random flavor; Swiss Fourmilab offers HotBits (http://www.fourmilab.ch/hotbits/), which reads radiation via a Geiger-Muller tube detector.

Writing about randomness recently on the rhizome.org list-serv, American artist Jeremy Zilar (http://silencematters.com/)doesn't believe in randomness per se. "It is the dialog that we have with the process of observing of ourselves. The observed self, or the object, performance, process being created is a clear reflection of ourselves, and when we are able to gain that distance, we become more aware of what is going on inside, we make changes,. and then we correct the reflection to mirror ourselves once again....Randomness does not occur. It is a controlled element that somehow figures in to the image that we have of ourselves. Even when things do happen by chance, we immediately incorporate that action into the image and judge it's relationship to the whole, and juxtapose the whole to ourselves. If it doesn't fit, we remove it."

The distance Zilar writes about here is analogous to the "disinterestedness" that nineteenth-century German philosopher Immanual Kant proposed as a integral element in the aesthetic experience. "...a judgment on the beautiful which is tinged with the slightest interest, is very partial and not a pure judgment of taste," Kant wrote in his Critique of Judgment. "One must not be in the least prepossessed in favour of the real existence of the thing, but must preserve complete indifference in this respect, in order to play the part of judge in matters of taste." In other words, if a painting is figurative, and of a sunset, concerning ourselves with the beauty of the sunset and not of the painting itself is a bad judgment call. The painting is not the sunset. The sunset is not the painting.

It's this idea that nudged art toward the nonfigurative. If the contemplation of an art object is a disinterested contemplation, i.e. if we are to consider the painting isolated from what the painting represents, then why produce mimetic art at all? Why not simply delve into a "pure" painting, completely divorced from representation? In the late nineteenth and early twentieth centuries, this became a central theme in the narrative of artistic development: freeing the art object from representation, and breaking through to the realm of pure being. Art, once a way to represent the natural, now strove to create the natural; instead of painting a sunset, the artist wanted to paint an experience that had all the impact of the sunset, but was its own phenomenon. Or, as Remko Scha asserts: "Esthetically motivated art...faces a curious challenge: if it is created by humans, it will always be inferior to nature! In the course of the twentieth century, this challenge has b! een taken up by
 many artists. Some of them have suggested that they are in fact natural forces, beyond the ken of ordinary humans. Others have tried to withdraw from their artworks, by developing objective art-generating processes which they initiate without controlling the final result."

Or, as Hakim Bey, perhaps the most eloquent proponent of randomness in art, has reassured us: "Everything in nature is perfectly real including consciousness..."

INTRODUCTION: The Bacchanates

"Noise – or random data, or interference – has long been an obsession of digital artists. That obsession reflects the Nietzschean idea of a creative tension between the Apollonian and Dionysian," maintains Peter Carty in his review of C6's (http://c6.org/)NEST, or Network Examination of Seredipitous Transfer. "First outlined in The Birth of Tragedy, the idea is that Dionysus represents fundamental primal energy, while Apollo stands for rationality, logic and structure. Noise is unbounded dissonance; it is Dionysian. Information which is structured and rendered directly meaningful by IT protocols is Apollonian." 

NEST (http://c6.org/nest/)is a peer-to-peer client with a twist. Unlike Morpheus or Kazaa, NESTers have access to only one file; a single audio file, corrupted by each pass through each user's computer. Instead of using TCP, C6 uses the notoriously unreliable UDP (User Datagram)protocol to hand data off from one client to another. UDP is unreliable because it performs very little error-recovery on the data passed through it; the experience is, as C6 themselves put it, "...much like the children's game of ‘Chinese Whispers’, where each client is linked to their closest geographical neighbour. Passing a ‘virtual whisper’ around the internet, each link in the chain can create new versions with each imperfect cycle."

Network art like this is ripe with entropy. There's just no telling what will happen to that audio file as it passes from me to you; but chances are, noise will distort it until it's no longer familiar to us. It's not "interactive" in the way most web and browser-based art has been; there are no Flash rollover buttons, no net video; its meat is the network itself, a network designed to incorporate flaws into its very hide. C6 style themselves as "conceptual marketers" (perhaps in keeping with the vogue of artists appropriating and aping corporate behaviors, and also, as are all such gestures perhaps, tongue-in-cheek), but what they've done here is more conceptual than marketable (fortunately!). NEST is unstable-network-as-aesthetic-experience.

Since random.org uses a radio tuned between stations as their entropy source, I often wonder what it would be like if they tuned in to rand()%(http://www.r4nd.org/). This net radio station is a randomness-hound's wet dream: all of the audio is composed at random in realtime, every time you tune in. Named for the ANSI C rand function, the audio programming is a collage of art-coders, including Lia and Carvalhais, Muio.org, Karlheinz Stockhausen (adapted for prime randomness by Georg Hajdu), and Pix. rand()% was developed by Tom Betts and Joe Gilmore as a commission by Media Centre Network of Huddersfield, England.

Listening to rand()%, one might often believe that one's computer is crashing. As with much random and chance art, you either enjoy it or you don't. Helen Valery Jamieson, in a recent post to the Netbehaviour list, expressed frustration with the flaws in randomly-generated artwork. She confessed: "I am not a big fan of randomly generated art; the concept might be interesting but I get bored by it fairly quickly. On the other hand, random elements within a work can be really inspiring. and there's computer random & human random - audience interaction with a programme or with a computer-mediated performance. It's random to the extent that you don't know what the audience is going to come up with, but it's within certain parameters. structured randomness perhaps." As rand()%'s only interaction with the audience is the act of tuning in, those less appreciative of random art might find the station annoying. There is, however, much to love in the program s! tream: pops, gurgles, grinds and
 static, peppered with actual notes in some cases, cascade through your PC's speakers when tuned to rand()%. When I was younger, I played in a lot of improvisational fusion bands, but after a few years found myself getting rotyally bored with the experience; it often seemed that all I was doing was going up and down scales in varying increments. rand()%'s artists seem to have solved that problem; but, while the works are composed anew with every listen, there can be an overwhelming sameness to the pieces, as if, in absence of hierarchical structuring, the works are so horizontal that they inspire no emotional interest. From an audience's perspective (and perhaps it's conditioning from so many years of absorbing more traditional 19th century narrative structures), the pieces may seem flat. 

Which begs the question: since random and chance works are relatively new developments in the history of art (John Cage, perhaps the most famous proponent of random compositional techniques, was, after all, a twentieth century figure), does work like this require a new kind of audience? Network art is even younger; remember, we couldn't even display images over the web until the mid-nineties. Works like NEST and rand()% are radical works in that they are, essentially, networks in themselves (radio was one of the first electric communication networks, along with the telephone system; imagine both of these works performed on analog networks; what if NEST was dependent on the snail mail system? It would resemble the Mail Art movement of the 80s, perhaps), and the concept of network-as-artform is so edgy an idea in 2004 that you might get a bloody nose just thinking about it. It will undoubtably take a decade or so before the aesthetics inherent in ! network art to leech into the
 mainstream, and perhaps longer than that for chance art to become commonplace. 

Until then; enjoy the entropy.






URLs and Works Cited

Bey, Hakim. The Temporary Autonomous Zone, Ontological Anarchy, Poetic Terrorism. Autonomedia Anti-copyright, 1985, 1991. http://www.hermetic.com/bey/taz_cont.html

Carty, Peter. "Deep Corruption on the Web." Metamute Web Exclusive: July 14, 2004. http://www.metamute.com/look/article.tpl?IdLanguage=1&IdPublication=1&NrIssue=24&NrSection=5&NrArticle=1432&ST_max=0

Hayes, Brian. "Randomness as A Resource," American Scientist, Volume 89, Number 4, July-August 2001, http://www.americanscientist.org/template/AssetDetail/assetid/20829;jsessionid=baadaKyCmrKiRl

Jamieson, Helen Valery. "Re: randomness." Post to Netbehaviour list-serv. July 2004.

Kant, Immanuel. Critique of Judgment, 1790, James Creed Meredith (translator), http://eserver.org/philosophy/kant/critique-of-judgment.txt 

Scha, Remko. "Readymades, Artificial Art, New Media", reprinted from Annette W. Balkema and Henk Slager (eds.): Exploding Aesthetics. L&B Series of Philosophy of Art and Art Theory, Vol. 16. Amsterdam & Atlanta: Rodopi, 2001, http://iaaa.nl/rs/Lier&Boog.html

Zilar, Jeremy."Re: randomness." Post to rhizome.org list-serv. July 2004, http://rhizome.org/



Lewis LaCook

net artist, poet, freelance web developer/programmer


XanaxPop:Mobile Poem Blog>> http://www.lewislacook.com/xanaxpop/

Stamen Pistol: http://stamenpistol.blogspot.com/


Sidereality: http://www.sidereality.com/



Do you Yahoo!?
Yahoo! Mail - Helps protect you from nasty viruses.


Lewis LaCook

net artist, poet, freelance web developer/programmer


XanaxPop:Mobile Poem Blog>> http://www.lewislacook.com/xanaxpop/

Stamen Pistol: http://stamenpistol.blogspot.com/


Sidereality: http://www.sidereality.com/




Do you Yahoo!?
Yahoo! Mail is new and improved - Check it out!
* Verspreid via nettime-nl. Commercieel gebruik niet
* toegestaan zonder toestemming. <nettime-nl> is een
* open en ongemodereerde mailinglist over net-kritiek.
* Meer info, archief & anderstalige edities:
* http://www.nettime.org/.
* Contact: Menno Grootveld (rabotnik {AT} xs4all.nl).