Travelogue: Australian Forum Panel Discussion

The focus of my panel discussion centers on Artificial Life, as encoded in three-dimensional computer generated space (3D space), with a specific orientation on the view into the Artificial Life virtual “world”; this view is predominantly mediated through the virtual camera.

Author(s)

Introduction

“A rich history has prepared the way for Artificial Life to make sense” writes Stefan Helmreich, [1 pp.6] reminding us the various dimensions through which to consider the Artificial Life are many. The discursive frameworks that tend to frame Artificial Life include; biology, the genealogy of cybernetics, the history of twentieth century generative art, often economics [2] and, more recently in relation to Islamic Art. [3, 4] It is not without irony then that Artificial Life visualisation itself as a specific dimension of the moving image is a somewhat impoverished discursive field with few notable exceptions such as. [5, 6] Reflecting the state of discursive impoverishment is the number of Artificial Life artworks that specifically explore Artificial Life in relation to the grammar of the moving image; Technosphere (1995) by Jane Prophet et al and Nemirovsky’s et al’s Emonic Environment (2001-2005) are but two exceptions.

How the view into Artificial Life ‘world’ frames our perception of the virtual world is the concern of this presentation.  Reframing Helmreich I argue, a rich history has prepared the way for us to view Artificial Life images; there are many visual grammars and interpretative regimes that inform our capacity to engage with Artificial Life virtual worlds. In a recent response to an abstract I submitted to a media arts history conference I was surprised to read one reviewer’s response; “The idea of the "virtual camera" in simulated settings is an interesting hold-over from the filmic world and deserves more exploration”.  The response was unexpected for a number of reasons, which I address here. Firstly and prosaically, the virtual camera is simply a de-facto protocol in all types of 3D simulation, filmic or not; it is the tool [7] or device through which to frame the image, window or view into 3D modeling software, VR simulation, FPS games, architectural visualisation, engineering simulation or, the artificial life virtual “world”.  Secondly, I argue there is a ubiquitous regime and protocol in Artificial Life image making that draws from both science and cinema even if these normative practices are not explicitly obvious to the practitioners creating Artificial Life images. Finally, I argue Artificial Life image making draws heavily upon the interpretative grammars and strategies developed in the scientific nature film. In summary, this paper outlines provisional research that situates Artificial Life visualisation, science and art, in both; the interrelated genealogies of scientific visualisation and cinema and, within a particular discursive orientation traced to Disney animation and nature films.

Science and the ‘long take’

Whilst much research energy has focused on computational techniques to generate lifelike behavior and emergence (Langton) the scopic regime through which to view artificial life “worlds” helps gives rise to the key themes of emergence and “lifelike behavior”; there are interpretive fields through which to view Artificial Life worlds. The scopic regime in Artificial Life visualisation is a “hangover” from scientific observation, vis-à-vis devices such as the microscope and telescope, and from film/cinema, vis-à-vis Andre Bazin’s device “the long take”; the perception of an uninterrupted view of the world underwrites both the arts of reality [8] and Bazin’s long take. Scientific objectivity and the long take function to ‘record’ an unmediated reality; this reality gives rise to the idea that one looks through a window into a ‘world’, in this context through a window into an Artificial Life world and not and at an image of Artificial Life (albeit there are exceptions).

The idea behind the virtual camera is embedded in the genealogy of analogue devices such as the microscope and telescope’; the virtual camera impassively enframes the ‘world’ whilst it simultaneously optimizes the credibility or factuality of the ‘world’. Moreover, similar to a photochemical camera (still or motion) the virtual camera ‘records’ or documents a temporal image of the ‘world’; in other words the virtual camera approximates Vertov’s “microscope or telescope of time.” [9 pp.213]

There are a number of ideas at work here worth considering. Firstly, the camera is employed objectively; as we know objectivity forms a tactical and an interpretive regime.

Secondly, similar to analogue telescopes and microscopes, the virtual camera operates in and is operated on in the ‘present tense’; Jenna Ng is instructive here observing that the long take also functions not just to record reality but presentness; Ng writes, “Pier Paolo Pasolini attributes presentness specifically to the long take as it is the shot which takes in the greatest amount of reality, and "reality seen and heard as it happens is always in the present tense": "the long take, the schematic and primordial element of cinema, is thus in the present tense. Cinema reproduces the present” [10 pp.133-134] or in the words of the film Director Aleksandr Sukorov's in describing his film Russian Ark (2002) “the present continuous.” [Sukurov in 10 pp.123] This present continuous has special resonance in Artificial Life image making; with few exceptions the key to Artificial life’s emergence [11] involves evolving live to a “global audience” (Scott Draves http://electricsheep.org/).

Thirdly, the virtual camera is routinely considered as a window into a ‘world’. Unlike analogue telescopic, microscopic and, photochemical images which are indexical to the physical world, in the case of photography and film they record imprints of light from the world, [10 , 12] Artificial Life ‘worlds’ are not indexical to the physical world, they are isomorphic to a specific computer model. Moreover, the computational models are mathematical expressions that interpret the physical world; they are neither indexical nor isomorphic to the physical world. Like all digital images Artificial Life images don’t ‘represent’ the physical world, digital images encode information and “computers produce tokens of numbers. [12 pp.131] That “Images are mediations between the world and human beings” [5 pp.9] is a important reminder that an image is not a window into a world it is an image. This point is critical when framed against the very premise of artificial life, which is predicated on “generating lifelike behavior [… and] focuses on the problem of creating behavior generators.” [11 pp.5] Frequently, the success Artificial Life visualisation is dependent on observing or deciphering emergent patterns in the ‘world’; what is perceived in the world or on the screen is what there is to perceive.  According to Flusser transposing the act of looking at an image into the act of looking at or into a world is “dangerous” business:

What one sees on them [technical images and by extension images created by the virtual camera] therefore does not appear to be symbols that one has to decode but symptoms of the world through which, even indirectly, it is to be perceived. This objective character of technical images leads whoever looks at them to see them not as images but as windows. Observers thus do not believe them as they do their own eyes. Consequently they do not criticize them as images but as ways of looking at the world … [this] lack of criticism of technical images is potentially dangerous at a time when technical images are in the process of displacing texts – dangerous for the reason that the ‘objectivity’ of technical images is an illusion. For they are – like all images – not only symbolic but represent even more abstract complexes of symbols than traditional images. They are metacodes of texts which … signify texts, not the world out there. [5 pp.15]

The problem with technical images, from art or science, is what you see is not what there is, what there are, are highly sophisticated models and concepts that require decoding and the virtual camera is strategically organized to impassively enframe these encoded models.

This account of the virtual camera in Artificial Life has particular resonance with the status of photography and film, from science and art, in the late nineteenth and early twentieth century; “Photographic images, according to the skeptic, were the automatic product of a machine, not of a mind” [13 pp.8] that is, the photographic machine “automatically [reproduced] whatever [found it’s] way in front of the camera lens.” [13 pp.9] The skeptical accounts were later confirmed in the actualités or documentary films of the Lumière Brothers [13 pp.9] and exquisitely represented in the photographic work of Karl Blossfeldt, a pioneer of the “New Objectivity” movement in the early 20th century. [14 pp.5-22] As is apparent in the images created by Blossfeldt, the framing of the ‘world’ is a tactical account of ‘reality’, an account that does not automatically record the natural landscape but frames it through various apparatus. Blossfeldt’s formal grammar is often manifest in artificial life image making, for example in Karl Sim’s Galápagos (1997) series and in Jon McCormack’s elegant study of the “computational sublime” in Bloom (2006) see http://www.csse.monash.edu.au/~jonmc/projects/Bloom/Bloom/bloom_main_page.html.

Disney Animal Animation and the Nature Film

A common feature in Artificial Life research, art and science, is the accompanying publication describing the research. The published material often includes both a description of the target system and a stylized fictional description of the system; the word “world” is such a stylized, rhetorical, descriptor. The beautifully evocative description of the homeostatic simulation Daisyworld (1983) suffices as a demonstration; published in the scientific journal Tellus Series B: Chemical and Physical Meteorology, the account reads; “Owing to a subtle change of climate, clouds appear on daisyworld. The clouds are light in colour. We will assume that the clouds form only over stands of black daisies because of the rising air generated over these warm spots.” [15] Obviously, stylized descriptions have properties that the models don’t. [16] These stylized accounts of the ‘worlds’ that describe ‘natural’ systems are nature stories, stories that would not appear out of place in David Attenborough’s Life series (1979-2010) nor in Disney nature stories.

The fictive re-imaginings of Artificial Life ‘worlds’ have much in common with the nature or wilderness stories popularized by Disney in the 1950s and 1960s and other production units in the 1970s, which retold stories of a mythic “western interior” as natural ‘progression’ and, anthropomorphized the subject matter. [17 pp.117-120] The similarities between the Disney stories and artificial life narratives, in how they both organize ways of thinking about the world through retelling of ‘eternal’ stories of birth, death, cycle, pattern and adaptation (progression), are many. Compare, for example, the role of the 1950s Disney filmmaker and Artificial Life practitioner; the filmmaker captures or “shoots” something on film and the Artificial Life practitioner captures life in the computational ‘world’ using the virtual camera, both are on the metaphorical hunt for the ‘novel yet familiar’ exotic life form in an ‘undiscovered land’. And similar to the Disney stories that did “something far more than reveal “nature’s mysteries”: [… and] spoke to us of a living and intelligible world beyond the fence of civilization, a world we could enter at will and experience in something like human time” [17 pp.118] Artificial Life proposes a comparable arrangement.

Whilst the reader might find the relationship between the Disney nature film of the 1950s and artificial life unpersuasive consider the 1993 commissioned report by Ars Electronica in which the author of the report, Roy Ascott, details the demarcated exhibition spaces at Ars Electronica. In a provocative gesture “Walt Disney, animal animations” are included as ‘artists’ in the “Artificial Life” section of the Ars Electronica exhibition. [18 pp.296] My reading of this provocation is twofold and interrelated; Disney animations and nature films attempt to simulate the ‘laws of nature’ - Disney animation “follows the laws of physics — unless it is funnier otherwise" observed Disney animator Art Babbitt whilst Disney nature films attempt to simulate life as we know it vis-à-vis the moral and political refractions of life as it is and life as it could be. Disneynature, the division of Walt Disney Motion Pictures Group that specifically releases Disney nature documentaries, renders explicit this interrelation between Disney animation and Disney nature film. Disneynature’s latest blockbuster African Cats (2011) is “An epic true story set against the backdrop of one of the wildest places on Earth, "African Cats" captures the real-life love, humor and determination of the majestic kings of the savanna [...] the story features Mara, an endearing lion cub who strives to grow up with her mother's strength, spirit and wisdom [...] Fang, a proud leader of the pride who must defend his family from a rival lion and his sons. Disneynature brings "The Lion King" to life in this True Life Adventure... An awe inspiring adventure blending family bonds with the power and cunning of the wild”. (http://disney.go.com/disneynature/africancats/#story/

Artificial life ‘world building’ is located in this trajectory of nature storytelling; cyberbeasts, virtual organisms and, agents are optimized in a similar vein to Mara, Fang and other ‘big cats’ to fight, breed, and die; moreover we observe them as they carry out the real life drama on the fitness landscape. This relationship between artificial life and the nature film is a theme repeated in artificial life researcher Terzopoulos’s conceptual framing of his artificial life work, for example in his paper Artificial Life for Computer Graphics Terzopoulos writes “computer animators can begin to play a role less like that of graphical model puppeteers and more like that of (National Geographic Society) nature cinematographers.” [19]

Artist and writer Michael Punt writes, “the gap between science and entertainment is much smaller than imagined. The interaction and exchange might occur not only in shared technologies but in the very imagination that seems necessary to negotiate our consciousness of the world as complex and ultimately unknowable. Science in narrative cinema and the cinematic in scientific research do seem to function reciprocally to help account for difficult things, providing images, metaphors, and useful descriptions for each other.” [20] Punt’s argument is clearly illustrated in Jon McCormack’s Turbulence (1995). Described by the artist as “a menagerie of synthesised forms, evolved within the computer using a process of artificial selection. A video laserdisc contains over 30 minutes of computer generated animation” [21] Turbulence is a series of pre-rendered animated sequences that are selected and viewed in no particular order. Whilst McCormack wrote the code to “allow certain algorithms and their graphic progeny to flourish through a recursive process of digital "procreation," and to terminate others—a case of survival of the aesthetically fittest” [22] the overall experience is expressed through the conventions of cinema as the user/viewer watches the pre-rendered movie files; the camera shots are composed by the artist director and; the narrative arc, regardless of the viewing order, reworks the story of Genesis creation or a “post-Fantasia reprise.” [22] In the words of Michelle Barker “In the end what we see is closer to a narrative cinematic experience complete with sound track.” [23]

Closing Remarks

In Artificial Life image making, a constellation of grammars from science and film are assembled as constituent of a particular ‘natural order’. Whilst this ‘natural order’ appears benign it is important to remember the ‘natural order’ in Artificial Life is neither neutral nor impartial; the capacities required to engage and interact with the metaphors, models and techniques of artificial life are cultural and political, they are recruited, so to speak, into Hayles’ posthuman world [24] in which the machine becomes the model for understanding the human.

At stake in Artificial Life screen based ‘worlds’ or images is agency both in the terms of the Artificial Life-form and being human; instead of looking at animats, virtual pets or “cyberbeasts” perhaps we should consider looking through the Artificial Life forms point of view. Virilio’s somber observation; “Once we are definitively removed from the realm of direct or indirect observation of synthetic images created by the machine for the machine instrumental virtual images will be for us the equivalent of what a foreigner’s mental pictures already represent: an enigma” [25] and Cubitts astute observation, “Machine perception and human perception are co-dependant and must co-evolve” [26 pp.108] are nice bookends to this topic.

“AI Life” researcher Margaret Boden argues, “An idea can be “possible” or “impossible” only with respect to a specific conceptual space. It is possible if the rules for generating new structures allow for it; impossible if they do not. The more clearly we can map the conceptual space, the better we can identify a given idea as creative, in this way or that.” [27 pp.269] If I understand Boden correctly she argues that an idea (as creative) is possible within a specific “interpretive regime”, an idea transmitted through a specific dispositif vis-à-vis structures of knowledge, discourse and power. It is through the institutionalised apparatus that I understand Cubitt when he states “Computers will talk to anyone, but only the wealthy teach them to speak, to define what perception might be and what is interesting.” [26 pp.47] The irony is this; in a contemporaneous media saturated landscape in which the “new space of mediated vision is post-Cartesian, postperspectival, postcinematic, and posttelevisual,” [28 pp.7] artificial life screen based work tends to orient around a single stationary view into the virtual ‘world’.

Acknowledgments

The author would like to thank the Australia Council for the Arts, ANAT (Australian Network for Art and Technology), and the curatorial team Vince Dziekan, Paul Thomas and Sean Cubitt for their kind support.

This research is supported by the Australian Research Council's Discovery Projects funding scheme (project number DP0772667).

References and Notes: 
  1. S. Helmreich, Silicon Second Nature Culturing: Artificial Life in a Digital World (Berkeley: University of California Press, 1998).
  2. J. Stapleton, 2004, "Black Shoals: A Meditation on Cosmology, Artificial Life and the Aesthetics of Political Economy,"  http://www.blackshoals.net/textpages/JamieText.html (accessed September 2011).
  3. L. U. Marks, Enfoldment and infinity : an Islamic genealogy of new media art (Cambridge, MA: The MIT Press, 2010).
  4. C. Alexander, A foreshadowing of 21st century art: the color and geometry of very early Turkish carpets (New York: Oxford University Press, 1993).
  5. V. Flusser, Towards a philosophy of photography (London: Reaktion, 2000).
  6. L. Wiesing, Artificial presence : philosophical studies in image theory (Stanford, Calif.: Stanford University Press, 2010).
  7. C. Paul, Digital Art, 2nd ed. (London; New York: Thames & Hudson, 2008).
  8. L. Mulvey, Death 24x a second : stillness and the moving image (London: Reaktion Books, 2006).
  9. N. Carroll, Theorizing the moving image (New York: Cambridge University Press, 1996).
  10. P.-S. Ng, “Mutations of Pastness: Time, Cinema, Ontology,” University College London, London, 2009.
  11. C. G. Langton, Artificial life:  the proceedings of an Interdisciplinary Workshop on the Synthesis and Simulation of Living Systems, held September, 1987, in Los Alamos, New Mexico (Redwood City, Calif.: Addison-Wesley Pub. Co., 1989).
  12. D. N. Rodowick, The Virutal Life of Film (London: Harvard University Press, 2007).