Final catch-up post from Perth, this time focusing on the DAC conference. It was probably fortunate that most presenters ignored the theme - "The Future of Digital Media Culture" - and as a result there was a wide diversity of art and thought presented. You can read some other (more punctual) responses from Mary Flanagan and Kristy Dena among others. Here are a few of my impressions.
Fox Harrell's GRIOT system is a generative poetry/narrative engine modelled on African diasporic orature. Harrell's work crosses cultural, linguistic and formal/computational domains with impressive ease, and offers a reflective and constructive response to the philosophical critiques of computing culture presented by Phil Agre (for example). He shows how a generative model or ontology can be constructed reflectively or critically, without losing sight of the pleasure and poetics of generativity itself.
Su Ballard gave an interesting paper on the work of New Zealand artist Douglas Bagnall, who was news to me. In Bagnall's Film-Making Robot (2004) webcams mounted on on Wellington buses collect video, then upload it wirelessly to a central server, where the "robot" - neural-network software - analyses the composition of individual frames, classifying them on the basis of some seriously received aesthetic wisdom. Out of this tongue-in-cheek Modernist engine comes abject, jittering, mundane "films" that nonetheless reflect the compositional training of the robot. Bagnall's deadpan, AI-powered deconstruction of the project of aesthetics continues in Cloud Shape Classifier and the latest Mimetic Television, a device that "watches" soap operas and synthesises new video based on frame-difference statistics. Sort of like Jason Salavon's Everything, All at Once, but with an embedded artificial agency; see also Nicolas Baginsky's connectionist robotic musicians, The Three Sirens.
Other generative-flavoured presentations included Jason Lewis of Obx Labs, who presented his interactive / generative digital poetry projects and the NextText library, a Java library for real-time dynamic typography apparently coming soon to Processing. Karl D. D. Willis presented his elegant interactive environment Light Tracer, and TwelvePixels, a pixel-art app for mobile phones; his related paper on "open interactions" is also worth a look.
Keith Armstrong presented a manifesto for "grounded media," a media arts practice that seeks to respond to an ecological crisis "perpetuated by our sense of separation from the material and immaterial ecologies upon which we depend." Armstrong's work foregrounds our material commonalities as well as their articulation with and through digital mediation and representation. His 2007 work InStep illustrates this elegantly; a foot bandage embedded with soft sensors transduces walking into haptic impulses sent to a separate, hand-held sculptural form. Working in pairs, participants gain a sense of another's "imprint upon the ground." Perhaps Armstrong's philosophy addresses the question I put to Toxi recently about sustainability and generative design?
Finally, Simon Penny's paper "Experience and Abstraction" made one of the strongest critical statements of the conference. It continues Penny's sustained critique of the ideologies embedded in technology, focusing especially on its tendencies towards disembodiment, abstraction and generality. Penny sets these against a sense of art as embodied, concrete and specific, and questions the ability of art practice to effectively work against this technological grain. Penny's position is a bit glass-half-empty for me, but it stood as an important challenge to some of the more technophilic tendencies at PerthDAC.
Wednesday, October 03, 2007
Pieces of PerthDAC
Posted by Mitchell at 11:59 am
Labels: AI, conference, generative art, theory
No comments:
Post a Comment