::fibreculture:: Data Politics and Automated Aesthetics – Workshop

[Ned Rossiter]

Data Politics and Automated Aesthetics  – Workshop
Digital Life research program
Institute for Culture and Society, Western Sydney University
Friday, 25 August 2017

Venue: Female Orphan School, Conference room 3.
Parramatta South campus

Organized by Liam Magee, Luke Munn and Ned Rossiter

Data, despite its claim to ‘pure’ technicity, is permeated all the way
through with the political. From the ‘ground truth’ of machine learning
to the manipulations of employee tracking, data systems and
architectures are infused with the extra-rational. This is not a party
politics, but a politics of small things with scalar potential. Design
decisions are embedded in our technical systems and digital devices.
These low-level decisions stack up to form asymmetries of power,
increasingly determining how architecture and our environments are
arranged (Airbnb), the social and labour relations that constitute work
(Uber), the imagery and information we consume (Facebook) and the ways
in which we produce and locate knowledge (Google). Politics here is not
spectacular and overt, nor aligned exclusively with the modern authority
of the sovereign state. Rather, data politics disappears as background
performance within operational structures and systems. In this sense
data politics can be considered protological (Galloway), or a
‘microphysics’ (Foucault) of power. To critically understand this
politics, we need new methods and analytical devices unconstrained by
traditional disciplinary boundaries.

Automation produces new forms of aesthetics. Think, for instance, of the
personalised social media feed as an automatically constructed object –
a particular arrangement of updates and info, viewpoints and video
designed to receive the highest engagement and the most likes. Or take,
for example, machine vision as a set of aesthetic practices that
understand the world as an algorithmic topology from which trackable
objects, readable numbers and parseable text might be extracted. As
Rancière reminds us, aesthetics are never purely about beauty or style,
but rather inherently political in that they determine who gets to speak
and who remains silent, what is visualised and what remains invisible.
Beyond the logic of representation is the politics of operation
(Mezzadra and Neilson).

Data politics and automated aesthetics determine futures. From
speculative fiction to phantasmic visualisations, the future is
materialised through imaginaries generated as much out of the
guru-driven enclaves of  Silicon Valley as the startup buzz of China and
its factories of gold farmers. The artist studio and the scientific
laboratory are replicated in silica in the data centre. This cultural
epoch will be one that integrates mechanical reproduction into wider
circuits of  machinic production, where novel visual, aural and sensory
effects are produced through data mining, patterning and optimisation.
These imaginaries forecast the termination of the human capacity to
intervene. But these fantasies never completely exhaust the
possibilities of automation, nor fully enclose constitutive outsides.
How might theory and practice, criticality and invention come together
to contour the forces of technological change?

This workshop probes aesthetic practices as a handle on algorithmic
regimes of governance and automated worlds.

Program
10-10.15am
Welcome and introductory remarks

10.15am-12.15pm
Professor Anna Munster, National Institute for Experimental Arts,
University of New South Wales
‘DeepAesthetics’

Professor Glenn Geers, Principal Engineer, Australian Road Research Board
‘Automated Vehicles and Future Urban Form’

12.15-1.00pm
Lunch

1.00-2.15  Tech workshop, Liam Magee and Luke Munn
Image Analysis with Processing: Approaching Media with Algorithmic Eyes
This short hands-on workshop introduces the basics of algorithmic image
generation using the popular scripting tool Processing. Participants
will load images, examine their colours, and loop over their pixels,
playing with these simple parameters in order to explore some of the
fundamental logics underpinning digital aesthetics.

2.15-2.30pm
Coffee/tea

Dr Josh Harle, Director, Tactical Space Lab, Sydney
‘Digital capture and the Aesthetics of Verisimilitude’

2.30-4.30pm
Dr Debbie Symons, Melbourne based artist
‘Communicating via the Database’

4.30-4.45pm closing remarks

This event is free. Please rsvp by 22 August: https://tinyurl.com/y7fx63yv

Framing Questions
What kind of politics emerge from data practices? How are the
conventional categories of race, sex and class reworked into new
classifiers generative of new political forces?

What is the trajectory of this politics? Automation, for example, is
deeply ingrained with an historical process of optimising labour,
significantly altering the conditions and experience of work.

How is this politics made operational? Can we empirically examine data
within machinic and computational settings in order to make legible the
techniques used, the architectures employed and the strategies that are
actioned?

How might aesthetics assist with the analysis and understanding of these
processes, which are often imperceptible or purposefully made inaccessible?

How might data and automation be instrumentalised by alternate logics,
intervening in the given and instead pointing towards new theories, new
methods and new aesthetics?

Abstracts and Participant Biographies
Glenn Geers
Abstract
‘Automated Vehicles and Future Urban Form’
During the late 19th century the advent of the railways induced changes
to urban form which has only accelerated with the rise of the
automobile. As we enter the 21st century we are at another cusp in
transport with automated vehicles set to influence almost every aspect
of how we live, how we interact with friends and family; and how we do
business. However, as history has shown new transport technology will
influence the way cities and towns are designed. In this talk I will
present some ideas about the way driverless vehicle technology may
influence urban form and whether the results will be as rosy as some
have suggested.

Bio
Glenn Geers is Principal Engineer, ITS at ARRB – Australia’s largest,
independent transport research organisation. He is an Adjunct Associate
Professor in the School of Computer Science and Engineering at the
University of New South Wales, and is a member of Society of Industrial
and Applied Mathematics and Association of Computing Machinery. He is a
member of the Scientific Advisory Committee (SAC) tasked with providing
advice on the current research strategy and future scientific directions
for the SMART Infrastructure Facility, University of Wollongong. Glenn
holds honours degrees in electrical engineering and theoretical physics
and received his PhD in the field of computational electromagnetism. He
is on the editorial board of GeoInformatica. From 1994 to 2005 Glenn
worked on biometrics, image processing and distributed systems at CSIRO
and in private industry. In 2005 Glenn joined National ICT Australia
(NICTA) as Systems Engineering Manager, Intelligent Transport Systems.
>From 2010 to 2015 he held the role of Technology Director,
Infrastructure, Transport and Logistics at NICTA. During the same period
he was a Director of ITS Australia.

Josh Harle
Abstract
‘Digital capture and the Aesthetics of Verisimilitude’
How do digital technologies attempt to make sense of the world? My
recent work Making Sense re-appropriates digital capture technology to
parody the motivations and ideal of the ‘objective observer’. As a
continuation of this research, I explore photogrammetry both from a
technical standpoint and in its contemporary use by Australian
archaeologists. This contextual discussion is broadened, examining the
opacity of the automated reconstruction process, its questionable claim
to verisimilitude, and the culturally sited standard for representing
sites. To conclude, I present Surrounds and Relics – two works which use
and subvert this process to create a very different aesthetic. I discuss
the importance of a fictocritical approach and the presentation of
broken results and conspicuous artifacts to reveal how alien the
algorithm’s understanding of the world is.

Bio
Josh Harle is a multidisciplinary researcher and new media artist with a
background in Computer Science, Philosophy, and Sculpture. Harle’s
practice explores the contemporary use of digital technologies to map
and make sense of the world, critiquing the opaquely ideological
practice of digital capture. His works take various established and
emerging mapping technologies – laser scanning, photogrammetry,
geolocation tracking – and re-appropriates them as expressive mediums,
altering their outcomes to introduce an affective element which is
normally absent.  The results embrace their performative origins and the
contingency of their creation. Through these radical cartographic
practices, Harle reveals his own place in a field of competing drives to
organise, stake-claims, and dictate boundaries: the map as a performance
of exploration, of trying to make sense of the world.

Anna Munster
Abstract
‘DeepAesthetics’
In the context of machine learning, the term ‘deep’ refers to the
addition of (potentially) thousands of layers to a neural network,
providing a training AI with opportunities for increased precision in
solving a particular task. But ‘deep’ must then imply the capacity to
run such networks. Both the computational and economic facilities
underpinning such processing have shifted to corporations capable of
servicing and financing massive parallel networks: in 2012, Google’s
research labs ran 10 million YouTube cat videos through neural networks
using Graphic Processing Units (GPUs) running parallel architectures
across hundreds of computers. Deep learning was thus ‘born’ in a
convergence of: i) a material infrastructural shift in computation –
from CPUs to GPUs; ii) the consolidation of networked corporatism and
‘intelligence’ research; and iii) the literal capture and harvesting of
contemporary visual culture by data centres.

This paper argues that ‘DeepAesthetics’ names this new kind of machinic
assemblage, and will propose that such an assemblage must be understood
transversally as the relationality of technical, political and cultural
claims to the production of automated forms of judgement and value. I
will ask: how have existing aesthetic paradigms been subsumed by
manufacturers of, for example, GPUs, such as Nividia? But the paper will
also consider what alternative aesthetic practices might yet be possible
as AIs themselves experiment with other kinds of deep aesthetics; ones
which operate out of machine learning’s own black boxing. We might
speculatively propose that the largely impenetrable operativity of such
vast neural networking might be the site of a radical unknowability and
affectivity for machine aesthetics.

Bio
Anna Munster is an artist, writer, educator and professor in art and
design, University of New South Wales. She is the author of An Aesthesia
of Networks (MIT Press 2013) and Materializing New Media (Dartmouth
University Press, 2006). Both of these examine aspects of artists’
engagements with networks and digital culture. Anna is also an artist,
regularly collaborating with Michele Barker. She has worked most
recently on the installations evasion (2014), and HokusPokus (2011)
using soundscapes, interaction and installation design to explore both
human and nonhuman movement and perception. She regularly contributes to
journals, writing on art, media, politics and culture and is a founding
member of the online peer-reviewed journal The Fibreculture Journal. Her
co-edited anthology, Immediations: Art, Media, Event, with Erin Manning
and Bodil Marie Stavning Thomsen will be published with Open Humanities
Press in 2017.

Debbie Symons
Abstract
‘Communicating via the Database’
This presentation dives into three recent artworks that explore the
destructive intersection of capital with the natural environment. World
Species Market co-opts a stock market board and replaces the customary
financial data with threatened species data. Counting One to Four:
Nature morte visualises the predicted consequences of our warming
atmosphere on the entirety of the Earth’s biodiversity through the use
of percentage formulas. The final work, Arrivals and Departures
interrogates the inextricable links between the demise of numerous
Australian native species impacted by the arrival of four species; the
fox, rabbit, cane toad and the Indian myna bird. By reusing the visual
language of financial and logistical systems, the catastrophic effects
of consumption and capital are made implicit. But reusing these
aesthetics also suggests new systems that might work to sustain, rather
than suspend forms of life within our fragile biosphere.

Bio
Dr Debbie Symons is a multi-disciplinary artist based in Melbourne,
Australia. She completed her PhD Anthropocentrism, Endangered Species
and the Environmental Dilemma at Monash University in 2014 with the
support of an Australian Postgraduate Scholarship. Symons collaborates
with scientific organisations such as the IUCN Red List to facilitate
the statistical data pertaining to her works. Symons’ works have shown
internationally; The International Urban Screen Association, The
Streaming Museum, New York and Galerie Prodromus, Paris. Locally,
Symons’ has exhibited her works at [MARS] Gallery, Craft Victoria,
Linden New Art, RMIT Gallery, Latrobe Regional Gallery. Her works are
held in a number of private collections within Australia.  She recently
completed the inaugural Carlton Connect residency for the City of
Melbourne as part of the CLIMARTE Art+Climate=Change 2015 Festival and
exhibited in ARTCOP21, the artistic response to the COP21 talks in
Paris, New York and Melbourne.

Suggested Readings
Jafarpour, Sina, Gungor Polatkan, Eugene Brevdo, Shannon Hughes, Andrei
Brasoveanu, and Ingrid Daubechies. ‘Stylistic Analysis of Paintings
Using Wavelets and Machine Learning’, Signal Processing Conference, 2009
17th European, IEEE, 2009, 1220-1224,
http://web.math.princeton.edu/ipai/dating.pdf

Maillardet’s automaton, https://www.fi.edu/history-resources/automaton

Machine Learning for Musicians and Artists,
https://www.kadenze.com/courses/machine-learning-for-musicians-and-artists/info

Mezzadra, Sandro and Neilson, Brett. ‘Extraction, Logistics, Finance:
Global Crisis and the Politics of Operations’, Radical Philosophy 178
(2013): 8-18.
https://www.radicalphilosophy.com/article/extraction-logistics-finance

Paglen, Trevor. ‘Operational Images’, e-flux (2014),
http://www.e-flux.com/journal/59/61130/operational-images/

PwC. ‘Managing the people and change aspects of implementing Robotic
Process Automation (RPA) in the workforce’, February 2016.
https://www.pwc.com.au/pdf/robotic-process-automation-people-change-and-robots.pdf


Srnicek, Nick. ‘Computational Infrastructures and Aesthetics’, in
Christoph Cox, Jenny Jaskey and Suhail Malik (eds.), Realism Materialism
Art (Berlin: Sternberg Press, 2015), 307-12.
https://s3.amazonaws.com/arena-attachments/799654/8ec57874e10c085956319ee899a42be2.pdf