Eurograd message

                ** Apologies for cross-posting **
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>The Cultural Life of Machine Learning: An Incursion into Critical AI Studies
<br>Preconference Workshop, #AoIR2018
<br>Montral, Canada
<br>Urbanisation Culture Socit Research Centre, INRS (Institut national de la
<br>recherche scientifique)
<br>Wednesday October 10th 2018
<br>
<br>Deadline for Abstracts: June 30th 2018  *Extended Deadline*: July 7th 2018
<br>
<br>
<br>Keynote: Orit Halpern (Department of Sociology and Anthropology, Concordia
<br>University)
<br>
<br>Machine learning (ML), deep neural networks, differentiable programming and
<br>related contemporary novelties in artificial intelligence (AI) are all leading
<br>to the development of an ambiguous yet efficient narrative promoting the
<br>dominance of a scientific fieldas well as a ubiquitous business model.
<br>Indeed, AI is very much in full hype mode. For its advocates, it represents a
<br>tsunami (Manning, 2015) or revolution (Sejnowski, 2018)terms indicative
<br>of a very performative and promotional, if not self-fulfilling, discourse. The
<br>question, then, is: how are the social sciences and humanities to dissect such
<br>a discourse and make sense of all its practical implications? So far, the
<br>literature on algorithms and algorithmic cultures has been keen to explore
<br>both their broad socio-economic, political and cultural repercussions, and the
<br>ways they relate to different disciplines, from sociology to communication and
<br>Internet studies. The crucial task ahead is understanding the specific ways by
<br>which the new challenges raised by ML and AI technologies affect this wider
<br>framework. This would imply not only closer collaboration among
<br>disciplinesincluding those of STS for instancebut also the development of
<br>new critical insights and perspectives. Thus a helpful and precise
<br>pre-conference workshop question could be: what is the best way to develop a
<br>fine-grained yet encompassing field under the name of Critical AI Studies? We
<br>propose to explore three regimes in which ML and 21st-century AI crystallize
<br>and come to justify their existence: (1) epistemology, (2) agency, and (3)
<br>governmentalityeach of which generates new challenges as well as new
<br>directions for inquiries.
<br>
<br>In terms of epistemology, it is important to recognize that ML and AI are
<br>situated forms of knowledge production, and thus worthy of empirical
<br>examination (Pinch and Bijker, 1987). At present, we only have internal
<br>accounts of the historical development of the machine learning field, which
<br>increasingly reproduce a teleological story of its rise (Rosenblatt, 1958) and
<br>fall (Minsky and Papert 1968; Vapnik 1998) and rise (Hinton 2006), concluding
<br>with the diverse if as-yet unproven applications of deep learning. Especially
<br>problematic in this regard is our understanding of how these techniques are
<br>increasingly hybridized with large-scale training datasets, specialized
<br>graphics-processing hardware, and algorithmic calculus. The rationale behind
<br>contemporary ML finds its expression in a very specific laboratory culture
<br>(Forsythe 1993), with a specific ethos or model of open science. Models
<br>trained on the largest datasets of private corporations are thus made freely
<br>available, and subsequently dtourned for the new AIs semiotic environs of
<br>image, speech, and textpromising to make the epistemically recalcitrant
<br>landscapes of unruly and unstructured data newly manageable.
<br>
<br>As the knowledge-production techniques of ML and AI move further into the
<br>fabric of everyday life, it creates a particularly new form of agency. Unlike
<br>the static, rule-based systems critiqued in a previous generation by Dreyfus
<br>(1972), modern AI models pragmatically unfold as a temporal flow of
<br>decontextualized classifications. What then does agency mean for machine
<br>learners (Mackenzie, 2017)? Performance in this particular case relates to the
<br>power of inferring and predicting outcomes (Burrell, 2016); new kinds of
<br>algorithmic control thus emerge at the junction of meaning-making and
<br>decision-making. The implications of this question are tangible, particularly
<br>as ML becomes more unsupervised and begins to impact on numerous aspects of
<br>daily life. Social media, for instance, are undergoing radical change, as
<br>insightful new actants come to populate the world: Echo translates your
<br>desires into Amazon purchases, and Facebook is now able to detect suicidal
<br>behaviours. In the general domain of work, too, these actants leave permanent
<br>tracesnot only on repetitive tasks, but on the broader intellectual
<br>responsibility.
<br>
<br>Last but not least, the final regime to explore in this preconference workshop
<br>is governmentality. The politics of ML and AI are still largely to be
<br>outlined, and the question of power for these techniques remains largely
<br>unexplored. Governmentality refers specifically to how a field is organisedby
<br>whom, for what purposes, and through which means and discourses (Foucault,
<br>1991). As stated above, ML and AI are based on a model of open science and
<br>innovation, in which public actorssuch as governments and universitiesare
<br>deeply implicated (Etzkowitz and Leydesdorff, 2000). One problem, however, is
<br>that while the algorithms themselves may be openly available, the datasets on
<br>which they rely for implementation are nothence the massive advantages for
<br>private actors such as Google or Facebook who control the data, as well as the
<br>economic resources to attract the brightest students in the field. But there
<br>is more: this same open innovation model makes possible the manufacture of
<br>military AI with little regulatory oversight, as is the case for China, whose
<br>government is currently helping to fuel an AI arms race (Simonite 2017). What
<br>alternatives or counter-powers could be imagined in these circumstances? Could
<br>ethical considerations stand alone without a proper and fully developed
<br>critical approach to ML and AI? This workshop will try to address these
<br>pressing and interconnected issues.
<br>
<br>We welcome all submissions which might profitably connect with one or more of
<br>these three categories of epistemology, agency, and governmentality; but we
<br>welcome other theoretically and/or empirically rich contributions.
<br>
<br>Interested scholars should submit proposal abstracts, of approximately 250
<br>words, by July 7th 2018 to CriticalAI2018 [at] gmail [dot] com. Proposals may
<br>represent works in progress, short position papers, or more developed
<br>research. The format of the workshop will focus on paper presentations and
<br>keynotes, with additional opportunities for group discussion and reflection.
<br>
<br>This preconference workshop will be held at the Urbanisation Culture Socit
<br>Research Centre of INRS (Institut national de la recherche scientifique). The
<br>Centre is located at 385 Sherbrooke St E, Montreal, QC, and is about a
<br>20-minute train ride from the Centre Sheraton on the STM Orange Line (enter at
<br>the Bonaventure stop, exit at Sherbrooke), or about a 30-minute walk along Rue
<br>Sherbrooke.
<br>
<br>For information on the AoIR (Association of Internet Researchers) conference,
<br>see https://aoir.org/aoir2018/ ; for other preconference workshops at AoIR
<br>2018, see https://aoir.org/aoir2018/preconfwrkshop/.
<br>
<br>Organizers: Jonathan Roberge (INRS), Michael Castelle (University of Warwick),
<br>and Thomas Crosbie (Royal Danish Defence College).
<br>_______________________________________________
<br>EASST's Eurograd mailing list
<br>Eurograd (at) lists.easst.net
<br>Unsubscribe or edit subscription options: http://lists.easst.net/listinfo.cgi/eurograd-easst.net
<br>
<br>Meet us via https://twitter.com/STSeasst
<br>
<br>Report abuses of this list to Eurograd-owner@lists.easst.net