BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//7.2.3.1//EN TZID:Europe/Berlin X-WR-TIMEZONE:Europe/Berlin BEGIN:VEVENT UID:695@hiig.de DTSTART;TZID=Europe/Berlin:20190515T130000 DTEND;TZID=Europe/Berlin:20190515T140000 DTSTAMP:20190513T143555Z URL:https://www.hiig.de/en/events/critical-ai-studies-the-epistemic-ensemb le-of-deep-learning-lunch-talks/ SUMMARY:Critical AI Studies & The Epistemic Ensemble of Deep Learning | Lu nch Talks DESCRIPTION:\nProf. Jonathan Roberge and Prof. Michael Castelle are guest r esearchers at the Alexander von Humboldt Institute for Internet and Societ y in May. In their joint Brown Bag Lunch\, they argue for the need for Cri tical AI Studies and ask what epistemic categories are needed to understan d Machine Learning. The lectures will be held in English. If you are inter ested\, please register using the form below.\n\n \;\nLunch talk with Jonathan Roberge &\; Michael Castelle\nWednesday\, 15 May 2019 · 1 pm · HIIG Kitchen\n \;\n\nJonathan Roberge\nCritical AI Studies. Making the Case for CAIS\nAutomated technologies populating today’s online worl d rely on social expectations about how “smart” they appear to be. Alg orithmic processing\, as well as bias and missteps in the course of their development\, all come to shape a cultural realm that in turn determines  what they come to be about. It is our contention that a robust analytical frame could be derived from a culturally-driven STS while focusing on Cal lon’s concept of translation. Excitement and apprehensions must find a s pecific language to move past a state of latency. Translations are thus c ontextual and highly performative\, transforming justifications into legit imate claims\, translators into discursive entrepreneurs\, and power rela tions into new forms of governance and governmentality. In this presentati on\, we discuss three cases in which AI was deciphered to the public: i) the Montreal Declaration for a Responsible Development of Artificial Intel ligence\, held as a prime example of how stakeholders manage to establish the terms of the debate on ethical AI while avoiding substantive commitm ent\; ii) Mark Zuckerberg’s 2018 congressional hearing\, where he constr ued machine learning as the solution to the many problems the platform mi ght encounter\; and iii) the normative renegotiations surrounding the grad ual introduction of “killer robots” in military engagements. Of inter est are not only the rational arguments put forward\, but also the rhetori cal maneuvers deployed. Through the examination of the ramifications of t hese translations\, we intend to show how they are constructed in face of and in relation to forms of criticisms\, thus revealing the highly cybern etic deployment of AI technologies.\nJonathan Roberge is Associate Profess or of Cultural and Urban Sociology at the Institut National de la Recherch e Scientifique\, where he also holds the Canada Research Chair in Digital Culture. He is among the very first scholars in North America to have cri tically focused on the production of algorithms\, a research agenda which culminated into a foundational text in this domain entitled Algorithmic  Cultures (Routledge\, 2016\, translated into German at Transcript Verlag\, 2017). He currently works on a manuscript entitled The Cultural Life of Machine Learning to be out early in 2020 at Palgrave MacMillan (together w ith Micheal Castelle).\n\nMichael Castelle\nExperiment\, Vector\, and Los s: The Epistemic Ensemble of Deep Learning\nThe fast-growing research fie ld of deep learning — the use of convolutional and/or recurrent neura l network architectures in machine learning — has been hailed as a “r evolution” by researchers and practitioners\, one which is sometimes co nsidered “unreasonably effective” and associated with a “black art ” of internalist knowledge. At the same time\, these models have been c riticized by social scientists for their interpretative opacity\, depend ence on classification schemes\, and capacity to reproduce social biases. Which methodologies are most appropriate for understanding these techniq ues and their increased deployment in everyday sociotechnical life? In t his presentation\, by focusing on three distinctive features of deep lear ning—its experimental method\; its vectorial or ‘structuralist’ on tology\; and the role of the ‘loss’ function through which model para meters are optimized—I will argue that a combination of historical\, e pistemological\, semiotic and interactional approaches are necessary for understanding deep learning. This allows one to understand this emergent field not as a revolutionary disruption but as a genre of technoscience which synthesizes aspects of past epistemic breaks—in this case behavio rism\, cognitivism\, structuralism\, connectionism\, and machine learning (as well as aspects of game theory and cybernetics)—together into what  I call an epistemic ensemble. This resulting perspective can permit ri cher engagements with these techniques on the part of social scientists a nd humanists\, who can thus draw on a rich historiography of previous cr oss-disciplinary engagements and\, I argue\, actually apply their own exi sting theoretical apparatuses to contribute to this unstable field for so me future ‘neural’ social sciences and/or humanities.\n \;\nMicha el Castelle is Assistant Professor at University of Warwick’s Centre f or Interdisciplinary Methodologies. His work is at the intersection of t he sociohistorical studies of science and technology and the economic soc iology of markets\, and is working to draw connections between present-d ay research developments in AI (such as generative network architectures and attention models) and 20th-century theories of language\, learning\, and creativity. He has written for the journals Philosophy &\; Techn ology\, Economy and Society\, Computational Culture\, and has presented at  EMNLP (Empirical Methods in Natural Language Processing) and SIGCHI (Sp ecial Interest Group on Computer-Human Interaction). He holds degrees in both Sociology (Ph.D.\, University of Chicago) and Computer Science (Sc. B.\, Brown University) as well as professional experience in computer gra phics\, computational neuroscience\, and neurology. ATTACH;FMTTYPE=image/jpeg:https://www.hiig.de/wp-content/uploads/2019/05/b ag-bags-brown-21436-Cropped.jpg CATEGORIES:Exploring Digital Spheres LOCATION:Humboldt Institute for Internet and Society\, Französische Straß e 9\, Berlin\, 10117\, Germany X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=Französische Straße 9\, B erlin\, 10117\, Germany;X-APPLE-RADIUS=100;X-TITLE=Humboldt Institute for Internet and Society:geo:0,0 END:VEVENT BEGIN:VTIMEZONE TZID:Europe/Berlin X-LIC-LOCATION:Europe/Berlin BEGIN:DAYLIGHT DTSTART:20190331T030000 TZOFFSETFROM:+0100 TZOFFSETTO:+0200 TZNAME:CEST END:DAYLIGHT END:VTIMEZONE END:VCALENDAR