Surveillance @ 4S Conference, San Diego

The 4S Conference in San Diego is slowly coming up, and there are some panels which are very interesting for surveillance scholars. Our friend Göde organizes a session on machine learning, which might be interesting for those working on algorithmic surveillance. Then, there's another session explicitely dedicated to surveillance and big data, organized by Torin Monahan and Anders Albrechtslund. The deadlines unfortunately have already passed, but whoever is in the area might be interested in joining the discussion.

The conference will take place in San Diego, Oct 9--12. Here's the conference website and the other open sessions. And here are the abstracts on machine learning and surveillance & big data:

Machine Learning Worlds: Politics and Practices

Organizers: Shreeharsh Kelkar (MIT); Göde Both (TU Braunschweig)

Machine learning (ML) is so pervasive today that you probably use applications based on it many times without knowing it. It enables search engines, spam detectors, video tracking systems, self-driving cars, automated trading, and credit card fraud detection. ML technologies help to settle disputes in sports, achieve situational awareness in robotics, and pick the right novel for your reading pleasure. Economists Brynjolfsson and McAfee (2011) suggest that the next few decades will see rapid advancements in the power of machines; the nature and political economy of work will thus be irrevocably transformed. They cite as evidence the recent advances in machine translation (e.g. Google) and software agents (e.g. “Watson” ). These tools are the result of ML practices involving large amounts of Big Data and computing power.

This panel is comprised of empirical and theoretical contributions that deepen our understanding of the politics and values of ML practices. We welcome, but not exclusively, discussions of questions like:

• How does ML research exist along with imaginations of people, societies, work and workplaces?
• How have different fields within and outside computer science -- language translation, computer vision, computational neuroscience and genomics, theoretical physics -- used ML?
• How does the work of ML “theoreticians” differ from those who practice "applied" ML?
• Classifying is a political act (Bowker/Star 1999). As classifiers, what kind of values and assumptions are inscribed into ML systems?
• What kinds of realities do these systems enact?
• Contrary to what the term “machine learning” suggests, doing ML actually requires a lot of “human intuition” . The system's designer must specify how the data is to be represented and the mechanisms used to model the data. How can we grasp practices such as calibrating sensors, tinkering with parameters, and adjusting models?
• What methodological innovations would be needed to study ML practices empirically?

Surveillance and the Mediation of Big Data

4S session(s) organized by Torin Monahan and Anders Albrechtslund

The "big data" paradigm signals an intensification and distribution of algorithmic surveillance across multiple organizational and geographical scales. More than an exponential advancement in storage and processing capacity, big data currently operates as a fluid metaphor for the potential of data analytics to intelligently predict and respond to the needs of individuals and institutions. Clearly STS inquiry could fruitfully deconstruct the technological deterministic slant of discourses surrounding big data so that attention could be drawn to the values being inscribed in algorithms, the profound materiality of cloud computing, the control dimensions of pervasive software, and the active cultivation of new subjectivities as people come to understand themselves through their data doubles. Surveillance is key to these processes, as the capture and processing of data is frequently oriented toward some form of intervention or control. Rather than viewing surveillance through big data as completely automated or neutral processes, this panel seeks to investigate the many forms of mediation and politics inherent in big-data applications.

Possible areas of inquiry might include:

• Data fusion, profiling, and prediction by security organizations.
• The crafting of new subjectivities as individuals embrace “quantified self” movements.
• The social and political effects of “filter bubbles” erected by various search platforms.
• Gamification of interaction with customers and clients as public and private organizations seek to capitalize on (and control) user involvement.
• Activist and civil-society harnessing of data repositories and sensing devices to achieve progressive outcomes.
• The optimization of urban infrastructures through “smart” information technologies.
• Health technologies used for documentation, analyses, predictions and recommendations.