Research Fellows and graduate students based here at MCC and at the Law School's Information Law Institute will host Governing Algorithms: A Conference on Automation, Computation, and Control on May 16-17. The faculty organizer is Professor Helen Nissenbaum, who holds appointments at both MCC and the NYU School of Law, where she is Director of the Information Law Institute.
On the eve of what promises to be a fascinating series of talks and conversations, two of the organizers, Sophie Hood and Malte Ziewitz, answered some of our questions about the event:
You highlight the fact that the definition of algorithm varies widely in the current scholarship; as you point out, it has been described as a technology, an epistemology, a mode of social ordering, a sociotechnical process. In hosting this conference are you hoping to arrive at a more precise definition?
Sophie: We hope to draw out the various definitions of algorithm, and the implications of the myriad use of the term.
I think we’re more interested in exploring a multiplicity of definitions that arriving at a singular one.
Malte: There is a weird dynamic at the heart of current talk about algorithms. On the one hand, they are increasingly invoked as powerful entities that govern, sort, shape, manage, regulate, select, and control everything from news media to financial markets. On the other hand, it is not really clear how “they” do all this. In fact, you can often substitute the word “algorithm” with others like “technology,” “machine,” or even "god” without changing the meaning considerably. So one important challenge is not so much to arrive at an ultimate definition of the term, but to understand what kind of work a word like “algorithm” does in different contexts.
Why do you think the algorithm has captured the imagination of such large and diverse segments of society?
Malte: Yes, there has been lots of interest in algorithms recently, but it is not really a new concept. In fact, computer scientists will tell you that many of today’s algorithms actually predate the era of “big data.” What might be new is that the concept captures something people feel increasingly strongly about. Namely that there is something happening that escapes human control and, increasingly so, understanding.
Sophie: In some respects, the current obsession with all things algorithmic was the natural next step after the buzz around “big data,” “analytics,” or “web 2.0.”
I am curious about your choice of the word "Governing" for the conference title. Are algorithms assigned a governing role, do they assume one, or does some outside entity govern algorithms?
Sophie: We’d go with answer “d” — all of the above. Do we govern algorithms, or do they govern us, in some sense? We don’t understand these as mutually exclusive.
Malte: Another advantage of “governing” is a very pragmatic one: it is broad enough to speak to different audiences. A legal scholar can relate to governance, but so can a sociologist or media scholar. At the same time, it indicates a shared concern with automation and control in all kinds of technological systems.
Tell us about the scholars who will present at the conference. What fields do they represent? How will their particular expertise or background shape some of the conversations you expect to take place?
Sophie: The scholars represent a wide variety of fields, which we hope will spark a rich interdisciplinary conversation. Media studies, Science and Technology Studies, information law, computer science, information systems, philosophy, history, sociology, anthropology — the field is really broad. Panel topics include, among other things, algorithmic financial trading, how algorithms are relevant to online speech, and how we should use algorithms to inform video surveillance programs.
Malte: There will be some really interesting talks at the conference. For example, I am very much looking forward to the opening talks on Thursday. Robert Tarjan, who is a living legend among computer scientists, will talk about “Algorithms as a Computer Scientist Sees Them.” And Claudia Perlich, an award-winning data scientist, will speak about her experience and applications in the digital advertising world. As you’ll see from the program, it’s a great mix of speakers.
The conference organizers bridge a media department and law institute at NYU. Why these two disciplines? How is this conference an extension of the questions your research team is pursuing?
Sophie: Algorithms are of interest to both media and law scholars, but in different ways. Given the breadth and complexity of this topic, we found it helpful to include people who think across disciplinary boundaries.
Malte: A good example of this kind of multidisciplinary work is the Privacy Research Group at NYU, which has long brought together students, professors, and industry professionals who are passionate about understanding privacy in the digital age. Another case in point is our co-organizer Solon Barocas, who is currently finishing up his dissertation on “Data Mining: Ethics, Ethos, and Episteme.” These are the kinds of topics that benefit vastly from a multidisciplinary perspective.