Algorithmic Identity: Networks, Data, and the terrible beauty of the black box

Annette N Markham, Holly Kruse, Jeff Hemsley, Molly Steenson

Abstract


_n_ $2 $2

: The distinct personality of an individual regarded as a persisting entity; individuality.

: The quality or condition of being the same as something else.

$2 $2 $2

:conceive of as united or associated;

:recognize or make a logical or causal connection

$2 $2 $2

: the attribution to yourself (consciously or unconsciously) of the characteristics of another person (or group of persons)

Computers, particularly those that seem to think on their own, have long fostered the perception that selfhood is as much about information as it is about individual bodies. In the last decade of the 20th Century, this idea was explored through the concept of virtuality.

In the first decade of the 21st Century, ubiquitous, mobile media interfaces replaced the computers that grounded our engagement in digital media. No longer sitting at desks or in front of devices to connect to the world through portals on our screens as virtual beings, we moved through complex networks of information flows, where the media we produced, consumed, and shared converged across a global range of stages. Now, we witness a third shift, sponsored in many ways by the infrastructures and algorithms that operate beneath the surface of interaction to co-construct self, identity, cultural categories, and meaning.

In this panel, we present papers that explore the idea of algorithmic identity. Across the four papers, we invoke the multiple notions of “identity,” “identify,” and “identification” to stress the complication of these concepts in contemporary social contexts. In each of our papers, we explore different terrain but agree: The ways we think about self and self-identity are complicated by how we are identified by the systems we use. Our interactions with algorithms both identify us and foster certain ways of identifying with others. We are increasingly defined by machine learning algorithms that sift through and aggregate the trace data we leave behind each time we post, visit or click within our networked virtual landscape. Yet most of us are largely unaware of the magnitude of data we generate or the algorithms that create an identity that marketers, governments and researchers assume reflects who we are. In our four papers, we discuss these points by talking about our research on our music habits, our DNA, the way we move through information spaces, and how researchers rely on algorithms that they may not understand to create categories within which we are located/situated, with or without our comprehension, control, or consent.

We believe this panel to be an interesting mix of cases that explore the idea of algorithmic identity. Three of the papers take the perspective of personal identity formation, focusing on the implication of our movement and interactions in data streams. The fourth paper illustrates the risks to knowledge production when researchers employ network visualizations as a sensemaking and data exploration tool without understanding the assumptions and design choices of the algorithms that create those visualizations. All four papers take a critical perspective on the terrible beauty of algorithms, as they shape and enable particular aspects of being and obscure or erase other possibilities.

Full Text:

PDF