Seeing mobility: how software engineers produce unequal representations
Pedro Sanches · Engineering Studies · 2016
Data produced by mobile networks are frequently presented as useful to understanding social phenomena related to human behavior. The risks associated with making use of massive datasets are often framed in terms of privacy, security, intellectual property, or liability. I show that the risks of mobile data sensing are not reducible to privacy, but are also related to how software engineers produce representations of individuals or populations. I characterize software engineering as a practice of producing abstractions and categorizations by articulating discursive and material elements. These may contribute to rendering certain groups of individuals invisible. If unaddressed, these risks can have consequences at least as serious as privacy violations.