The Challenge of the Black Box: Technosocial Predictive Analytics (TPA) and the Real World

The term "black box" traditionally refers to understanding systems in terms of input and output characteristics. In Technosocial Predictive Analytics, the term most often is applied to models, since their specific internal workings either are unknown or need not be considered. This paper examines the consequences of approaching the real world itself as a black box, and how some current verification and validation challenges may be simplified thereby.