[nexa] Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability - Dec 13, 2016