My dissertation, The Algorithm at Work: The Reconfiguration of Work and Expertise in the Making of Similarity in Art Data, was an ethnography of a start-up firm commonly referred to as the 'Pandora for Art.' Organized around a similarity matching algorithm and a large, continuously changing database of art images, the technology provided an expert-oriented framework for end-users to discover their tastes and enter the emerging online market for contemporary art.


I studied the work practices of a team of art experts, who were hired to render art image data legible to the similarity matching algorithm and render the algorithm’s output legible to those with knowledge about art. Observing in-person and online interactions in the moments of technological breakdown allowed me to see the contention and negotiation involved in the ongoing, distributed repair of the algorithm by the art experts. Despite limited technical skills, the team was continuously mobilizing competing claims about how the algorithm worked in order to recognize and repair breakdowns between the algorithm's output and their expectations. As they experimented with different explanations and repairs, a novel form of expertise emerged that created opportunities for collaboration and conflict on the team.


My findings contribute to existing literatures on standardization, expertise, repair work, and teams, as well as the emerging literature on the sociology of data science. More importantly, the study raises questions about the undervalued skills increasingly required by the expanding market of contingent knowledge labor associated with data analytic technologies. Finally, my findings suggest that although we, as users, may know very little about "how the algorithm works," in practice we develop accounts that prove useful in context. These practices of accounting are central to understanding the implications of these technologies in work and market settings as well as in daily life. 

Journal articles

Sachs, S.E. 2019. The algorithm at work? Explanation and repair in the enactment of similarity in art data. Information, Communication, & Society.


Manuscripts in preparation


“Repair and the development of algorithmic expertise.”


“How knowledge workers manage the credibility of algorithms.” 

“Standardizing team practice through the creation and control of productive friction.” 

© 2018 by S.E. Sachs

S.E. Sachs   |   Cornell University, Department of Science & Technology Studies   |