Algorithmic rationality

Epistemology and efficiency in the data sciences

Bibliographic Data

ID5260679
AuthorsIan Lowrie (0000-0002-7216-6849, Rice University, corresponding author)
Year2017
Volume4
Issue1
Pages205395171770092
Publication date2017-06-01
Peer ReviewedYes
Open AccessYes
TypeARTICLE
VenueBig Data & Society (JOURNAL)
Journal identifiersISSN: 2053-9517 • E-ISSN: 2053-9517
PublisherSAGE Publications Inc (PUBLISHER)
DOI10.1177/2053951717700925
OpenAlexW2600290815
LanguageEN
Citations received36
References cited39

Recently, philosophers and social scientists have turned their attention to the epistemological shifts provoked in established sciences by their incorporation of big data techniques. There has been less focus on the forms of epistemology proper to the investigation of algorithms themselves, understood as scientific objects in their own right. This article, based upon 12 months of ethnographic fieldwork with Russian data scientists, addresses this lack through an investigation of the specific forms of epistemic attention paid to algorithms by data scientists. On the one hand, algorithms are unlike other mathematical objects in that they are not subject to disputation through deductive proof. On the other hand, unlike concrete things in the world such as particles or organisms, algorithms cannot be installed as the objects of experimental systems directly. They can only be evaluated in their functioning as components of extended computational assemblages; on their own, they are inert. As a consequence, the epistemological coding proper to this evaluation does not turn on truth and falsehood but rather on the efficiency of a given algorithmic assemblage. This article suggests that understanding the forms of algorithmic rationality employed in such inquiry is crucial for charting the place of data science within the contemporary academy and knowledge economy more generally

Ethics and Social Impacts of AI · Information Systems Theories and Implementation

Tools

Open DOISci-HubOpen Access