We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CY

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computers and Society

Title: The Objective Function: Science and Society in the Age of Machine Intelligence

Authors: Emanuel Moss
Abstract: Machine intelligence, or the use of complex computational and statistical practices to make predictions and classifications based on data representations of phenomena, has been applied to domains as disparate as criminal justice, commerce, medicine, media and the arts, mechanical engineering, among others. How has machine intelligence become able to glide so freely across, and to make such waves for, these domains? In this dissertation, I take up that question by ethnographically engaging with how the authority of machine learning has been constructed such that it can influence so many domains, and I investigate what the consequences are of it being able to do so. By examining the workplace practices of the applied machine learning researchers who produce machine intelligence, those they work with, and the artifacts they produce. The dissertation begins by arguing that machine intelligence proceeds from a naive form of empiricism with ties to positivist intellectual traditions of the 17th and 18th centuries. This naive empiricism eschews other forms of knowledge and theory formation in order for applied machine learning researchers to enact data performances that bring objects of analysis into existence as entities capable of being subjected to machine intelligence. By data performances, I mean generative enactments which bring into existence that which machine intelligence purports to analyze or describe. The enactment of data performances is analyzed as an agential cut into a representational field that produces both stable claims about the world and the interpretive frame in which those claims can hold true. The dissertation also examines how machine intelligence depends upon a range of accommodations from other institutions and organizations, from data collection and processing to organizational commitments to support the work of applied machine learning researchers.
Subjects: Computers and Society (cs.CY)
Cite as: arXiv:2209.10418 [cs.CY]
  (or arXiv:2209.10418v1 [cs.CY] for this version)

Submission history

From: Emanuel Moss PhD [view email]
[v1] Wed, 21 Sep 2022 15:05:54 GMT (3637kb)

Link back to: arXiv, form interface, contact.