Richard Groß // Situating Machine Learning. Cooperative Pattern Recognition and Calibrated Problems
Table of contents
Contact
Fellow
NameRichard Groß
Send encrypted mail via the SecureMail portal (for TUD external users only).
Visiting address:
Z21, Room 216b Zellescher Weg 21
01217 Dresden
Räume 205, 206, 216 + 217, jeweils a+b
Postal address:
TUD Dresden University of Technology Schaufler Kolleg@TU Dresden
01062 Dresden
Research project: Situating Machine Learning. Cooperative Pattern Recognition and Calibrated Problems
Discipline: Sociology
Doctoral advisors: Prof. Dr. Dominik Schrage and Juniorprofessorin Dr. Susann Wagenknecht
Machine learning has proven to a challenging object of study for sociological research, both in theoretical and methodological terms. As an applied technology proliferating across many domains of social life, it provokes social theory especially in terms of its social capacities. Linguistic communication, for instance, no longer seems to be a human privilege, and algorithmic modeling affects how knowledge is generated and disseminated.
Research on machine learning in the humanities offers a variety of assessments of the subject that vary considerably in their analytical verdicts. They range from enthusiastic speculations about untapped more-than-human potentials to sober observations of the mediocrity of the algorithmic meaning modelling to emphatic warnings about the dangerous effects of the unregulated real-world use of machine learning technologies. Picking up on these current controversies, I approach machine learning empirically and explore it as a situated practice in my dissertation project.
Based on ethnographic field work, I research current applications of machine learning in science and art by means of two case studies. I approach machine learning by studying situations of its practical realization. Inspired by pragmatist philosopher John Dewey, I understand situations as problematic episodes of practices that interrupt routines. Resolving situations requires creative solutions.
Within situations, technology – however smart or intelligent – does not appear as an isolated entity but as an integrated and codependent element of practices. In this sense, I do not conceive of machine learning as a material computational artifact or an autonomous individual agent but use the term to denote successful situational cooperation in practical efforts of machine learning. My analyses focus on how problems are identified and resolved in the interaction of different heterogeneous machine learners. Following Adrian Mackenzie's definition, I understand machine learners as a variety of social entities, human as well as non-human (including devices, programs, data, graphs, algorithms, tables, and many more), that are involved in machine learning.
Situating machine learning ethnographically, my research offers empirical insights through a problem-centered perspective on so-called artificial intelligence as practice. As a pragmatist contribution to current debates in social theory, my dissertation spells out various issues of situated cooperation to characterize machine learning as a social technology.
CV
From 2018 | Edition of Complete Works of Arnold Gehlen (Project Coordinator), TU Dresden |
2018 | Diploma degree (M.A.) in Sociology, Art History and Musicology, TU Dresden |
2017/2018 | Study Visit, New School for Social Research, The New School, New York |
2013 - 2017 | Research and Teaching Assistant, Institute of Sociology, TU Dresden |