The Cultural Life of Machine Learning

This book brings together the work of historians and sociologists with perspectives from media studies, communication studies, cultural studies, and information studies to address the origins, practices, and possible futures of contemporary machine learning. From its foundations in 1950s and 1960s pattern recognition and neural network research to the modern-day social and technological dramas of DeepMind's AlphaGo, predictive political forecasting, and the governmentality of extractive logistics, machine learning has become controversial precisely because of its increased embeddedness and agency in our everyday lives. How can we disentangle the history of machine learning from conventional histories of artificial intelligence? How can machinic agents' capacity for novelty be theorized? Can reform initiatives for fairness and equity in AI and machine learning be realized, or are they doomed to cooptation and failure? And just what kind of "e;learning"e; does machine learning truly represent? We empirically address these questions and more to provide a baseline for future research.Chapter 2 is available open access under a Creative Commons Attribution 4.0 International License via link.springer.com.

This book brings together the work of historians and sociologists with perspectives from media studies, communication studies, cultural studies, and information studies to address the origins, practices, and possible futures of contemporary machine learning. From its foundations in 1950s and 1960s pattern recognition and neural network research to the modern-day social and technological dramas of DeepMind’s AlphaGo, predictive political forecasting, and the governmentality of extractive logistics, machine learning has become controversial precisely because of its increased embeddedness and agency in our everyday lives. How can we disentangle the history of machine learning from conventional histories of artificial intelligence? How can machinic agents’ capacity for novelty be theorized? Can reform initiatives for fairness and equity in AI and machine learning be realized, or are they doomed to cooptation and failure? And just what kind of “learning” does machine learning truly represent? We empirically address these questions and more to provide a baseline for future research.

1. Toward an End-to-End Sociology of 21st-Century Machine Learning.- 2. Mechanized Significance and Machine Learning: Why it Became Thinkable and Preferable to Teach Machines to Judge the World.- 3. What Kind of Learning Is Machine Learning?.- 4. TheOther Cambridge Analytics: Early “Artificial Intelligence” in American Political Science.- 5. Machinic Encounters: A Relational Approach to the Sociology of AI.- 6. AlphaGo’s Deep Play: Technological Breakthrough as Social Drama.- 7. Adversariality in Machine Learning Systems: On Neural Networks and the Limits of Knowledge.- 8. Planetary Intelligence.- 9. Critical Perspectives on Governance Mechanisms for AI/ML Systems.