COMPUTING AND AI: HUMANISTIC PERSPECTIVES FROM MIT
Science, Technology, and Society
Eden Medina and Dwaipayan Banerjee
Eden Medina, MIT Associate Professor of Science, Technology, and Society; photo by Allegra Boverman
"A more global view of computing would demonstrate a broader range of possibilities than one centered on the American experience, while also illuminating how computer systems can reflect and respond to different needs and systems. Such experiences can prove generative for thinking about the future of computing writ large."
— Eden Medina and Dwaipayan Banerjee, MIT Program in Science, Technology, and Society
Series | Computing and AI: Humanistic Perspectives from MIT
Eden Medina is an associate professor of Science, Technology, and Society at MIT. Her research uses technology as a means to understand historical processes, and she combines history, science and technology studies and Latin American studies in her writings. She is the author of Cybernetic Revolutionaries: Technology and Politics in Allende's Chile (MIT Press, 2011) and the co-editor of Beyond Imported Magic: Essays on Science, Technology and Society in Latin America (MIT Press, 2014).
Dwaipayan Banerjee is an assistant professor of Science, Technology, and Society at MIT. His research centers on the social contexts of science, technology, and medicine in the global south. He has two book projects now nearing completion: Enduring Cancer: Health and Everyday Life in Contemporary India (forthcoming in 2020, Duke University Press) and Hematologies: The Political Life of Blood in India (forthcoming in 2019, Cornell University Press; co-authored with J. Copeman).
. . .
Q: How might domain knowledge, perspectives, and methods from your field fruitfully shape the research and curriculum of the Schwarzman College of Computing?
Recent developments in fields such as data science, machine learning, and artificial intelligence (AI) have drawn attention to how computer systems shape and are shaped by understandings of justice, fairness, privacy, truth, and democracy. However, scholars in fields such as the history and anthropology of computing have long recognized that computer systems can perpetuate inequalities, shroud flawed decision making in the guise of automated objectivity, and enable forms of surveillance that undermine human freedom and expression.
While computers and their advancing capabilities often prompt us to look to the future, the past reminds us that the design, implementation, and use of these machines can affect different people in different ways — sometimes with lasting consequences. When we look back in time, computers cease to be abstract machines with change-making potential; they become cultural artifacts and as such provide a lens through which to see the ramifications of human decisions made in particular times, places, and cultures.
Many people have already remarked publicly that an appreciation for context should be central to the research and teaching mission of the Schwarzman College of Computing (SCC), and there is one aspect of context that we especially wish to emphasize here, namely the geographic bias found both in computer history and in current discussions of AI and machine learning.
Today, these accounts and discussions disproportionately foreground the United States, particularly the actions of tech entrepreneurs such as Mark Zuckerberg and Jeff Bezos, the companies they created, and the activities of places such as Silicon Valley and academic institutions such as MIT.
Dwaipayan Banerjee, MIT Assistant Professor of Science, Technology, and Society; photo by Jon Sachs
Action
"We highly recommend that research and teaching in the MIT Schwarzman College of Computing adopt a global approach — one that views the U.S. experience as one among many."
A substantial part of the contemporary digital world can certainly be traced to these sources, but this is not the whole story. As scholars of science, technology, and society, we study how people interact with computers in a full range of contexts — including the past and the present, as well as areas of the world outside of the United States. Given this background, we highly recommend that research and teaching in the SCC adopt a global approach — one that views the U.S. experience as one among many.
A global view of computing would demonstrate a broader range of possibilities than one centered on the American experience, while also illuminating how computer systems and the decisions of those who design them can reflect and respond to different needs, contexts, political projects, and informational infrastructures. Such experiences can prove generative for thinking about the future of computing writ large.
An example from Chile
Consider an historical example from Chile. In the 1970s, the Chilean government embarked on an ambitious political experiment to bring about socialist transformation through peaceful democratic means. As part of that project, it developed a computer system — Cybersyn — to manage the national economy. Operations research scientists created models of state-run factories and collected factory data from across the country to predict production trends.
Since the government viewed raising employment levels as central to its political platform, Cybersyn’s creators set out to use information technology to improve production in ways that would not lead to worker dissatisfaction or layoffs. Perhaps even more radical from a contemporary perspective, the system only collected data on a small subset of factory activities. This was due to the limited computer capabilities of the day. However, this technical limitation required factory modelers to be thoughtful from the outset about what data they actually needed to understand production at the national level and to develop innovative ways to do more with less.
A military coup brought Chile’s socialist experiment and work on the system to a premature end. However, the Chilean system illustrates that we can make different choices about the kinds of computer systems we build, including decisions about the values and relationships they promote. This is especially important as concerns grow about how new computing developments will shape the future of work and the ramifications of collecting large volumes of data for ends that are not always apparent or foreseeable. A more global approach to computing shows that there are many ways to build our digital world. This can open our eyes to new possibilities.
"As global citizens, we must learn from one another as we build new information infrastructures that will endure and extend beyond the boundaries of any single nation, culture, company, or community."
An example from India
Adopting a global approach to computing also shows the dangers inherent in aiming to fix social problems through purely technical solutions. Consider another example, from India. Since 2009, the Indian government has pioneered an effort to establish the world’s largest biometric database. Developed by computing pioneers in India and celebrated by technology leaders worldwide, the project aims to leverage computational infrastructures to further social welfare.
However, serious problems have plagued the system since its inception: point-of-service machines have been prone to malfunctions, while migrant workers, the elderly, laborers, and the infirm have had difficulties getting their biometrics into the system. To make matters worse, the database has had names duplicated and deleted, its security breached, and its data illegally bought and sold on the internet. The consequences of these errors include the denial of food rations, pensions, wages, and medical services — leading to deaths, identity theft, and unauthorized state surveillance. As the SCC seeks to accelerate efforts in computing to address social problems, an attention to such global examples would spur students to consider how social and political contexts affect the deployment of computing technologies.
As global citizens, we must learn from one another as we build new information infrastructures that will endure and extend beyond the boundaries of any single nation, culture, company, or community. For that reason, we encourage the SCC to closely follow the diverse, complex, and nuanced ways that computers intertwine with society and how methods from fields such as history and anthropology enable the study of these relationships.
Suggested links
Series | Computing and AI: Humanistic Perspectives from MIT
Eden Medina
Dwaipayan Banerjee
MIT SHASS Program in Science, Technology, and Society
MIT Schwarzman College of Computing
Related Publications
Beyond Imported Magic, Essays on Science, Technology, and Society in Latin America
co-edited by Eden Medina (MIT Press, 2014)
Cybernetic Revolutionaries: Technology and Politics in Allende's Chile
Eden Medina (MIT Press, 2011)
Dwai Banerjee receives 2019 Levitan Research Prize
Hematologies: The Political Life of Blood in India
Dwaipayan Banerjee and Jacob Copeman (Cornell University Press, forthcoming)
Series prepared by SHASS Communications
Office of Dean Melissa Nobles
MIT School of Humanities, Arts, and Social Sciences
Series Editor and Designer: Emily Hiestand. Communication Director
Co-Editor: Kathryn O’Neill, Associate News Manager
Published 23 September 2019