School of Social and Political Science

Description

With Srravya Chandhiramowuli  and Dr Amir Anwar

In the 1970s the rich world started to export it's data processing needs to low income countries, starting with India. This outsourcing of the tedious behind the scenes work of our digital lives and business continues to grow, and has come to include moderating our social media and training our AI system. The modern giants of Indian IT industry are some of the descendants of this trend, but more diverse model of outsourcing were developed in the early 2010s - newly connected and education populations in many countries could now do this relatively low paid work, and become 'part of the global digital economy'. However the conditions of this work have been the source of considerable controversy. Big Tech, from Meta and Google to OpenAI have resorted to armies of subcontracted workers working in precarious conditions to clean up the toxic material that floods the internet, to rank websites for search, or train the huge models of the AI age. They are asked to do mind-numbing work, such as producing deliberately bad coding for AI programming tools, tag endless pictures of road scenes for autonomous vehicles, or even watch people shopping while pretending to be “AI”.   In these talks to of our researchers tell stories from the frontline of this work - in India and southern Africa -  does this work offer economic  opportunity, and pathways to a career, does it allow the trickle down of money from Silicon Valley, or does it trap people in poor working conditions doing tedious work?  How do data-work centres change the social order in the towns and villages, and challenge social and cultural values?

Friday 3:19-5:30pm Newhaven Lecture Theatre, 13-15 Old College Street Edinburgh

Get the ZOOM link from Eventbrite

Srravya Chandhiramowuli,  PhD candidate in Design Informatics

Data annotation for AI: Following ‘impact’ sourcing within global supply chains    

AI systems have made the headlines quite frequently over the last couple of years, most often for their remarkable computational capabilities. Lesser known and rarely acknowledged is the human labours involved in producing datasets for training and supporting these celebrated AI systems. Thousands of workers, particularly in global south regions, are engaged in creating large-scale annotated datasets to sustain AI’s research, development and use under the banner of ‘impact sourcing’. Yet little is known about what their work entails. What do data annotators do when they label data for AI? And how is this work organised within the AI supply chain to continually feed into, train and sustain AI systems? I aim to answer these questions by drawing on ethnographic study of data annotation work conducted in outsourcing centres in India. Through this talk, I hope to share insights into the everyday work practices, organisational hierarchies, norms, and values that shape data annotation for AI and reflect on how they are caught in global flows of resources, rhetoric, and relations of power. 

Bio: Srravya is a technology researcher and PhD candidate in the Institute for Design Informatics at the University of Edinburgh. Her current research examines the work of data annotation for AI, paying particular attention to systemic challenges and frictions, to envision and inform just, equitable futures of AI. Her research builds on scholarship in Human Computer Interaction (HCI) and Science and Technology Studies (STS) and seeks to shape the design, policies and practices surrounding emerging technologies.

Readings 

Srravya Chandhiramowuli, Alex S. Taylor, Sara Heitlinger, and Ding Wang. 2024. Making Data Work Count. Proc. ACM Hum.-Comput. Interact. 8, CSCW1, Article 90 (April 2024). https://doi.org/10.1145/3637367 [pre-print: https://arxiv.org/abs/2311.18046]

Milagros Miceli and Julian Posada. 2022. The Data-Production Dispositif. Proc. ACM Hum.-Comput. Interact. 6, CSCW2, Article 460 (November 2022), 37 pages. https://doi.org/10.1145/3555561 

Ding Wang, Shantanu Prabhat, and Nithya Sambasivan. 2022. Whose AI Dream? In search of the aspiration in data annotation. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems (CHI '22). Association for Computing Machinery, New York, NY, USA, Article 582, 1–16. https://doi.org/10.1145/3491102.3502121 

Noopur Raval. 2021. Interrupting invisibility in a global world. Interactions 28, 4 (July - August 2021), 27–31. https://doi.org/10.1145/3469257 

Dr Mohammad Amir Anwar,  Lecturer in African Studies and International Development; Programme Director MSc International Development

From self-checkouts to self-driving cars to social media newsfeed to surveillance systems, AI is part of our lives whether we know it or not. Some advocates proclaim AI as ‘the greatest technology humanity has yet developed’ and it will reshape our society. Others say AI will contribute about US$ 15 trillion (roughly the size of the EU economy) to the global economy by 2030. Yet, AI remains an enigma: little is known about how AI systems are made, and that human labour is behind them. For machines dependent on AI to work effectively in identifying patterns and making predictions, they need to be trained first. This is called machine learning (the science of getting computers to make specific decisions), which depends on behind the scenes human labour. As a result, various forms of data work (such as training cameras for autonomous vehicles) have emerged as new employment prospects for many, including those in the 'majority world'. There are already discussion of the transformative potential of AI on economic development, including the growth of digital jobs and poverty reduction. In this discussion, I challenge some of these assertions and highlight that data work is defined by precarity and vulnerability and reflect on what this means for workers in the age of AI.

Readings: 

Work Without Workers. Phil Jones. Verso.
Automation is a Myth. Luke Munn. SUP.
The Digital Continent. Amir Anwar and Mark Graham. Open Access Book, OUP.

Bio: Amir is a Lecturer in African Studies and International Development. At Edinburgh, he teaches on a range of Postgraduate (both online and on campus) and Undergraduate courses.In the past he has convened the courses: Politics and Theories of International Development and Governance, Development, and Poverty in Africa, and Researching International Development. As a testament to his teaching excellence, he has been awarded the Fellowship of the Higher Education Academy (UK). He is a Senior Research Fellow (Honorary) at the British Institute for Eastern Africa and Senior Research Associate at the School of Tourism and Hospitality, University of Johannesburg. He was also a Fellow of the World Economic Forum’s Global Future Council on Digital Economy and Society.  He holds a Ph.D in Geography from Trinity College Dublin. He has extensive experience of conducting research both in India and Africa. Before coming to Edinburgh, Mohammad was a Researcher at the Oxford Internet Institute, University of Oxford, where he was a part of the ERC-funded project Geonet: investigating the changing connectivities and potentials of Sub-Saharan Africa’s knowledge economy. As a part of the project, he lead the research on digital outsourcing and gig economy in Africa. At Oxford, he also taught a Masters course on Economic Development in Digital Capitalism. Before joining the OII he worked as a Post-Doctoral Research Fellow at the University of Johannesburg (UJ). At UJ, his work examined at the political economy of India’s engagements in Africa. He also briefly worked as a Research Assistant at Trinity College Dublin, for an Irish Research Council funded project on the role of information and communication technologies in enterprise development and industrial change in Africa. He is the author of 'The Digital Continent', (2022) published by Oxford University Press. He has 20 peer-reviewed publications including highly-reputed journals, such as Environment and Planning A, Competition and Change, Gender and Development, Canadian Journal of Development Studies, Review of African Political Economy, Journal of Modern African Studies, International Labour Review, Singapore Journal of Tropical Geography, First Monday, and Urban Forum. Peer-reviewed book chapters have been published by Sage and Cambridge University Press. He regularly contributes to public debate through blogs, articles for online news sources, and radio interviews. His articles have appeared in New Statesman and The Conversation.  Mohammad has received funding the British Academy/Leverhulme Trust, the Economic and Social Research Council (ESRC), the Foundation of Urban and Regional Studies. His PhD was funded through Trinity Research Studentship, Trinity College, University of Dublin.