Overview
Some advocates proclaim AI as ‘the greatest technology humanity has yet developed’ and claim 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 where and how AI systems are made, who are the key actors involved, and what are the socio-economic and ecological costs associated with AI.
AI systems often thought to be automated are actually dependent on human workers labelling and cleaning data to train algorithms when they give incorrect answers. ChatGPT and driverless cars would not exist without this behind-the-scenes human labour for AI. Large Silicon Valley firms rely on the outsourcing of data enrichment tasks (commonly known as ‘data work’) to different parts of the planet via a host of suppliers and labour platforms. Thus data work value chains (different stages of production, and distribution of goods) can go anywhere across the planet both to Global North countries (i.e., high-income) such as the United States and the United Kingdom and Global South (i.e., low and middle-income) such as India, Colombia and Kenya, among host of others. The project will map these data work value chains to show how the planetary AI systems are made.
AI systems are also deeply material in nature. Not only they depend on planetary networks of resource extraction from DRC to Chile to Indonesia, they require large physical infrastructures such as data centres that are rapidly emerging. These data centres are implicated in diverting million of gallons of water and tremendous amount of electricity away from communities affecting lives and livelihoods of people living nearby. This project studies the environmental costs associated with data centres and resistance to them in the Global South.
- i. In the context of an emerging AI and development agenda, the overall objective of the project is to develop grounded theoretical and analytical insights into the production networks of AI by placing data work and data centres at the centre of its enquiry.
- ii. To develop and apply innovative methodological tools combining multidisciplinary approaches from geography, anthropology, political ecology, computational data science, etc to investigate value chains of AI.
- iii. To produce original empirical evidence showcasing (a) different types of data work, actors, and its implications on local labour markets and workers and (b) various socio-environmental costs associated with data centres in the Global South.
- iv. To study production networks of AI in India, Colombia, Kenya and Uganda.
- v. To co-produce findings from the Global South and adopt a diverse range of dissemination strategies across multiple audiences and platforms.
- vi. Build capacity among early career researchers across the Global South.