All-pervasive AI
The furtive, ever-increasing influence of artificial intelligence (AI) is shaping almost every aspect of modern life. Madhumita Murgia’s Code Dependent: Living in the Shadow of AI delves deep into the intricate web of algorithms, data and human behaviour powered by AI, compellingly exploring an increasingly code-driven world. Murgia, the first AI editor for the British daily Financial Times, steers clear of narrative extremes by neither glorifying technological feats nor painting a dystopian future controlled by machines. Instead, she sinks her teeth into the murky underbelly of algorithm building, exposing the impact of AI on the lives of ordinary people around the globe.
The book’s strength lies in this focus on the human stories behind the algorithms. The experiences of individuals far removed from technology hubs bring a human dimension to the often abstract discussions surrounding AI. At the core of Code Dependent lies a profound exploration of how algorithms—ostensibly neutral and objective—are in fact embedded with their creators’ biases and values. An example that the book illustrates is the apparent bias against African Americans in medical analysis tools that use AI. It also explores “data colonialism”, or the exploitation of vulnerable communities by the Western world’s Big Tech. We learn about the data labellers of Kenya, who tag images to train self-driving cars that are now emerging in the streets of the US.
Although their meticulous work forms the invisible foundation for a technology touted as revolutionary, they remain severely underpaid and devoid of agency. But such jobs provide a modicum of financial stability to data workers, who are all mostly from the vulnerable Global South or refugees from war-torn countries with few prospects. These workers are kept disenfranchised and far removed from understanding the value of what they do so that they may be exploited as cheap labour— a mere step away from sweatshops. Murgia also does not shy away from exposing the grim consequences of new-age algorithmic systems. For instance, she explains how the increasing abuse of AI image manipulation or “deep fakes” disproportionately impacts women and people of colour.
In particular, she uncovers the terrifying business of deep-fake pornography targeting women. Such explicit content is not termed illegal, because the images are not “real”. However, they leave an intense psychological impact on those who are targeted. It is also impossible to completely scrub this content from the web because of the internet’s decentralised nature. But this complex technology and its targets are given little attention by lawmakers even in the Global North. Murgia further exposes the inherent biases embedded within AI systems. She tells a harrowing story from the Netherlands, where social workers unduly harass a mother and child after the latter is flagged by an algorithm as a “potential criminal”. The book highlights the dangers of such automated decision-making based on flawed data sets, starkly reminding us how AI can perpetuate existing social inequalities and discriminate against marginalised groups. The book also looks at how AI’s role in the gig economy is growing.
Murgia explores the lives of those working as delivery partners for mobile applications offering food and other services, who depend on opaque, ambiguous algorithms to determine their workload and income. The human cost of convenience becomes apparent, raising questions about worker exploitation and the erosion of traditional labour protections in the face of automation. Murgia argues against the notion of technological determinism—the idea that our destinies are shaped inexorably by the code we create. She emphasises on the agency and responsibility that lie in the hands of those who wield the power of code. But at the same time, it is not just another critique on AI; the book acknowledges that technology can help bridge geographical divides and improve access to essential services. For this, it takes us to Chinchpada, a small village in Maharashtra where a majority of the residents are from the Bhil community. During the novel coronavirus (COVID-19) pandemic, a doctor in a mission hospital heavily relied on an AI-based mobile application to quickly determine whether her patients had tuberculosis or covid-19. While the balanced narrative fosters a nuanced understanding of AI, acknowledging its potential for positive change, the focus consistently remains on the vulnerable and the marginalised population groups, and how AI impacts them—sometimes even without their knowledge. For instance, Murgia shares how local data gatherers in India such as ASHAS (accredited social health activists), collect clean, reliable data on rural communities that help shape Big Tech’s algorithms. These communities and the data gatherers that emerge from them help build systems for the affluent in the West, but in return, they suffer exploitation as cheap labour. All the narratives in Code Dependent are a mix of personal experiences and Murgia’s expert analysis. She is able to distill complex technical concepts into digestible prose, which helps keep the book accessible even to those unfamiliar with technology. Her journalistic background and thorough research also shine through in the interviews with data scientists to activists and different people impacted by AI. Code Dependent’s call to action is clear: we must critically examine the impact of AI on our lives. The book is a much-needed nudge to reclaim our agency and engage in a proactive dialogue about the future of AI.
This was first published in the 16-31 July, 2024 print edition of Down To Earth