All is fair in AI Development(?) 

 

Who is marginalized in this new AI (Artificial Intelligence) era? Why are their voices so often unheard?  To what extent do the economy and politics shape our very existence?   

At first, these questions might seem distant, mere abstractions—until you move closer to the realities they uncover. In the post-pandemic era, I heavily rely on digital technologies and (e-commerce) platform economy to ease my work and cope with stress, despite being neither tech-savvy nor a digital worker.Without a grain of doubt, AI has woven itself into every dimension of life since the pandemic, from healthcare to business logistics and supply chain, education to research, and beyond.AI in the Post-Pandemic World

This surge of AI adoption and reliance on digital technologies in the post-pandemic era has reshaped not only how we navigate our day-to-day life but has even revealed layers of hidden narratives in the global digital economy — stories which books like “Digital Work in the Planetary Market” bring to light and challenge us to delve deeper into the worlds of technology, labor, and inequality that are intertwined. The book discusses digital labor and micro work in Sub-Saharan Africa and Southeast Asia, Amazon’s business model, and the heterogeneous nature of economic geographies in the light of digital labor. It provides a tabulated overview of the themes highlighting complexities in the context of digital labor, especially where it applies to the disadvantaged worker and the global economy. 

This understanding often raises complex questions such as, Who is marginalized in this digital world? And perhaps most pressing—are the economy and politics dictating our very existence?

 

Marginalization:

It is a multidimensional phenomenon involving not only deprivation of material resources, but also involuntary exclusion from political, economic and social processes. These patterns of exclusion are often compounded by discrimination and social differences, categorizing people as ‘marginalized‘. It involves more than material poverty; it includes deprivation of knowledge, political rights, recognition, power and capacity (Jenson, 2000). This is further exacerbated in this digital age that has been shaped by a new global political economy with its platform economy, gig economy, and aspects of a neo-capitalist framework.

 

Gig economy associated with AI development:

The gig economy typically includes people working on short-term contracts or as freelancers, often receiving hourly wages. But how does the gig economy connect to AI and algorithms?

Shadow side :

Invisible Labor Powering AI:

AI development isn’t fully automated from the start; it requires significant data entry and annotation by human labor, often referred to as “ghost workers,” “cloud workers,” and “moderators.” These data workers provide the foundational labor necessary for algorithm training and data analysis. Since AI companies often lack the resources and expertise to ensure quality and precision, they frequently outsource this work through business process outsourcing or digital labor platforms, creating complex networks of AI production (Muldoon et al.,2024). These external services provide flexible and cost-effective labor for large-scale AI data projects.

While AI appears to simplify decision-making, it relies on significant human oversight and intervention. This aspect of the gig economy comes with its challenges. Unlike traditional employment with unions and established labor protections, the conditions for data workers in the gig economy are often far more precarious and unregulated. These workers are typically hired from remote locations, paid hourly at minimal rates, and subject to high workloads without fixed hours. Some describe them as the ‘Slaves of Modern-Day‘ (Real Stories, 2024; Graham & Ferrari,2022).The “Modern-Day Slaves” Of The AI Tech World .

 

Hurdles confronting AI workers:

This less discussed sector—training AI, large language models (LLMs), and content moderation—requires intense, manual labor, often involving exposure to disturbing content and scenarios. A qualitative study highlights the severe psychological impacts, including cognitive changes and symptoms such as intrusive thoughts, distress, depression, burnout, and even Post Traumatic Stress Disorders-PTSD (Spence et al., 2023). Despite these challenges, many are drawn to these jobs due to difficult family circumstances—such as the loss of an earning family member, single-parent responsibilities, and health issues—or because of the flexibility it offers. 

Additionally, the technicalities of this digital labor market — such as access to the internet, English communication skills, and lack of dedicated workspace, widen the gaps for refugees as freelancers who depend on gig income (Graham & Ferrari,2022). As prof. Amartya Sen rightly asserts, challenging circumstances can restrict individual opportunities, well-being, and personal freedom, thereby affecting overall development (Kuhumba, 2018).See Internet Encyclopedia of Philosophy.

 

Bright side :

On the other hand, the developing nations view this proliferation of the digital (gig) economy as an opportunity to strengthen their economy and labor market despite the informal nature of its operation (Surie and Huws, 2023). In diverse and densely populated countries like India, many tech companies, such as ‘Karya’ are working to bridge the gap between training AI and regulating the data work involved. According to local people, this approach may empower communities and help them keep up with a rapidly changing world. The Workers Behind AI Rarely See Its Rewards. This Indian Startup Wants to Fix That. 

 

The New Marginalized in the Age of AI:

In this evolving digital economy and expanding AI landscape, factors like limited economic resources, knowledge gaps, restricted political rights (refugee AI workers), lack of recognition, and power imbalances contribute significantly to the marginalization of these workers. This imbalance will undoubtedly continue to shape the future of AI labor and those within it, implicitly  influencing human development. This dual nature of AI —-both a tool for empowerment and exploitation — represents a double-edged sword that requires vigilant handling. All’s Fair in Love and War —- now in AI (development)?

 

Addressing issues surrounding AI development:

A proactive approach is needed to raise awareness about gig work, particularly in the AI sector, and advocating for fair labor practices is essential This can be achieved by supporting companies that prioritize ethical practices and fair treatment of workers (refer to my comment at the bottom of Navigating the digital economy – Leading the way with your voice. – Artificial Inequality). 

Furthermore, engaging in discourses that support ethical AI initiatives, promoting digital literacy, advocating for policy change, sharing (AI tech workers’) personal experiences, participating in research, hosting a webinar or panel discussion, and encouraging mindful consumption—  these initiatives collectively facilitate the development of ethical AI. Additionally, prioritizing the mental health of AI tech workers is of paramount importance (Stieger et al.,2021). Effective interventions could include incorporating holistic wellness programs in workplaces or online and providing psycho-social support through individual counseling, mindfulness therapy, and ensuring adequate breaks between work sessions.Here, I conclude this blog with a few complex questions (mentioned above) for you to contemplate, along with suggestions for active engagement in the digital landscape of AI progression to create a fair and inclusive world.  

 

In my upcoming and concluding blog, I will explore the future impacts of AI by reflecting on two compelling YouTube broadcast interviews with Nobel prize laureate and ‘Godfather of AI’, Professor.Geoffrey Hinton. It will be an interactive blog post, I invite all of you to share your thoughts and insights.

 

 

Featured Image:©[ Image generated via Canva.com]

References:

Graham, M & Ferrari, F. (eds.) (2022): Digital Work in the Planetary Market.MIT Press.

Jenson, J. (2000). Backgrounder: Thinking about marginalization: What, who and why?. Ottawa: Canadian Policy Research Networks.

Kuhumba, S. (2018). Amartya Sen’s capability approach is the theoretical foundation of human development. Journal of Sociology and Development, 1(1), 127-145.

Muldoon, J., Cant, C., Wu, B., & Graham, M. (2024). A typology of artificial intelligence data work. Big Data & Society, 11(1). https://doi.org/10.1177/20539517241232632.

Muldoon, J., Graham, M., Cant,C. (2024,6 July) Meet Mercy and Anita–the African workers driving the AI revolution, for just over a dollar an hour, The Guardian. Retrieved November 11, 2024, from https://www.theguardian.com/technology/article/2024/jul/06/mercy-anita-african-workers-ai-artificial-intelligence-exploitation-feeding-machine.

Spence, R., Bifulco, A., Bradbury, P., Martellozzo, E., & DeMarco, J. (2023). The psychological impacts of content moderation on content moderators: A qualitative study. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 17(4).

Steiger, M., Bharucha, T., Venkatagiri, S., Riedl, M., Lease,M. (2021). The Psychological Well-Being of Content Moderators: The Emotional Labor of Commercial Moderation and Avenues for Improving Support. In CHI ’21: CHI Conference on Human Factors in Computing Systems. https://doi.org/10.1145/3411764.3445092

Surie, A., & Huws, U. (2023). Platformization and informality: Pathways of change, alteration, and transformation. In Platformization and Informality: Pathways of Change, Alteration, and Transformation (pp. 1-12). Cham: Springer International Publishing.

Featured Image Source:©[ AI Image generated via Canva.com]