The human cost (and resistance) of the platform economy

The rider’s world is reduced to a GPS line and a ticking clock. Somewhere, far away, another invisible worker sits at a screen, drawing pixelated boundaries around an image of a road—training an algorithm that promises to make that rider obsolete. Between them lies a web of unseen code, capital flows, and fragmented realities: the platform economy. The rider is not just some Korean meal courier. She is Dr. Seonok Lee, a researcher from the Rijksuniversiteit Groningen.

In this AI-driven world of labor, convenience reigns supreme, and the cost is buried deep within the folds of an algorithm. Platforms like Uber Eats, Thuisbezorgd, and Fiverr present themselves as sleek interfaces, frictionless experiences, and the epitome of modernity. But beneath their glossy surfaces lies a labyrinth of invisible hands—gig workers, data annotators, and logistical cogs—keeping the machine running. Platforms are active participants, shaping behaviors and economies through the design of their algorithms. They empower customers while leaving workers to fend for themselves. Abuse, harassment, and theft—all flow downstream, unchecked, to the worker.

We delved into the invisible labor behind these platforms during the Noorderlicht symposium Digital Mirror: Invisible Labor. Dr. Seonok Lee shared her experience cycling through the algorithmic maze of Thuisbezorgd with the talk ‘Challenges of Platform Labor.’ Lee’s embodied research revealed the emotional and physical toll of navigating customer demands and algorithmic pressures. As a rider, she felt the relentless weight of being monitored by a faceless system, and her value was reduced to a rating.

These ratings are the currency of survival. One worker bluntly remarked, ‘My rating is everything,’ emphasizing how a strong score determines not just current earnings but access to future work. (1) Another described their rating as even ‘more important than money’ because without it, the jobs simply dry up. (2) The platforms empower customers to rate and review workers but provide little accountability for harmful behaviors. The worker, stripped of autonomy, is placed under constant algorithmic surveillance, subject to the whims of customers with no one to advocate for them.

Workers have the most freedom at the start of a task—choosing whether to log in, accepting or rejecting gigs—but this autonomy erodes with each step. By the end, they are wholly subject to the algorithm and the customer’s judgment. These narratives remind us that the platform economy is not a neutral space. It is a contested battleground, where power dynamics play out between algorithms, workers, and consumers. Customers, too, play a role in shaping this dynamic. Every ride we hail, every meal we order, every gig we outsource implicates us. A five-star rating may seem trivial, but for the worker on the other side of the screen, it can determine survival. Platforms champion a ‘customer is king’ ethos, but unchecked power comes with responsibilities. Patience, fairness, and transparency are no longer acts of kindness but necessities in an economy built on precarity.

Pixel Perceptions artist Nicolas Gourault offered another lens into this world, where the boundaries between human labor and machine intelligence blur into obscurity. His work Unknown Label (video, 17’, 2023) focuses on the hidden labor of data annotators—those who painstakingly label and classify images to train AI systems like those used in self-driving cars. The film begins with stark, monochromatic animations that mimic the segmentation process these workers perform, gradually unraveling the human presence behind what we often perceive as automated magic. These early scenes confront us with the monotonous and repetitive reality of this labor, performed largely by individuals in the Global South under precarious conditions. As the narrative progresses, the visuals shift, moving through pixelated landscapes and city-scale 3D reconstructions. These digital worlds echo the detachment inherent in automation while simultaneously illuminating the vast, interconnected human effort that supports it. The aesthetic choices—geometric abstractions and layered digital textures—mirror the dehumanized, mechanical nature of the work itself.

The AI industry's insatiable demand for data labeling has created an ecosystem reliant on invisible labor, often sourced from economically vulnerable regions. Platforms have transformed places in the global South into hubs for so-called ‘Ghost Work,’ as anthropologist Mary Gray and computational social scientist Siddharth Suri describe it. (3) Platforms enforce customer-centric policies with little accountability, leaving workers like her vulnerable to algorithmic oversight and unpredictable pay. This digital labor pipeline echoes historical patterns of exploitation, as AI's global development relies on the labor of those who remain at its margins, rendering their contributions invisible even as they underpin the industry's rapid growth. (4)

Gourault’s work dismantles the illusion of AI as autonomous, objective, and frictionless. By focusing on the human cost of automation, he exposes the dependency of so-called intelligent systems on human judgment and labor. His art confronts viewers with the uneasy truth that the promises of automation often obscure the global inequalities embedded within its supply chain. These aren’t merely technical systems—they are socio-political constructs shaped by global power dynamics, capitalism, and exploitation. Workers push back in the margins. Gourault’s artwork highlights informal acts of small resistance: WhatsApp groups where riders exchange tips, VPNs to spoof locations and access higher-paying gigs, and collaborative hacks to game the system. These are the cracks where a little bit of the workers’ agency seeps through—small rebellions in a landscape of submission, citing French scholar Michel Foucault ‘Where there is power, there is resistance’. (5) Yet this resistance has limits.

Nicolas Gourault’s Unknown Label challenges us to confront the ethical contradictions of an economy reliant on invisible labor. Unlike technology, which seeks to solve problems and streamline processes, Gourault’s art compels us to linger in discomfort. It destabilizes our assumptions, transforming abstract systemic issues into tangible realities. His work doesn’t provide answers but instead asks us to question the societal structures that sustain and legitimize these systems. By making the unseen visible, Unknown Label becomes an act of resistance—an invitation to rethink our relationship with the technologies that shape our lives and the labor that makes them possible.

AI promises progress, efficiency, and transformation—but at what cost? The gig worker on the bike, the annotator drawing boxes, the customer tapping five stars—they are all nodes in a network that precariously balances exploitation and opportunity. The platform economy exists in a ghostly state of half-presence—its labor invisible, its workers fragmented, its power diffuse. Yet its contours are beginning to emerge. The cracks widen, and the questions multiply.

Gourault’s evocative work and Dr. Seonok Lee’s immersive research remind us that technology is not neutral. It mirrors the structures of power, control, and exploitation that underpin it. As we navigate this rapidly evolving landscape, the challenge is not merely to optimize these systems but to humanize them. Platforms must prioritize fairness and accountability, while consumers must recognize their role in shaping this ecosystem. Just as algorithms dictate the value of labor, our choices—how we tip, rate, and interact—directly influence the livelihoods of others. Ultimately, the platform economy reflects the world we choose to build. Do we prioritize convenience at any cost, or do we strive for a system that values the dignity of labor, recognizes invisible workers, and embraces the complex humanity that no algorithm can replicate? The answer lies not in automation, but in our collective capacity to care.

The Noorderlicht symposium Digital Mirror: Hidden Labor with Dr. Seonok Lee & Nicolas Gourault was organized by Noorderlicht in collaboration with the Jantine Tammes School, part of Rijkuniversiteit Groningen on Thursday 19 December 2024. Left to right: Roosje Klap, Dr Seonok Lee, Nicolas Gourault, and Dr Rosa Wevers during the Digital Mirror symposium Invisible Labor on 19 December 2024 at House of Connections, Groningen.

Footnotes:

1. Lindsey D. Cameron, Hatim Rahman. ‘Expanding the Locus of Resistance: Understanding the Co-constitution of Control and Resistance in the Gig Economy’. Organization Science 33, pp. 38-58. 2021

2. Ibid.

3. Gray, Mary L., and Siddharth Suri. ‘Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass.’ Houghton Mifflin Harcourt, 2019.

4. Karen Hao, ‘Inside the Secretive World of Data Labelers Who Train AI,’ MIT Technology Review, April 20, 2022 https://www.technologyreview.com/2022/04/20/1050392/ai-industry-appen-scale-data-labels/, accessed 23 December 2024.

5.  Michel Foucault, ‘The History of Sexuality,’ trans. Robert Hurley. New York: Pantheon Books, 1978