With the continuous evolution of technological advancements, algorithm-driven systems have assumed a profound position in shaping the way individuals interact with the world. This dynamic interplay, marked by a reciprocal interaction between users and algorithms, has resulted in the transformation of our societal landscape. Yet, the role that algorithms have increasingly taken on as intermediaries between humankind and the world not only highlights their transformative potential, but also brings to the forefront the social, political and ethical implications of these systems. Importantly, this role illuminates a myriad of issues arising from algorithmic systems. One such issue is their opaque nature, as hiding behind the veil of complexity, algorithms create a ‘one-way mirror’ effect – a metaphor used by Frank Pasquale in The Black Box Society: The Secret Algorithms that Control Money and Information. This opacity limits understanding and scrutiny from the people affected by the decisions made by these systems. Moreover, there is a significant risk of exacerbating existing inequalities, as algorithms, if not carefully designed, can perpetuate biases. Additionally, the pervasive adoption of algorithm-driven systems poses a threat to democratic values, prompting apprehensions about their potential influence on public discourse and democratic processes.
In light of these issues, concerns emerge about the power dynamics between these systems and society at large, which has found itself subject to algorithmic power and its implications. Naturally, such concerns lead to the question: can algorithmic power be resisted and, if so, how?
Embedded within this evolving landscape of algorithmic power is the theoretical concept of ‘everyday resistance’ introduced by James C. Scott in his book Weapons of the Weak: Everyday Forms of Peasant Resistance. Deriving from the daily struggles observed in peasant societies, the concept encompasses quiet, day-to-day acts of resistance, which Scott alternatively refers to as ‘infrapolitics’. Rather than framing resistance strictly as organised and confrontational dissent, it redirects attention to less conspicuous forms, devoid of revolutionary implications. According to Scott, these acts include ‘…foot dragging, dissimulation, false compliance, pilfering, feigned ignorance, slander, arson, sabotage, and so forth.’ They serve as tactical measures employed by exploited individuals to endure, challenge, and subvert oppressive domination.
While the concept originated from the struggles of peasant societies against the dominant system, it proves valuable for understanding agency in objecting various forms of power, including algorithmic power. Building on the theoretical concept of everyday resistance, this article delves into day-to-day acts of resistance against the prevailing governance of algorithms. By examining media practices of repair and the anti-face recognition movement, it aims to illuminate the nuanced ways individuals navigate and contest the influence of algorithmic systems in their daily lives. This exploration contributes to a broader understanding of the challenges and possibilities that arise in the context of ‘the black box society’.
Media Practices of Repair: Navigating Algorithmic Biases
‘…even though power might run through algorithms, resistance to the governing power of algorithms, and in particular to the politics of attention that they conduct, takes place from within the logic of the algorithm.’
Julia Velkova and Anne Kaun
In her book Algorithms of Oppression: How Search Engines Reinforce Racism, Safiya U. Noble critiques the political-economic framework and representative discourse surrounding racial and gendered identities on the internet. Noble argues that data discrimination is a real social problem, perpetuated by the misrepresentation carried over from traditional media to digital platforms, resulting in algorithms favouring whiteness while simultaneously discriminating against people of colour.
Despite the power dynamics embedded in socio-algorithms and the overarching goal of the information age to disembody users or to mitigate the hegemonic implications of the technological revolution, Noble shifts attention to the ways African Americans have adapted and contextualised technology within significantly different frameworks. Her work reflects upon radical interventions in socio-technical systems, urging a more constructive and thoughtful approach that avoids further marginalisation of those already marginalised.
In addressing this question of user agency in algorithmic culture, Julia Velkova and Anne Kaun propose media practices of repair. This concept centres the user as an agent in shaping socio-technical systems and serves as a strategy to correct deficiencies in algorithm-driven systems, not by creating alternative pathways but by addressing existing shortcomings. Velkova and Kaun illustrate it through a case study on Swedish artist, Johanna Burai, and her ‘World White Web’ project.
In 2015, while researching images of hands for an assignment, Burai – a graphic design student at the time – noticed racial bias in Google Images search results, predominantly featuring images of white hands. In response, she undertook an initiative to create a collection of racially diverse hand images that would top search outputs. As Velkova and Kaun explain, Burai understood that to influence Google’s search results, her pictures needed to be uploaded and shared by websites with a high PageRank in the system. After consulting experts in the field, she learned that international news outlets tend to have a very high PageRank and their content is considered important by Google’s system. As a result, Burai launched a media campaign. By promoting the ‘World White Web’ on press releases and social media platforms, the project garnered impressive media coverage from international news outlets and, within a short period, her pictures surged to the top of Google Images search results for hands. Burai’s case thus provides a framework for understanding how users can exercise resistance against the current algorithmic power by reshaping and influencing algorithmic outputs.
The Anti-Face Recognition Movement: Defying Algorithmic Surveillance
Patricia de Vries and Willem Schinkel
In the aftermath of the terrorist attacks on 11 September 2001, the global landscape witnessed a swift increase in the acquisition of information for security purposes, marking the genesis of what scholars have aptly termed ‘globalised surveillance’. Subsequently, the pervasive adoption of biometric technology has become increasingly evident across various sectors, with notable applications in law enforcement, security, and finance. One particularly prominent subspecies of biometric technology is face recognition. Although it was initially conceived to enhance human welfare, as Xiaojun Lau and Pei-Luen Patrick Rau assert in their article for Computers in Human Behavior, its widespread implementation has raised profound ethical and legal concerns, encroaching upon individual privacy and infringing upon fundamental societal principles.
Prompted by concerns surrounding this technology, recent years have witnessed a growing anti-facial recognition movement, including a variety of artistic interpretations of the issue. In their article for the journal Big Data & Society, Patricia de Vries and Willem Schinkel explain that when modern artists engage with facial recognition algorithms, they often centre their works on the human face, resulting in the emergence of masks and camouflage attire as artistic responses to conceal and make the human face unrecognizable to facial recognition technology.
One notable project within this movement is ‘HyperFace’, developed by German artist and technologist, Adam Harvey. This project uniquely challenges facial recognition systems through printing patterns on textile that algorithms recognise as human faces. As elaborated by Vries and Schinkel, by presenting false positives to these systems, the project ‘…reduced the confidence score of the true face (figure) by redirecting more attention to the nearby false face regions.’
Impactful Resistance: Doubts and Realities
In examining instances of everyday resistance against algorithmic power and addressing the inherent issues within algorithm-driven systems – such as their opaque nature, the risk of exacerbating inequalities and the looming threat to democratic values – a crucial question emerges: do day-to-day acts of resistance truly have an impact?
For example, Velkova and Kaun detail how, following the launch of the media campaign for the ‘World White Web’ project, Burai’s pictures of racially diverse hands almost immediately topped search results on Google Images. For a long while after, her pictures consistently featured on the first search result page, albeit not always in the same position. However, in my recent search on 20 December 2023, I noticed a significant absence of Burai’s images. Not only did they fail to appear on the first search result page; they did not appear at all, regardless of how long I scrolled down for.
Considering this, it is not surprising that doubts may arise regarding the long-term impacts of day-to-day acts of resistance. In addressing these concerns, it is essential to recognise that, when introducing the concept of everyday resistance, Scott did not suggest that individual acts could bring immediate change or overthrow the governing power. Instead, he argued that the true potential of everyday acts of resistance lies in their cumulative effect, gaining significance only when ‘multiplied many thousandfold’. Although individual acts may seem small and easily overlooked, their collective influence over time and across a population holds the power to effect substantial long-term changes. When sustained collectively, these actions can challenge established power dynamics, question oppressive norms and contribute to broader societal or political transformations. Through persistent and widespread resistance, cultural attitudes can be reshaped, policies influenced and systemic changes achieved.
As we recognise the potential of everyday acts of resistance and their cumulative effect, what trajectory does this interplay between resistance and algorithmic transformation set for our vision of society?
Towards ‘No Place’
‘The defining characteristic of utopianism is that it is a political theory specifically directed towards the creation of human happiness.’
Goodwin and Taylor, The Politics of Utopia: A Study in Theory and Practice
When we actively envision a society where cultural attitudes have been reshaped, policies influenced and systemic changes achieved, we delve into the essence of the concept of ‘utopia’. From Thomas More’s seminal work, Utopia, to the contemporary landscape, the concept has consistently depicted an ideal society marked by transformative change and collective efforts towards a more just and inclusive world. Over time, diverse ideas have emerged on the path to utopia. As Arthur O. Lewis notes, Western thinkers have explored various avenues, with a notable emphasis on the belief that the quickest way to utopia lies in technological advancement – a notion eagerly embraced particularly in the United States, where algorithms are seen as crucial for realising utopian ideals. However, numerous problems emanate from these systems, shedding a critical light on the notion that technological advancement alone is the panacea for utopian aspirations.
The continuous evolution of algorithms introduces intricate problems. Ironically, the very systems intended to shape a utopian society have, at times, become sources of contention, raising questions about inherent issues in their design and deployment. Therefore, as articulated by Andreas Sudmann in Media Infrastructures and the Politics of Digital Time: Essays on Hardwired Temporalities, preserving the utopian potential of algorithmic systems today requires leading them ‘out of the mode of speculative discourse’ and confronting tangible challenges embedded within these systems. This necessitates a shift from idealised narratives to a grounded exploration of the social, political and ethical dimensions of these systems, acknowledging both their promises and pitfalls as we navigate the path towards a utopian future.
Engaging in everyday resistance becomes crucial in the pursuit of a utopian society. The synergy between active resistance against algorithmic problems and a strategic approach to addressing existing shortcomings within algorithm-driven systems serves as a linchpin for transformative change. The resistance strategies, exemplified in instances like media practices of repair and the anti-face recognition movement, showcase the potential of individual and collective actions in redirecting algorithmic trajectories. By actively challenging biases, opaque structures and threats to democratic values, individuals engage in a form of everyday resistance that transcends mere opposition. It is a conscious effort to co-opt algorithms, steering them to the realisation of utopian ideals. Through persistent efforts to rectify deficiencies and challenge the current algorithmic culture, individuals contribute to the gradual transformation of societal attitudes – a pivotal step towards realising a utopian vision through the positive utilisation of technological advancements.
This article was originally published in OPR’s Issue 12: Utopia.
Labëri Leci is a Master’s student of International Politics at Leiden University in the Netherlands.