Algorithms of Resistance: The Everyday Fight Against Platform Power By TizianoBonini and EmilianoTreré, Cambridge: MIT Press, 2024. 257 pp. US$ 30/£29 UK. ISBN: 9780262377485
本书探讨平台用户（如零工工人、内容创作者）如何通过日常策略理解和对抗算法控制，提出“算法能动性”和“算法抵抗”概念，适合关注平台权力、数字劳动和算法治理的读者。
The book ‘Algorithms of Resistance: The Everyday Fight against Platform Power’ by Tiziano Bonini and Emiliano Treré focuses on users' agency, which refers to the ability of platform users such as gig workers, social media users and content creators to make choices, take action and push back against the power of algorithms. The authors emphasise that resistance is necessary because algorithms often prioritise the interests of platform owners over platform users, shaping behaviour in ways that are biased or exploitative. While digital platforms such as X (formerly Twitter), Spotify, Uber and Deliveroo use algorithms to control what users see, do or how they work often without transparency, Bonini and Treré argue that users are not entirely powerless. User agency is about how people find ways to understand, navigate and sometimes outsmart these systems. As digital sociology and platform studies scholars, we are aware of the influence of algorithms in our everyday life. A dearth of literature has focused on ‘algorithmic power’, which refers to the way algorithms control how work is done on digital platforms. For example, an algorithm can decide how a task is allocated and restrict the users to use the app by blocking their account if they do not accept or complete a certain number of tasks. In many cases, these algorithms now act like a digital boss replacing the role a human manager would have traditionally played (Lata et al. 2023). However, Tiziano Bonini and Emiliano Treré challenge this monolithic discourse and argue that platform workers and users are able to resist algorithmic power. The book is divided into seven chapters including the introduction chapter. In the Introduction, building on Gidden's structuration theory, Bonini and Treré argue that ‘human agency and algorithmic infrastructures mutually shape each other’ (p. 7) and they ‘frame the relationship between users and algorithmic infrastructures as a symbiotic one’ (p. 7). They refer to this symbiotic approach as ‘algorithmic agency’, which means that people can influence or shape what an algorithm does. The authors explain that while algorithms can change the way people act or make decisions, people can also push back and try to change how the algorithm works, for example, by learning how to achieve better outcomes such as requesting rideshare customers to give a five-star rating. However, there is a key difference: humans have reflexive ability—we can reflect, learn and change intentionally. Although algorithms do not think or reflect like humans, they can still react. When people try to ‘game’ or manipulate an algorithm such as by using GPS spoofing apps to make it appear that Uber Eats drivers are located near busy restaurant hubs even though they are actually far away, the algorithm can adjust itself automatically in response, by changing the way it ranks or displays results. Chapter 1 discusses how this book contributes to media studies, cultural studies, sociology of work, science and technology studies and critical digital humanities by advancing two concepts—‘algorithmic agency’ and ‘algorithmic resistance’ before demonstrating the contradictory moral economies of platforms and users in Chapter 2. ‘Algorithmic resistance’ refers to ‘(1) an act, (2) performed by someone upholding a subaltern position or someone acting on behalf of and/or in solidarity with someone in a subaltern position, and (3) (most often) responding to power through algorithmic tactics and devices’ (p. 23). In Chapter 2, expanding on Thompson's concept of moral economy—the idea that people judge actions based on what they believe is fair or right, Bonini and Treré introduce the terms ‘platform moral economies’ to describe the values embedded in digital platforms and ‘user moral economies’ to capture users' perceptions of acceptability and legitimacy. Later, they connect this conceptualisation with de Certeau's distinction between tactical and strategic actions, which often rely on the different availabilities of time, money, expertise and control among users regarding the action they can take to resist algorithms. Building on this, they describe four different ways people interact with algorithms, either by trying to work with or against the platform's built-in values, and either in a planned (strategic) or more reactive (tactical) way. The following chapters focus on individual and collective actions adopted by gig workers, creative industry workers and politicians, their supporters and other activists. Chapter 3 explores the creative tactics, tricks and individual and collective actions adopted by couriers working for online food delivery service platforms in China, India, Spain, Italy and Mexico. Drawing on interviews with gig workers from these countries and using case studies from both the Global South and the Global North public press, the authors demonstrate how gig workers use different tactics to resist algorithmic power. For example, to increase their incomes, sometimes delivery riders enter the app as a customer and place an order and then they wait in front of the restaurant where they placed the order so that the order can be automatically assigned to them. This strategy usually works and they get the order. As soon as they pick up the food from the restaurant, they switch to their second smartphone, where they act as the customer and cancel the order to get a refund from Uber Eats. Bonini and Treré refer to these tactics as ‘tactical algorithmic agency’, where users can use everyday tactics to game algorithms. In Chapter 3, Bonini and Trerè demonstrate how Uber and Lyft drivers utilise ‘surge clubs’ tactics, where they all simultaneously log out from Uber apps to raise the price of a taxi ride. In this way, they game Uber and other similar platforms' algorithms. While Uber drivers use ‘everyday resistance’, Deliveroo riders use both everyday resistance strategies as mentioned earlier as well as collective tactics such as strikes against delivery companies. For example, in 2016, when Deliveroo UK replaced its pay structure of £7 per hour plus £1 per delivery with a fixed rate of £3.75 per delivery, over 100 Deliveroo workers protested outside the company's London headquarters, demanding the retention of the hourly wage. The same forms of resistance that are observed among the couriers of the online food delivery economy can also be found in many other domains of the gig economy such as cloud work. Bonini and Trerè further explain how some couriers strongly adhere to the moral economy of the platform that emphasises that workers should follow the rules provided by the platforms. However, as discussed earlier, many workers do not adhere to platforms' rules and use individual or collective tactics to break those rules. Interestingly, those who adhere to the moral economy of the platform believe in the platform's promises, which focus heavily on the idea that platform work provides autonomy and flexibility, even when their earnings fall. They also blame the couriers who reject the moral economy of the platform because they think they are breaking the law and do not work hard like them. As Bonini and Treré note (p. 102), ‘This group is convinced that if a courier does not receive any orders, the blame lies with him [or her] or the area where he decided to hang out… They never, ever, question the platform or the algorithm, which they consider to be highly meritocratic and fair’. The authors also highlight new, alternative food delivery platforms like Robin Food in Italy. Unlike big commercial platforms (such as Uber Eats or Deliveroo), Robin Food is a worker cooperative, which means the delivery riders who work there have a say in how the company is run. Workers are treated with respect and fairness and decisions are made democratically. Hence, the founder of this coop claims that their initiative is ‘green, ethical and local’. Like ‘Robin Food’, many cooperatives have been established around the world recently. However, the authors express uncertainty about whether these cooperative models can achieve sustainable growth while still ensuring secure employment and more democratic participation in company management. Nonetheless, they recognise these initiatives are the beginning of a new wave of resistance. In Chapter 4, Bonini and Trerè demonstrate how the rise of platforms is changing the traditional cultural industries through reshaping cultural creation, distribution and consumption—a process Nieborg and Poell (2018) refer to as ‘platformization of cultural industry’. In this, they refer to the new platform-related practices where cultural content producers are continuously grappling with seemingly serendipitous changes in platform governance, ranging from content curation to pricing strategies. The chapter shows cultural work has become more precarious due to the reliance on visible labour. Online platforms have developed a technical infrastructure that could calculate, datafy and commodify visibility. That means, platforms have power to influence the visibility of a content creator's work. The chapter demonstrates that although online platforms developed a technical infrastructure that could commodify visibility, users of platforms whose visibility is at stake are able to artificially manipulate and reappropriate their visibility. For example, music producers can artificially increase their ratings by tricking Spotify's algorithms into thinking that they are dealing with genuinely emerging hits that deserve a place on the more popular playlists. Bonini and Trerè explore the manifestation of algorithmic agency and resistance in the realm of politics, introducing the notion of ‘algorithmic politics’ and situating their discussion within the broader literature of data politics in Chapter 5. Drawing on case studies from Europe, Latin America, the US and Asia, the chapter demonstrates the use of institutional or strategic types of algorithmic politics to manipulate public opinion, spread propaganda, create an illusion of popularity and destabilise digital dissent. For example, politicians now employ automation to artificially boost the visibility of and generate an inflated image of popularity for themselves. The perceived popularity of these candidates can then influence other people to follow and like them on social media platforms, which further increases their algorithmic visibility and the related illusion of popularity. The chapter further presents case studies that showcase how the dynamics of movements have been restructured by algorithmically mediated environments. The chapter concludes with the argument that both conservative and progressive social movements use algorithms to manufacture consensus to support their cause. In particular, activists can utilise ‘the algorithmic opportunities of digital platforms to push conservative, racist and right-wing agendas’ (p. 139). Bonini and Trerè refer to this practice as the ‘agnosticism’ of algorithmic activism. For example, they present the technopolitical practices of the Spanish movement 15 M (also known as the Indignados, or the anti-austerity movement) to demonstrate how these protesters use social media algorithms to disseminate information, organise protests, build powerful narratives and influence both national and international media coverage. Spanish activists employed digital tools such as collaborative notepads for collective writing and selecting a list of possibly successful hashtags to build the momentum of the protest. Once selected, these hashtags were paired with carefully crafted sample tweets, which were then distributed to fellow campaigners via internal communication channels. To disseminate their messages widely, activists turned to public-facing platforms like X (Twitter), using them to amplify content and achieve their intended reach. They state that one of the most effective tactics of algorithmic activism used by these activists was the systematic creation of trending topics on X (Twitter). Through this way, activists can create ‘an alternative moral economy that justifies the artificial creation of popularity from below’ (p. 145). In the final chapter, Bonini and Trerè discuss in depth the importance of considering the different aspects of algorithmic agency, the presence of various moral economies within the platform society and the significance of different, everyday tactics in constructing more organised models of resistance against algorithmic control. Bonini and Trerè also remind us that tech companies are very critical of these resistance strategies which are popularly known as ‘gaming the algorithm’ and have labelled these practices as ‘immoral or illegal’. In this chapter, Bonini and Trerè argue that what platforms deem immoral or illegitimate may represent the only viable course of action for others who use them to work or increase their visibility to support their political cause. Rather than a single dominant moral framework, platform society is shaped by multiple, often competing moral economies with blurred boundaries. While both users and platforms may operate under the logic that ‘the ends justify the means’ (p. 166), their goals and methods frequently diverge. Overall, this book demonstrates that platform society is undergoing two opposing dynamics. On one side is the platformization of workers, consumers, activists and users more broadly, a process that disciplines and socialises them into the norms and expectations of the platforms' moral economy. On the other side is a growing resistance to this moral economy, which is increasingly taking the form of organised individual and collective actions. However, such structured resistance can only emerge once users develop a critical awareness of how platforms extract value from their data, bodies, emotions and performances. Through this discussion, this book contributes to the literature on ‘everyday resistance’ (Scott 1985) and critical data and algorithm studies presenting narratives on the everyday practices of workers and users of platforms within the structural constraints imposed by platforms' producers. This book transcends the silos of media and platform studies to expand our understanding of algorithms and algorithmic resistance by connecting findings with broader debates on artificial intelligence and automation. ‘Algorithms of Resistance’ will appeal to readers who are interested in platform studies, critical data and algorithm studies, media studies, science and technology studies and sociology.