Meaningful human oversight over AI requires a critical look at the temporal dynamics of how AI enters our lives. I argue that ideologies such as “move fast and break things” do just that—move fast and break things. Instead, what if we could slow down and contribute to crafting empowering futures? 

While working at a number of technology organizations in Silicon Valley, I’ve found that what has kept me true to my own values is to counterbalance a widely spread fascination with the science fiction at the center of technology innovation with a notion of speculative friction centered on the margins of those considered intended and unintended users of the products of technology innovation. My fascination with speculative friction is not about speculation as in financial investment in stocks, property, or other ventures in the hope of gain with the risk of loss. It is also not about friction that results in a waste of time, harm to vulnerable populations, value extraction, and disempowerment. Instead, it is about  the use of speculation as in critical thinking, asking questions, and imagining alternatives, not in a distant future but in the present moment. 

For example, to prevent viral misinformation from spreading, in 2020 Twitter introduced friction by nudging users to read an article before retweeting it. Other nudges on the platform ask people to pause before they post something potentially harmful. I argue that similar types of friction in the interactions between people and recommender system algorithms could trigger a generative process that empowers new models of engagement among human and algorithmic actors. 

To start with, there are many kinds of  friction we are faced with everyday, for example, friction makes a wheel move, gives rise to fires, and starts arguments. Anthropologist Anna Lowenhaupt Tsing describes friction as “the awkward, unequal, unstable, and creative qualities of interconnection across difference.”² Documenting the historical events of how corporate deforestation halted in Indonesia, at least temporarily, she shares that the process was one of “collaboration not as consensus making but rather an opening for productive confusion,”³ observing that knowledge “grows through multiple layers of collaboration—as both empathy and betrayal.”⁴ 

Inspired by her fieldwork, I suggest the need for us to collectively articulate and negotiate the social, political, and environmental aspects of friction in the context of algorithmic systems. Furthermore, in response to the strong economic incentives to move quickly in the field of AI, we aim to speculate about the benefits from slowing down and learning from educators who’ve leveraged both science fiction and speculative fiction as a tool to open up new imaginaries in AI-driven tech and tech policy innovation.⁵ 

Understanding friction

Acknowledging the complexities of using the term AI,⁶ we need to consider the temporal, spatial, and social aspects of how we relate to AI-driven sociotechnical systems. To analyze the friction among stakeholders involved in these relations, including impacted communities, civil society advocates, technology companies, regulatory bodies, and environmental ecosystems, we take inspiration from the fields of anthropology, the critical anthropology of design, and speculative and critical design. 

Lucy Suchman takes us on a journey through her 20 years of experience at Xerox PARC, sharing her reflections on the problems she encounters in the enactment of innovation.⁷ Innovation as technology production through laborious reconfigurations,⁸ exemplifies the friction between the old and the new.  In her critical scholarship, Paola Ricaurte articulates the friction between AI-enabled technology and the territories and the bodies who bear the costs, including power asymmetries, unfair labor practices, opaque supply chains, and the historical process of extractivism and dispossession.⁹ Furthermore, Sasha Costanza-Chock’s work has inspired communities globally to dismantle structural inequality through a design justice approach led by marginalized communities.¹⁰ 

What unites these interdisciplinary scholars is a critical approach to the tensions between technology innovation, design, power, and social justice, which we now consider in the context of the field of Speculative and Critical Design (SCD).  Anthony Dunne and Fiona Raby describe SCD as a type of design practice that creates friction, it aims to challenge norms, values, and incentives, and in this way has the potential to become a catalyst for change.¹¹ 

In the table below they juxtapose design as it is usually understood with the practice of SCD, highlighting that they are complimentary and the goal is to facilitate a discussion. SCD (the B side of the A/B comparison below) is not about problem solving but about problem finding, asking questions, provocation, and creating functional and social fictions that “simultaneously sit in this world, the here-and-now, while belonging to another yet-to-exist one.”¹²