Andreu Belsunces Gonçalves researches the intersections between sociology of design and technology, politics of fiction and collaborative practices. He is interested in the entanglements between discursive practices and material phenomena in technology development, especially focusing on the convergence between narrative artifacts, infrastructures, data, epistemology, democracy and industrial outlets. He is the co-founder of Becoming, an emerging scenarios research studio, and associate professor on critical thinking, artistic and design research, and speculative design at Escola Massana, Elisava and Istituto Europeo di Design.
In the field of art and culture, he has performed, researched with, or collaborated with MACBA, Hangar, CCCB, Tecnopolítica-IN3/UOC, Sónar +D (Barcelona), STRP (Eindhoven), UNESCO (Montevideo), The Wrong Biennale (International), Medialab Prado (Madrid), Primer (New York) among others. His work has been published in several media studies and design journals.
Fiction is usually understood as an information protocol that draws the line between what can beconsidered truthful from fantasy. Nevertheless, it can also be considered as an agent, a thingthat makes things, both in the symbolic, cultural and epistemic spheres, as well as in thematerial realm.
This second approach that overcomes dualities (like artificial and natural, cultural and material)and recognizes the agencies of fiction, is rooted in the assumption that both human andnon-human entities affect and constitute our world in meaningful (although different) ways.Doing so, it dialogues with Actor-Network Theory (Callon 1998; Latour 1996), post-humantheory (Braidotti 2016), speculative realism (Barad 1998, 2003; Haraway 2016), and ObjectOriented Ontology (Harman 2005, Bogost 2012, Morton 2016), recognising also that any object can be considered as an event, in the sense that there are always interaction within it, as Bennett (2010) puts it.
Following this path, some works point out that social institutions such as money, the state or thenation are made partly by fiction (Shaw and Reeves-Evison2017). At the same time, fictionoperates as an institutionalizing practice, as happens with the credit system, for example. AsBeckert (2015) explains, in a capitalist context, applying for a loan is based on adecision-making process articulated by personal and social expectations, that is, a fiction thatcannot be exhausted by rational calculation. By performing this role, fiction becomesinfrastructural for the symbolic functioning of capitalist markets.
Likewise, fiction plays a key role in sociotechnical imaginaries (Jasanoff 2015), in the sense thatany technoscientific configuration embeds promises about what will happen, like when scientists promise to find a vaccine for Covid19 in a very short time. By doing so, they are trying to mobilise resources (finding funding and support) in the present to fulfill these promises in the future. Based on this neo-materialist approach to fiction, this paper will investigate AI, from the perspective of a long-standing emerging sociotechnical phenomena that has been deploying fiction to stabilize itself. To explore this hypothesis this resarch will analyse the phenomenon called the AI winter (Hendler 2008, Florid 2020).
The idea of AI winter describes hype cycles where excitement and disappointment related to AIdetermine funding coming both from public administration and venture capitals. In the currentpositive hype cycle of AI, autonomous functionalities have been performed by humans(Sadowski 2016) -a famous example of this being the Mechanical Turk- to amuse capitalistinvestors to carry on with the fundings. This strongly resonates with the CCRU (Nort, 2015)concept of hyperstition, where a future event tries to realise itself in the present throughfeedback hype looops.
Here, the abstraction and opacity of financial trade and AI are entangled with the act ofspeculation in different senses. Based on this analysis, this research will question the verynotion of speculative fiction and how it is defined. It will then use the following categorisation todraw the relationship between speculative fiction and technology industry. (1) Technicalspeculation: AI, as a technology needing huge efforts of innovation, is dealing with fiction andspeculation in its development processes, in the sense that, as long asit is not functioning, it ismore of an hypothesis than a fact, (2) Financial speculation: trade markets create computationalsystems to face risk and predict future events that will guide their investments in order toincrease their capital. Because of this, the act of @‘futuring’ has a fictional nature in the senseof Beckert (2015). And (3) Fantastic speculation, which is the fictional (in a traditional sense)representation of AI, that feeds the hype cycles while structuring favorable sociotechnicalimaginaries or collective visions of desirable futures.
This work will explore the role of these three forms of speculation in the sociotechnicalstabilization of AI in order to keep developing in the infrastructural nature of fiction incontemporary societies