A Structural Plan for Imitation: Engines of Differentiation

Photo: Nicos Avraamides, 2024

2-channel video installation

in collaboration with Alexia Achilleos
dimensions variable
2024

A Structural Plan for Imitation: Engines of Differentiation considers how the concept of the model plays out in recent contexts surrounding artificial intelligence, looking at discrepancies between the promises made about AI and the realities of how it materially acts on the world. With the current emphasis on methodologies that rely heavily on the implementation of models, it’s crucial to consider what assumptions, power dynamics, and ideologies are embedded in these practices. The project explores this idea through models’ structuring of relationships between visual perception, real-world phenomena, and the traditional systems of value that have culminated in the present pervasiveness of artificial intelligence.

 

models are ways of exploring the world or accomplishing certain goals. We use the model to pose and answer questions. A model is useful, successful, accurate just insofar as it achieves a purpose.

— Alva Noë, 2015

While immaterial themselves, models, algorithms, and computation are capable of structuring the world in significant ways. Visual culture is increasingly influenced by statistical models based on the large-scale analysis of data, and even of visual culture, itself. This holds resounding implications for the aesthetics, significance, and ideological value of images. Creating visualisations based on computational models of the world causes a shift in which images that result from such methods act as interfaces between visual and non visual, human and nonhuman, informational and material. Behind a given image lies a massive amount of data, computation, infrastructure, and resource consumption, in addition to a complex web of ideological and ethical implications that are difficult or rather impossible to disentangle.

As we struggle to come to terms with AI’s affordances and perils, it is especially relevant to consider to what extent these methods and their outputs actually align with the way they are often described in public discourse. If the news headlines are to believed, artificial intelligence is taking over the world, it’s a threat to our humanity, art, and jobs, yet it will paradoxically make the world a more equitable place in the process. Detractors highlight artificial intelligence’s tendency towards bias, extractivism, and resource-intensiveness, among a long list of other issues. But in some cases, even displays of skepticism play into the interests of tech companies themselves, inflating the importance, power and visibility of artificial intelligence rather than addressing its genuine potential to inflict harm.

Working from the idea of the model, A Structural Plan for Imitation: Engines of Differentiation focuses on several central talking points that are often seized upon, recycled, and used to serve various interests, whether criticising or bolstering artificial intelligence in public perception. The installation and the video will look into the material basis of artificial intelligence, something that is often discussed as being omnipresent, yet difficult to locate. Looking at the actual infrastructure (material and conceptual) behind the generation of images through machine learning models, the project seeks to develop a more concrete, grounded perspective on the subject than the generalisations that are commonplace currently.

 

 

This work was realised within the framework of a European Media Art Platform (EMAP) residency program at NeMe in Limassol, Cyprus with support from the Creative Europe Culture Programme of the European Union.