The exhibition “Calculating Empires, A Genealogy of Technology and Power, 1500-2025” curated by researchers Kate Crawford and Vladan Joler at the Osservatorio of Fondazione Prada in Milan, exemplifies a focused inquiry into the current technological landscape, particularly artificial intelligence. The exhibition, 23 Nov 2023 – 29 Jan 2024, seeks to elucidate the mechanisms by which power is concentrated and the consequential impact on various sectors, as well as the environmental toll associated with AI technologies.

Venturing into the heart of Milan beneath the iconic Galleria Vittorio Emanuele II is consistently a rewarding choice. A visit to the Osservatorio provides a unique opportunity to appreciate the glass roofs from an unconventional vantage point.

Artificial intelligence, as portrayed in the exhibition, is analyzed for its capacity to centralize power, induce collective perceptions, and disseminate misinformation. The environmental implications of AI, including substantial energy, water, and mineral consumption, are underscored. The objective of “Calculating Empires” is to provide viewers with a comprehensive understanding of the intricate interplay between technology and power dynamics.

The exhibition employs maps as visual aids to convey the complexity of communication systems and their intersection with social structures. From classifications in prisons to pervasive surveillance, and from temporal considerations to educational contexts, the maps serve as analytical tools rather than purely visual representations.

A notable aspect of the exhibition is the exploration of Amazon Echo (Alexa), revealing it not only as an artificial intelligence device but as a system intricately connected to Earth’s resources, extracting data from human communication, culture, and connections. The exhibition underscores the role of individuals in daily AI training and advocates for increased awareness of this responsibility.

As of now, you still have a few days left to visit.

Here are two reflections from the curators: