Cyprus as AI Saw it: Perpetuating Colonialism

Left: image generated by custom AttnGAN model.
Right: generated by AttnGAN model trained on Microsoft COCO dataset

Text input: 

“The ruins of ancient cities offer no attraction to the traveller in this island, as nothing is to be seen upon the surface except disjointed stones and a few fallen columns of the commonest description”

2020 – ongoing

Custom trained AttnGAN model trained on Cypriot landscapes and excerpts from the book “Cyprus as I saw it in 1879” by Samuel Baker.

Cyprus as AI Saw it in 1879: Perpetuating Colonialism is concerned with AttnGAN Text-to-Image synthesis as a tool to critically comment on issues of bias in historical and digital colonialism. The ongoing project aims to give agency to a geographic region that is under-represented in present day AI research.

Textual descriptions of the Eastern Mediterranean island of Cyprus, taken from the book ‘Cyprus, as I Saw it in 1879’ by Sir Samuel Baker, where the British official presents his impressions of the British Empire’s newly annexed colony, have been fed to two different AttnGAN Text-to-Image synthesis models. One model has been trained on pre-configured Common Objects in Context (MS-COCO) dataset, created by Microsoft in the US – one of the leading countries in the research and use of AI worldwide. The other AttnGAN model has been trained on a custom dataset of Cypriot and Eastern Mediterranean landscapes.

Baker’s colonial, orientalist gaze is reflected in his descriptions of the newly annexed colony as a degenerate country whose people are in need of salvation by the British empire. These descriptions are juxtaposed with the images generated by the AI model. Some bias is also evident in the images generated by the MS-COCO dataset, many of them bearing little, if any, resemblance to the landscape of the Eastern Mediterranean region that the input texts refer to. Instead, the generated images are reminiscent of Western or North-European landscapes and architecture. Bias has been an issue that has been flagged in a number of AI-related systems. Google’s machine vision system has labelled black people as gorillas, Beauty.AI, an AI-judged beauty contest that has been shown to significantly favour white individuals of Caucasian origin, while facial recognition AI models have shown not to recognise black skin because they have been trained on images of faces with lighter skin.

Who creates and who consumes the world’s digital infrastructure? AI technologies are expected to shape the future of our societies, yet they are being developed and deployed by a very small group of powerful, and predominantly western, countries/companies, using proprietary algorithms. What is the future place of countries such as Cyprus in the world map of this emergent digital era? 

Installation shot at Electropixel #12, Nantes, France, 2022. Photo: Thanos Vozikis
Installation shot at Electropixel #12, Nantes, France, 2022. Photo: Thanos Vozikis