Learning from atrocity: When machines regard the pain of others

The images that have emerged from Gaza, capturing catastrophic violence inflicted upon the Palestinian people by the state of Israel, are atrocious, haunting, injurious. Gaza-based photojournalist Motaz Azaiza came to the fore in the early months of the onslaught as a critical source of on-the-ground information, capturing and posting a steady stream of horror to his Instagram account. Social media outlets are replete with images like his, paired alongside pleas to ‘Stop scrolling!’, ‘Keep looking!’ and ‘Don’t look away!’ This article questions what it means to look at images of atrocity within contexts defined by their financialization and instrumentalization, wherein images, now operational, work to fuel online economies and inform developments in machine learning and artificial intelligence. Reflecting on the images and commentary posted to Azaiza’s Instagram account and building on existing efforts to theorize intersections between spectatorship and photographic representations of atrocity, the article investigates the technical appropriation, financialization and instrumentalization of the atrocious image online, exposing its complicities and identifying how functions of the image that were once lauded are precisely those now being exploited.