[A Group of People Standing]
Contesting Algorithmic Vision
A group of people standing is a generic alt-text description on Instagram, based on AI, designed for accessibility. The starting point of the workshop is that today image circulation depends less on human emotions and meanings, and more on machine vision mechanisms that tag, classify and rank visual content. However, describing what appears in an image is an interpretive act that is inherently unstable and often contested. Research by Prof. Anat Ben-David and Noa Turgeman found that Instagram's algorithmic alt-text consistently flattens cultural, social and political meanings for optimization and safety, avoiding controversy and maintaining neutrality. What are the social and political implications of algorithmic image interpretation? How do these invisible mechanisms shape collective imagination and understanding of political and climate crises? Beyond that - what are the broader implications of these mechanisms in the development of artificial intelligence, when algorithmic alt-text is fed into language models and in a feedback process reshapes our knowledge and interpretation of reality? This workshop seeks to create a synthesis between theoretical research and communication design, establishing visual knowledge design as methodology for understanding invisible mechanisms, developing visual critique, and promoting digital literacy. It is organized by the M.Des program in Visual Communication, Bezalel Academy of Art and Design Jerusalem and held in collaboration with the Goethe Institute's residency program.
