Digitised archives now have the potential to build sustainable connections with source and stakeholder communities well beyond Europe, as well as within it. Stakeholder sensitivities, copyright laws, and restitution agreements regulate the use of archives. Yet social media and AI have opened up new ways of circulating digitised archive materials in ways that are hard to regulate. Meanwhile, historians, memory and media scholars are still grappling with the ‘analogue’ complexities of reconstructing historical contexts and explaining the contested use of visual sources in the present.
Mass digitisation of visual archives and the proliferation of AI raises the questions: who are digital archives for? What social and ‘scientific’ purposes do visual AI tools serve? And how might this impact how images are used as representations of the past? To what extent are AI tools actually driving digitisation of visual archives rather than serving this process? Computational methods have great potential for generating new taxonomies and turning ‘primary sources’ into ‘data’ that can be analysed at scale. But is a technology-driven framework producing ‘better’ history, and who is it improving accessibility for? Generative AI already has applications to harvesting ‘stock’ web-based images to produce new or edited visual icons of violence, atrocity, genocide, colonialism or war. If the extant icons of violence already circulate free from historical context, and constitute a narrow range of examples to begin with, what will new ‘averages’ produce, and what does that mean for the way these histories are perceived?
Attendance is free, but registration is required or email m.a.van.maanen@vu.nl.
Supported by: Supported by: CLUE+, Peace and Conflict Studies Centre (Vrije Universiteit Amsterdam) and The History, Memory and Decolonial Futures Collective.
Image credits: Wulia, T, 'Behind the scenes of Absence' in Substantia series (2023). Instagram @tintinwulia.