Algorithmic biases: An archive against the odds

Description

In a time in which artificial intelligence co-shapes our everyday lives, a new archive is emerging: Not on shelves, but in datasets, server rooms, and models. This archive has no consciousness and no morals, only probabilities. It does not know what it does not know. But it helps decide what remains visible. Queer and non-white perspectives are algorithmically considered improbable and are missing. Where knowledge critical of discrimination is lacking, the archive of the future also remains silent. In this silence lies a new form of invisibility: it is technical, methodical, and frighteningly friendly. In my master’s thesis, a resistant position emerges from this, which does not simply want to make visible, but rather wants to make the struggle for visibility tangible: How do we fight back against forgetting?