The rise of image generation Artificial Intelligence (AI) models trained on an unedited corpus of data from media sources including news outlets and archives has revealed the direct effects of media misrepresentation through the eyes of generative machines.
For underrepresented groups, this leap into the future further deepens the digital and social divide and increases the risk of misclassification and underrepresentation while strengthening long-held stereotypes. With the first magazine cover generated by AI already on shelves, machine learning models imbued with the biases that media have historically been composed of are again being used on the very populations their training datasets have discriminated against.
Through an interactive installation, we ask, how can we use data and dialogue to harness our collective knowledge to arrive at an alternative approach to the way underrepresented groups are portrayed in the media and how this informs the way images are generated by machines and used?
Babusi Nyoni (1988) is a creative technologist and founder of Tripleblack Agency who develops new work at the nexus of emerging technology and the needs of underrepresented communities.
affect lab is an award winning research practice and creative studio led by Dr. Natalie Dixon and Klasien van de Zandschulp that offers key cultural insights through design research, immersive storytelling and experimental field work.
Cigdem Yuksel (1989) is a photojournalist and winner of the prestigious Silver Camera prize (2016) with her poignant portrait series of Syrian refugee child labourers in Turkey.