Twentieth International Conference on New Directions in the Humanities

Abstract

While humanity has surrendered, not without initial resistance, to automation of processes and workflows, the idea of ‘general AI,’ technology with the capacity to learn, create, and stand in place of human relationships in imperceptible and unpredictable ways are still met with mixed feelings of anticipation and fear, provoking contemplations on the redefinition of ‘human.’ In this context, the digital humanities can play a crucial role in the imagination and critique of such technologies, and decisively influence their development into the next stage for the decades to come. Critical Code Studies, as a field within the digital humanities, provide an intervention that can potentially highlight the gendered, racial and class ideological foundations of affective AI technology, code writing and its amplifying discrimination mechanisms, and demonstrate its imbalanced impact across populations. Based on this code-centered interpretive methodological framework, my presentation will take a ‘friendship’ chatbot, Replika AI, as a case study of neural network architecture machine-learning programming, its history and imaginations, and the multiplicity of sites that researching it can occupy. To early mythological imaginations of such affective technology and the contested territory that constitutes its training data body, the question of affect, rather than feelings or emotions, offers an emerging space of action for humanities research and activism. This is because its modern exclusion from bodies of knowledge has allowed it to thrive within underepresented in technological spaces groups, but also because its transmission and translation still constitute equally unpredictable as market-valued spaces with ambivalent practical implications.