Hxr Voice explores various ways race, gender and development of algorithms in contemporary computational systems such as machine learning and artificial intelligence are intertwined with each other. With a critical perspective on how these algorithms shape transformations in art and culture and how in turn art and culture can shape algorithms, the project studies different tactics to engage with inherent biases in data and technology in creative and community-focussed ways. With a focus on staging encounters with algortihms, Hxr Voice is conceived as a series of works to bring to the forefront the role of active supervision and collaboration between human experience and the potential of algortihms. For this colloquium, the artists are focussing on the growing sub-disciplines of voice, accent and speech profiling systems. They will present their recently produced audiovisual performative piece [DataTone] where they take a closer look at the impact of machine learning on the everday use of voice recognition technology, revealing the sources of inherent bias of techniques such as automated speech recognition and automated speech analysis. Through an expanded lecture performance, they will attempt to spark a dialogue on the development of computer listening systems that are implicitly or explicitly biased about the accent of the voice, but can be made aware of the subtle and subjective complexities of the soundscape of human voice.
Upcoming: live collaborative performance at HEK Basel, 7 – 9 PM, 5th April 2023.
Leave a Reply