Send message to Authors

Long abstract:

Artificial voices are now part of our media environments and public spaces. As anthropomorphic objects, they contribute to naturalizing technology and convey social representations regardless of the gains made in terms of listening comfort or clarity of enunciation. With deep fakes, the latest generation of synthesized voices is even more causing confusion in our digital identities. While these technologies rely more than ever on very large databases, made up of thousands of speakers, they simulate individuality and erase the collectives that served as data sources.

In contrast, the Bestiorobot project explores the hypothesis of an artificial voice that would sound 'collective', revealing as well as representing the groups involved in its design. To this end, we are bringing together researchers from Ircam's Analysis/Synthesis team and french pupils to work on a joint project: the creation of the artificial voice of Bestiorobot, a hybrid and monstrous character invented in class. Our contribution to the panel will provide an opportunity to share different visual and audio versions of the bestiorobot, each exploring different ways of (mis)using the available voice technologies, to open up the path of what we could name ‘a democratic handcrafted voice’.

This project, at the crossroads of communication studies, anthropology and design, explores central issues in STS: sociotechnical imaginaries, relationship between scientists and society, transparency and democratization of digital technologies as well as the political significance of the aesthetics of sensitive interfaces.