Skip to main navigation Skip to search Skip to main content

Here is a dataset

Joel Stern, James Parker, Sean Dockray

Research output: Non-textual formAudio / Visual RecordingResearch

Abstract

It's a widespread myth that data is mined, like gold, coal, oil, or diamonds. This way of thinking sees data as just another natural resource in our world. But that's not true: data isn't simply mined; it's created. For example, when actresses read texts to train an AI to recognize emotions. When they perform a happy, sad, sly, or fearful demeanor for the algorithm. This audio piece is based on extensive training data for the AI. To what extent does this way of speaking shape us for the algorithm? To what extent does it mold our language? And our humanity?
Original languageEnglish
Place of PublicationGermany
PublisherDeuschlandfunk Kultur
Size36:01 minutes
Publication statusPublished - 2025

Cite this