Facebook users are presented with newsfeeds generated by an AI algorithm that is fed with different kinds of personal and contextual data. Such data determines what content is displayed in the feed and in what order. But what kind of data would users want the system to use to adjust their newsfeeds contents, if they had a choice? This installation lets people reflect on what data they share and want to share with the Facebook algorithm.
In this research project, we developed a public interactive installation to invite people to reflect on how their personal data is used by the Facebook algorithm to adjust their newsfeed contents. The installation is designed to make data more tangible and to give control back to the user. Participants can decide what data the algorithm should be fed with, by activating or deactivating certain data points. By flipping the installation, participants can operate on two different levels: (de)activating entire categories of data, or individual data points.
The installation provides the user with visual feedback on the expected quality of their newsfeed based on the data they select, through a change in light intensity. The more data points are deactivated, the lower the light intensity, which means the less engaging the newsfeed content is expected to be. Participants are encouraged to explore the best balance between control and engagement, what is acceptable to them, and what degree of control they would like to have.
If you were given a choice, what data would you want Facebook to use?
