Market, Testbed, Backroom: The Redacted Internet of Facebook's Discover

This paper uses document theory to analyze Discover, a partially free-to-use application developed by Facebook's philanthropic initiative Connectivity and released in the Philippines in May 2020. Discover's design is predicated on the conviction that access to valorized forms of technology-in this case, popular websites viewed via the Internet-promises benefits to marginalized users who are presumed to lack resources needed to fully participate in contemporary informational capitalism. Document theory in HCI provides a framework that allows us to analyze the most popular websites as rendered by Discover. We argue that Discover's logic of redaction and form moderation reproduces the very structural inequality that access to the Internet frequently claims to ameliorate. We conclude by pointing to potential applications of our approach in research at the intersection of HCI, ICT4D, and political economy.

• Human-centered computing → Empirical studies in HCI.