"I can't lie anymore!": The implications of location automation for mobile social applications

Abstract

Human factors research has shown that automation is a mixed blessing. It changes the role of the human in the loop with effects on understanding, errors, control, skill, vigilance, and ultimately trust and usefulness. We raise the issue that many current mobile applications involve mechanisms that surreptitiously collect and propagate location information among users and we provide results from the first systematic real world study of the matter. Our observations come from a case study of Jaiku, a mobile microblogging service that automates disclosure and diffusion of location information. Three user groups in Finland and California used Jaiku for several months. The results reveal issues related to control, understanding, emergent practices, and privacy. The results convey that unsuitable automated features can preclude use in a group. While one group found automated features useful, and another was indifferent toward it, the third group stopped using the application almost entirely. To conclude, we discuss the need for user-centered development of automated features in location-based services.