Conversations are a daily human activity, and a crucial part of face-to-face collaboration in virtually any context. It is also a highly engaging activity that requires the full attention of participants involved in it. This is why computers have generally been perceived as intrusive in the world of human conversation, for they take some of their user's attentive focus, reducing their capacity to engage with the other. However, computers today are no longer limited to pieces of technology that we place in front of us or hold in our hands while we interact directly with them via keyboards, touch screens or other input devices. Some computers now hide in our environment, avoiding our attention, achieving whatever function is required of them without us even knowing they are there, and leaving us to focus on the tasks that are important to us. We present a system to explore the role computers can take in face-to-face conversations within the context of this new computing paradigm. Our interactive table, which we call Reflect, monitors the conversation taking place around it via embedded microphones and displays relevant information about member participation on its surface in a discreet and unobtrusive manner. We raise several questions about how such a device can be used to improve the quality of face-to-face collaboration. In particular, we explore whether or not this system is capable altering user behavior and under what conditions this is possible. We also examine whether or it is possible to achieve a change in user behavior while remaining unobtrusive. In addition, we look at the use of such a device outside the scope of face-to-face collaboration by examining its role in the world of communication training. Finally we study the transition process and the design changes needed to bring such a device out of the laboratory and into the real world. To answer these questions, we describe two user studies conducted on the Reflect table. In the first study, we show how the table can be used to promote balanced participation and we examine the conditions under which this is possible. In the second study, we test such a system's ability to change the way people speak during a conversation, and show some of the difficulties in achieving that, as well as some differences in how male and female users respond to such a device. We then take the Reflect table outside of the laboratory and explore its use in the real world. We explore the changes to the system design that are needed for such a transition to take place. We also show how the table is perceived by users outside the scope of a laboratory study.