Abstract: Computer methods and systems allow for selective control of the audio and visual output of users in a virtual conference room. For example, users may be grouped into groups, and users within a group may see and hear other users within that group more prominently than users outside the group. Groups may be associated with other groups as peripheral, and users in a group associated as peripheral may see and hear other users in the peripheral group less prominently. Additionally, a user may be designated as a main speaker, and all users may see and hear that main speaker more prominently than other users in the virtual conference room. Further, a user may direct audio input to one or more users through a whisper. AI may be used to improve the system.