In the current situation, which means i’m in home office since mid of march, all video conference platforms are lacking major social interation parts like looking at someone, having multiple conversations run in parallel, etc.
What i’d like to have is a “real conference” option that allows you to interact nearly natual (clearly it will lack gestures). Two example szenarios:
- You are in a idea finding meeting and one central discussion breaks up in multiple smaller ones, but you can at any time join another discussion at the table to toss in your thoughts to that idea
- You are at lunch/dinner with some colleagues, which normally will result in 1-3 parallel conversation-topics, but you can still leave a topic and shortly join another one when its more interesting to you
Things such a video call need to tackle
- permanently updates conversation groups in a meeting (figure out who is interacting with whom)
- show this grouping to the user. who is looking at you and who is listening to you (2 differnt indicators)
- reduce audio and video signal for other groups, but not mute/disable video
- hands free - e.g. just use eye focus to determine the groups
This is of course no easy problem and i already saw you have social interaction as an example challenge on this months Hackfest (https://mattermost.com/hackfest2020/)
Probably you’ll need real-time ML for eye-tracking and clustering the groups.
In such a meeting it might also be better to not share a screen (i.e. grouping will be disabled when someone shares the screen) or have more than X people at the same time, because they all need to fit on everyones screen.
What do you think about it? - I guess a lot of companies would like it