I have a feeling that this sort of live video facial analytics will spur the creation of live filters that will mask your actual response. After all, if your b

I have a feeling that this sort of live video facial analytics will spur the creation of live filters that will mask your actual response. After all, if your boss might act on what an AI says your reaction is to a conversation they are having with you (such as scepticism, reluctance, disgust, sarcasm, boredom, insincerity, exhaustion, etc.), then you have a strong personal motive to use a solution that adjusts the results of that detection to put your 'best foot forward', or at least that guides you toward projecting the desired affect (IOW, enthusiasm, agreement, sincerity, excitement, alertness, etc. EDIT: eg. https://www.newscientist.com/article/2267298-men-who-are-bad-at-public-speaking-can-get-help-from-a-virtual-clone/ https://www.newscientist.com/article/2267298-men-who-are-bad...). The popularity of adjusting trivial aspects of how you present yourself (ie. background replacement) can probably serve as an indicator for the eventual popularity of filters for more actionable aspects of personal video streams. In turn, I would then expect filter detection capabilities to gain a foothold. A nice little adversarial arms-race with the potential for creating a fortune or two, and possibly a significant new non-gaming marketing niche for graphics/AI acceleration hardware. EDIT 2: Who knows, this sort of functionality (both analytics and filters) might have applicability to other situations too, like negotiations, where a poker-face (or it's opposite) or reading 'tells' could come in handy.