Ok, putting aside for a moment questions of 'Why/What-the-F'?! the technology behind the Direct-To-Gob targeting is pretty ingenious - and sound-driven.
There's some seriously impressive distance and directional information extrapolated from what would appear to be a relatively lo-tech binaural microphone setup listening out for a vocal trigger... it got the cogs whirring a bit on a Monday morning... perhaps you could combine aural targeting with directional audio delivery too?
In public spaces, it'd be possible to track a voice and deliver highly localised audio information...
Or maybe in a studio situation, your audio monitors could be motorised and their precise alignment to the listener be tracked and targeted by just talking back to them?
Hmmm, I can see that being an anonymous heckler in an audience suddenly becomes a bit more risky too ;-]
Interesting use of audio tech, though.