keeps the models relevant. you need to capture that human data, you need to capture that natural movement, their nuances, their body language, their posture, their gesticulation, the personal space, it gives you a sense of that person in that space. so, even if the avatar doesn't look like you, people can still look at it and be, like, that is your walk, that is the way you stand, i know that is you. it draws you in, it keeps you immersed in the experience and makes it more believable and enjoyable, as a consumer. what we've built on top of the ai are a number of processing engines on top. we have biomechanical rules written into them so there so there are kinematic models, holonomic constraints which are what the human body can do, in terms of how the joints can move. so, for instance, if i bent my elbow like this, you can't bend it past a certain point of extension. rules like that are written into the system. we can work out where your feet are, and when you are moving around the volume, what that allows us to do is basically create a map of the volume and you within it. there are six cameras in this setup, and that is enough to track about five people in this space, although they may bump into each other a bit.