I find concept videos like this often grossly over-simplify the problem of contextual computing, or even personalization for the matter. I think today we alrea

I find concept videos like this often grossly over-simplify the problem of contextual computing, or even personalization for the matter. I think today we already have plenty of context -- A paypass-equipped credit card can already provide context. And NFC-equipped phone, or even your wifi or bluetooth fingerprint. All of these things are already capable of providing the same level of uniqueness describing your person (Or at least arranging the transfer of enough uniqueness to qualify you). I think the biggest problem that every entrepreneur always misses is: "How do we define context?" "How do we organize it?" "How do we make sure that your personal information is open, tailored to you, and not limited in so much as Microsoft, Apple, or Facebook is willing to sell you that expression of identity which you will be LIMITED to being, as they see fit." I think it comes down to a much more broad question that I don't think the industry has yet to answer at all. What is the GRAMMAR of this context -- that will make it so fluid that everybody can integrate with it, seamlessly? This is not something the industry has tried to answer, because we're still arguing over Apache's "Web of Things" vs a Json "Web of Things". We haven't even prescribed machine-understandable meaning to these free structured Apis yet -- let alone write a system that can automatically understand what a trunk is, how it works, and the entire context and meaning of a car. This is not a move to context -- it's a limited gesture, and one that isn't based in the realm of context any more than a key understands a door.