Notes: Microsoft’s Craig Mundie on NUIs, Avatar Kinect, Surface 2 and telepresence

Craig Mundie is the Chief Research and Strategy Officer at Microsoft Corporation. He gave an interesting talk tonight titled “More Like Us: Computing Transformed” at the University of Melbourne. Mundie discussed natural user interfaces, Avatar Kinect, Surface 2 and telepresence, and I’d like to share my brief notes and thoughts below.

An apology first though: I am not a big fan of brain dump blog posts full of bullet points (sorry Luke :)) – yet here I am:

  • Mundie’s most frequently used word was “capability”, which seemed to underline the technology-orientated nature of the talk.
  • He was still talking about “tasks”, “integration” and “automation” … I thought those were the buzzwords of the previous era of computing?
  • “The current web seems like it’s nouns only, we want to add the verbs.” This sounded really interesting but I couldn’t quite grasp the practical meaning.
  • There were several mentions of “computing power”, which in my view is already irrelevant to the larger audience.
  • The new kid on the block is “Natural User Interfaces”. I still don’t like the term – the natural connotation (pun intended) is that some computer interfaces are more natural than others – however it looks like we’re stuck with it.
  • Mundie spent most of his airtime on Avatar Kinect. Having not heard of it before my immediate reaction was “Matrox HeadCasting – 10 years later.” I’m not sure if the human interface problems of those days are fully resolved yet, however the potential reach and audience for Avatar Kinect are huge.
  • The technology is interesting and the live demo was great, but I still can’t see immediate value in telecasting in productivity applications. It’s hard enough to get value-for-time in a live four-way meeting (let alone in one with more people), and the issues are only compounded when the meeting is mediated through technology.
  • (An idea: A real killer app would be taking a meeting/discussion/telecast recording and “cutting out the boring bits” automatically. Mechanical Turk for video editing anyone?)
  • (An idea: We need to think beyond the current “single-limb” gestures like the three-finger drag – or ten finger drag – and consider them in context of and in combination with other expressions such as facial ones. How does “frown-and-drag” sound? What about “two-shoulder shrug”? You heard them here first 🙂 Bonus: Check out the brilliant Grimace project.)
  • Using Kinect as a 3D scanner is way cool. In fact, Mundie seemed to be very much in favour of Kinect Hacks and other similar projects, and suggested that further SDKs will be released within the next few months.
  • Mundie’s suggested continuum of telephone > television > telepresence is interesting. Perhaps telepresence will revolutionise the way we communicate, one day.
  • An interesting detail about Microsoft Surface 2: the “fourth component”, an IR sensor in addition to the standard RGB outputs for each pixels, allows detecting “touch” above the surface. What does this mean in practice? We regain hover! Flash on iPad 3 anyone?
  • There was a great question from the audience comparing the (perceived) ease of gesture-based interfaces to the relatively limited power they (still) have compared to old text-based interfaces. I think Mundie’s answer missed the key point but he did note that “programmers create the framework” [of use] for these new interfaces. I think this risks further division between “consumers” and “producers” in the digital society.
  • (A question: When will we see the first self-hosting 3D user interface? Self-hosting in the appropriated sense that its developers would do most of their development work using the 3D interface itself, rather than another environment.)
  • Odd spot: Somehow Mundie had managed to make his Kinect avatar look like Jakob Nielsen 🙂

Big thanks to Craig Mundie for his thought-provoking talk and for the University of Melbourne for hosting the session!

A virtual meeting organised using Microsoft's Avatar Kinect
Image:
8-way meeting organised using Avatar Kinect. Source: Microsoft

Author: Jussi

Jussi Pasanen is the founder and principal at Volkside