Interesting microsoft virtual world developments?

Yes this is a blog written by IBMers, but it would be wrong to not comment on things happening in an industry that we helped energize simply because a perceived competitor is doing some unusual things.
Remember, these are my thoughts not IBMs.
The last few days have seen a plethora of virtual world pitches, reports, articles and blog posts around certain types of virtual world platform.
The first was over at @monkchips a.k.a James Governor analyst blog around a visit to microsoft to see about the ESP platform. This appears to be a high fidelity simulation platform and toolkit.
The second was widely reported but Wagner’s new world notes is the one most of the metarati will have read on the matter. This centres around some statments by Craig Mundie that avatar based interaction was of limited interest and really it was photosynth that was the way forward, modelling the real world from photos.
The third is the current blackout on Xbox Live for a system upgrade, widely expected to be the bassis for the new avatar based experience (a mix of Mii and PS3 Home).
The fourth is the OpenSim powered project manhattan, with some very clear Microsoft flavour to it.
So comments I have left around various places on all this are worth tying up.
Whilst there may appear to be conflicting messages, in particular the quote about limited interest in avatar based interactions yet it really may just come from a large corporation not having a central voice. Who is the microsoft metaverse evangelist. Maybe they need to appoint one?
That aside, the interesting thing here is that all the dicussion is not about why would anyone want a virtual world, but instead what sort is best. That seems to be progress?
Also many of the photosynth and ESP elements are geared around mirror worlds. A quadrant of the industry, but by no means the only reason for virtual worlds to exist. They form a great basis too for augmented reality. There is no need to rebuild the world as it is today, but we can enhance it by seeing the data around us. Human communication and in particular the non verbal communication of standing near someone may be far more engaging than a fully realistic model of somewhere that already exists. Though that is a different reason.
The realistic nature of interactions in a model may or may not be important. Yes in a flight sim or fire training it is important accuracy and fidelity counts, but a cartoon style interaction may be all that is needed to get a point across in a meeting. As we really should link more to the games industry, not all games are photo realistic. Not all films are purely based on real world locations. In a user generated environment if the backdrop is too real, you just can build any old thing as it would look out of place.
So it seems there is room for everything and platform arguments are just the usual IT problem that we face, which again sounds good to me, we are in normal territory.
I again point back to my reverse ICE model of interaction different modes of interaction have very different needs across the spectrum. Its not worth dissing any one of them, they are all relevant.

13 thoughts on “Interesting microsoft virtual world developments?

  1. Revolution in Interactions for the Virtual World

    The success of Virtual Worlds for business services will be dependent upon the strength of interactions. The reality today, is that most services virtual buildings are empty except for events. This is not conducive for visitors who want to interact.

    Why is this happening?

    The costs of staffing just one avatar is simply to expensive for a 24*265 service. Each Avatar needs to be supported by say a staff of five people plus a supervisor costing $300,000+ a year to cover 24*365. But a service with just one staffed avatar would not work. So even a small team of avatars is going to cost over a $1m a year in staffing costs!

    The economics and indeed logistics are even more worrying when the avatar represents a subject matter expert who then needs to handover to other experts when appropriate.

    Interactions are fundamental to Real World (refer to McKinsey paper ‘The Next Revolution in Interactions’ or the Economist paper ‘The Future of Marketing: from Monologue to Dialogue’) and the Virtual World.

    The Virtual World has four primary types of interactions to power Avatars:

    1. humans (the norm today) – natural language or scripted language

    2. artificial intelligence broad and shallow – natural language

    3. artificial intelligence broad and deep – scripted language (my specialization)

    4. artificial intelligence narrow and deep – specialized ai application

    Interactions are supported by powerful socioeconomic formulae and apply to the Virtual World in the same way as for the Real World.

    Virtual Worlds inevitably will be an ecosystem of avatars powered by humans and artificial intelligence.

  2. I agree that there is a cost to staffing, but that is primarily one for a particular event or store or public facing presence. The IBM business centre in SL is integrated to the rest of IBM.com, so you can call the next available IBMer.
    AI does and will play a significant part. In games the mass of AI’s albeit relative simple in many cases add to the experience and immersion. Human to human competition though has yet another flavour to it.
    There is therefore room for all disciplines. Just as virtual worlds have introduced non gamerse to 3d, so they would start to introduce people to expressions of AI.
    For my focus on the internals of the enterprise the people are already here, they are already interacting. What they will need is boththe channels to communicate and the intelligence support to help understand, share and solve their business problems.
    I like the eco system statement too 🙂

  3. I agree fully. My response above was narrow in context to one element of the earlier blog “…………..statments by Craig Mundie that avatar based interaction was of limited interest …… “. I simply reacted negatively to a statement that does not reflect business needs for the services sector.

    Keep up the good work as we need all the champions we can get helping move towards a greater multi dimensional experience.

  4. As both an AI enthusiast and an AI skeptic 🙂 I would love to see even one example of a usefuli AI-powered avatar in some virtual world somewhere. The closest I’ve ever seen just puts out a pre-canned paragraph when someone utters any of the corresponding pre-chosen keywords in its hearing; very “A”, not at all “I”, and no more useful than a “click here for more information” menu.

    It’s great fun to talk about using AI in virtual worlds, but it’s not going to work until we can actually do AI. 🙂 And we can’t yet.

  5. David, to reply on your comment, AI in virtual worlds is just starting to exist. A company developed a virtual pet (dog) that completely is beyond nintendogs for example. This virtual pet already functions on the multiverse platform where it was developed in first instance and if I am correct is currently being worked on to be compatible with RealXtend and should become to opensim as well in the future. The company who made the virtual pet which can be trained, evolves in skill as the owner of the pet spends time with it is going to be open sourced. This will be a first example of AI inside a virtual world. Things like this still have a long way to go but they will become possible and new technical solutions will come out of them.

    I looked it up for you, the company is Novamente you can read more about AI here http://technology.newscientist.com/article/mg19726495.700-virtual-pets-can-learn-just-like-babies.html

  6. Pingback: UgoTrade » Blog Archive » Tim O’Reilly: “Instrumenting the World”

Comments are closed.