Thursday, February 11, 2010

Secure natural commmunications - elaboration

Until now, my description of secure natural communications has been conceptual, with a few examples. Before we go further it is useful to tease out more of what the words mean in the original statement describing secure natural communications:

Any player (person or device) can have a rich communications experience at will with any target without encumbrances of distance or additional devices.

Here are some elaborations of the above goal.

A rich communications experience. Included in this would be:
• Finding someone I’ve never communicated with
• Having a record of the relevant threads of information that I previously have with this group.
• Having the universe of social networking capabilities at my fingertips.
• In particular, sharing all types of data, using all types of media
• Being able to tie in the history of communications that are relevant.
• A range: from purely communicating to requesting a complex computing task with sophisticated semantics
• Security and privacy

at will. Included in this would be:
• I don’t need to plan to find addressing conventions to locate the other player.
• I can use my natural means of communicating and a sensory rich world will understand my physical gestures and understand the semantics
• Speech recognition, gestures, natural language, are supported
• All media types are supported
• There are means (such as pervasive video screens or wearables) for me to get responses

with any target. Included in this are groups, so it would include:
• Being able to create new groups with security and privacy
• Being able to create ad hoc groups to achieve some new task without a great deal of overhead – and still have the rich communications experience.
• Being able to join groups and get access to their shared history

without encumbrances of distance. Included in this is:
• Assuming the very high bandwidth backbone in which bandwidth is not a constraint.
• Assuming the very high bandwidth local wireless network that gives adequate untethered bandwidth to get to the high bandwidth fiber backbone
• Developing protocols that work well with this unlimited bandwidth
• Recognizing that for certain purposes there are still constraints of propagation delay

or additional devices• Implies that there are natural user interfaces (speech, gesture) that can be relied on; i.e. the technology is good enough
• Implies that when using such natural interfaces in a noisy area that my specific sentences and commands will be heard
• Understands that this might require nano-scale wearable microphones
• Any language
• Recognizes that many people will want to continue to use keyboard for input
• Recognizes that a variety of display and storage devices will continue to be needed

This provides a more complete description by example of the type of capability that is desired. Next posting I’ll provide a little more general background, before moving to a more detailed discussion of the constructs that can deliver this and the technologies that are needed.


  1. Some things you may want to consider:
    1) Without addressing how will a user be able to to locate another user? Example: I use a cellphone and gtalk. I am accessible on both these places. So, how will the system try to contact me? You might say cellphone, but if I forgot the cellphone in another room, will I be contacted on the gtalk. Will the system be intelligent enough to consider scenarios like, my cellphone is low in battery then contact me on gtalk.
    2) You say support for all languages/gestures, but these languages/gestures are culturally influenced. You mentioned semantics in one of the previous posts. But, it will be hard to find one to one mapping across all languages/gestures.
    3) You mentioned the ability to tie in the "history of communications that are relevant". How will the relevancy of the communications be decided? Example: Yahoo supported transmitting video from one user to another. Now if one is using mike and the other is using both video camera/mike then how will the threads be formed? Although this may not seem to be very relevant thread of communication, but consider a distance education learning program where a professor is transmitting video lectures and simultaneously receiving questions via cellphone.
    4) Wireless is a natural resource. Remember the large spectrum wars which the telephone operators undergo, or the intervention of market designers to ensure fair auctions.
    5)Since we talk about the rich natural communication, a user can access his information using any possible device. Access to the threads will be limited by the device?
    6) Internet is also means of communication. So, are you thinking of search engines build into such system. Example: I think of some song, and if that song is free/available on one of my stored stuff, the song will be played.
    Or consider another scenario

  2. Chiku,
    These are all fantastic examples and points for consideration. In general we will build intelligence into the infrastructure to address these issues and your list of issues go right to our list of problems to solve! Some solutions are completely available today - others are largely available today, and others will require years of effort to be usable. We'll develop this more as we go along - as I say in the last sentence - there are technologies that are needed. Here are my thoughts on your specific considerations:

    1. We locate users with search today. Adding some customization to list preferences on "how to be contacted" seems solvable.
    2. Multiple languages will be difficult to perfect. Ultimately we need participants from many cultures to characterize their languages/gestures.
    3. Relevance is already addressed - imperfectly - by search engines. I see this as usable today, with plenty room for improvement over time.
    4. Yes.
    5. Today we already customize capability for devices (e.g. Web on Smartphone). Your point is the natural extension of this.
    6. Yes.
    Thanks for your feedback