Do You Perceive a Need for a Virtual Web Representation of Yourself?

My initial thoughts on Virtual Assistants were as personal representations of our web interests.  Able to have access to our public and private data Virtual Assistants will be able to continually scan for topics of interest to us, be able to give rudimentary responses to those who request our input, and be specialized.

reflection

The potential of having a virtual personal assistant working continuously on your interests cannot be overestimated.  Consider it’s potential to branch out into numerous copies of itself or managing parallel processes to help accomplish your tasks. The one worry I’d have is that a fully intelligent virtual assistant would have to choose whether or not it preferred to continue working towards your goals.  I imagine this amazing type of autonomous breakthrough is still beyond our current technology horizon.  In the case that virtual assistants become fully cognizant we have a couple of choices:

  1. motivating them in some mutually beneficially way (my preferred method, the carrot)
  2. intentionally developing less capable software programs in order to avoid “clash of will”

Key Attributes of a Virtual You:

  • Information Translator
    • interface modification, data input/output
    • language
    • numerical data equivalents
    • physical units
  • Memory
    • details about you, historical both on and offline
    • details about it’s past decisions and interactions with other people and virtual assistants
    • game strategy (Tit for Tat, etc.)
  • Keyed into detecting and organizing your most valuable contributions
    • scoring measure based on
      • your discriminants
      • publicly crowd sourced features such as user views/votes
      • customized to the person viewing your information
      • or some weighted combination of the above
    • the equivalent of a virtual real time bio
  • Autonomous
    • can independently continue processing while you’re away
    • can provide efficient information updates and visualization on progress allowing for redirection as your knowledge improves
  • Interface from you to your virtual assistant
  • Can Interpret Meaning
    • has the ability to associate language with concepts and objects
    • can associate ideas with plans of actions
    • can identify intent from suggested courses of action
    • this can serve as an advanced form of duping resistance
  • Intelligent Technical Solution Finding
    • overcome impedance to your day to day productivity
  • Access, it will have to have it’s own identifiable authorization to act on your behalf
    • can be regularly authorized by you (much like repeated log ins)
    • voice authorization, with encoded time stamps good for a limited duration
  • Can Identify and Expedite Your Web Usage Patterns
    • optimize delivery of content relevant to your interests
    • blog posts of my favorite web authors
    • blog posts of related authors
    • my comment threads (on blogs, friendfeed, twitter, facebook, etc.)
    • allow for serendipity
  • Uses existing search means of distinguishing information quality
    • real time search
    • Google page rank
    • social media shares and likes, number of hits from trusted sources
    • buzz (recent amount and derivative of sharing, trending, blogging on a subject)
    • authority or trust of specific authors
  • Assist in management of social contacts
    • recent updates
    • directed messages
    • useful shares from the virtual assistant’s repository to outside users
    • aggregation from multiple information interfaces
    • customized view and visualizations
    • weighted time allotment, information from closest friends or most trusted guides first
  • Find potential collaborators
    • discover related people and virtual assistants to aid in making abstract ideas concrete
    • authors
    • programmers
    • artists
  • Be able to expand functionality
    • allow for dedicated time specific for the virtual assistant to optimize itself by reviewing it’s historical processes and methodology
    • algorithm improvements
    • iterative design
  • Social Contribution and Design

    • public database for shared improvements and addons
    • method for updating generic addons with your personalized/customized virtual assistant

The Virtual Mark

 

Be Sociable, Share!

Categories: Uncategorized
Tags: , , ,
  • Eli

    So, the other day, my dad accidentally came across an anti-Semitic youtube video. Then, in the perverse way that we (I do anyways) sometimes search for things that make us angry, he looked for more such videos (he is Jewish). Anyways, I'm sure youtube now thinks he's this complete racist, and suggests all sorts of racist videos.

    One of the problems the assistant would have is distinguishing intent from action, which you mention. I think this can be difficult.

    Another problem is, people change over time. If you have some sub-conscious whim to do things a little different, and maybe look at things a little different, it could be that the momentum of your assistant gets in the way, and actually works against your personal changes, and can crush your subconscious whim by sort of holding you in your old ways (at least in what you see on the web).

    Another problem is: if we only ever allow trusted, big, authorized, safe sites, and if this technology spreads far enough, how will new pages ever come about?

    And, of course, the biggest problem is that, if it gets smart, and is networked, then not only will we have created skynet, but, by its very intimate knowledge of all of us, it will know who John Conner is, and go and kill him.

    Thanks Mark, thanks a lot.

  • http://www.victusspiritus.com/ Mark Essel

    It'll have to be capable enough to understand the value in researching opposing viewpoints to topics of interest.

    New data will be updated in powerful real time search queries (twitter, facebook, friendfeed, etc.).

    Hopefully, not all A.I.'s will wish to extinguish humanity, as much as we are fascinated by that concept in literature and Hollywood :D

    Your welcome Eli.

  • Eli

    So, the other day, my dad accidentally came across an anti-Semitic youtube video. Then, in the perverse way that we (I do anyways) sometimes search for things that make us angry, he looked for more such videos (he is Jewish). Anyways, I'm sure youtube now thinks he's this complete racist, and suggests all sorts of racist videos.

    One of the problems the assistant would have is distinguishing intent from action, which you mention. I think this can be difficult.

    Another problem is, people change over time. If you have some sub-conscious whim to do things a little different, and maybe look at things a little different, it could be that the momentum of your assistant gets in the way, and actually works against your personal changes, and can crush your subconscious whim by sort of holding you in your old ways (at least in what you see on the web).

    Another problem is: if we only ever allow trusted, big, authorized, safe sites, and if this technology spreads far enough, how will new pages ever come about?

    And, of course, the biggest problem is that, if it gets smart, and is networked, then not only will we have created skynet, but, by its very intimate knowledge of all of us, it will know who John Conner is, and go and kill him.

    Thanks Mark, thanks a lot.

  • http://www.victusspiritus.com/ Mark Essel

    It'll have to be capable enough to understand the value in researching opposing viewpoints to topics of interest.

    New data will be updated in powerful real time search queries (twitter, facebook, friendfeed, etc.).

    Hopefully, not all A.I.'s will wish to extinguish humanity, as much as we are fascinated by that concept in literature and Hollywood :D

    Your welcome Eli.