Loading ...
Sorry, an error occurred while loading the content.

My 2 cents on More Preparation Notes

Expand Messages
  • pam rostal
    I think a reset in computing makes sense, but it may be starting without us. The use of software agents to mediate between humans and their computing
    Message 1 of 1 , Apr 8, 2001
    • 0 Attachment

      I think a reset in computing makes sense, but it may be starting without us. The use of software agents to mediate between humans and their computing technology may allow people to satisfy their system needs much more naturally. (If existing approaches don't do it, we should look for something that does.)  If a user can tell agents what goals he/she want to achieve, the agents can probe their environment to discover whether or how the goals have been implemented previously and then deliver proposed process definitions to the user. If the user approves one and meets security and operational criteria, the agent could create or find a user interface to match, and the user could determine how to integrate the process into his/her environment. If the user fails to meet the criteria, the agent could provide instructions for resolving the situation (and perhaps alert those who created the criteria of a possible mismatch between responsibilities and constraints within the organization). If the process doesn’t exist yet, the user may be asked to justify it as part of continuous process improvement in his/her department.

      Once the user has an adequate interface, it could be incorporated into a workspace using some variation on the gadget theme used by Plumtree personalization software. Plumtree allows people to try their software online at

      www.corporateportal.com. (They are unusual in that they seem to have heeded parts of the Cluetrain Manifesto – their discussion thread gadget features actual user postings, including one that recommends an evaluator choose a different product.)

      Although there is nothing revolutionary about this, it does reflect a slightly different world than we live in today.  It's almost like building software inside out.  Users would determine when and how to use technology in their environment. By using pre-packaged interactions, they are assured that the applicable business rules have been implemented and security policies have been followed (software agents could also ensure this). If a particular process is found to be very useful, users themselves may form process exchanges to proselytize the most efficient way they’ve found to do their work. (Plumtree lets people submit their adapters for certification and then hosts an adapter exchange for companies that want to use the certified products.)

      Some developers might find themselves acting as evangelists for user empowerment by living with the users as resources (instead of the usual ‘send a user over to live with us for a while’ model). These developers could advocate for the development of additional integration events and/or behaviors or might be able to implement the enhancements themselves in a matter of weeks because the integration platform and possibly the required information already exist.  The new interactions would then become additions to the behavior bank on which all agents can draw as they try to achieve their goals in a continually evolving environment.

      By separating out the agent-computer communications from the agent-user communications, developers (and possibly languages) could specialize in agent-integration platform interactions, application-integration platform interactions or user-agent interactions.  This might let people find a better match with what they want to do with their talents, while also letting them evolve to different environments as they progress in their careers.

      Pam

      P.S.  There is an implicit assumption that our primary goal in using technology is to empower users in their environments, as opposed to having a primary goal of monitoring them or enforcing corporate policies.  If this is not true, then a different kind of reset should take place.

    Your message has been successfully submitted and would be delivered to recipients shortly.