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

Re: Self-Aware Solutions (was: The moral questions)

Expand Messages
  • Ali Arsanjani
    To me, self-awareness implies the existence of at least two levels: first a level at which the software ordinarily functions, through the application of
    Message 1 of 4 , Dec 9, 2001
    • 0 Attachment
      To me, "self-awareness" implies the existence of at least two levels: first
      a level at which the software ordinarily functions, through the application
      of [business] rules and another level in which we have meta-data, which
      describes the first layer.

      When a software component (e.g., an Enterprise Component) finds itself in a
      given context (checks its state and determines this based on its
      Configurable Profile which stores its meta-data, it then determines which of
      a set of [Business] Rules to apply at three levels: macro (workflow/business
      flow) across subsystems boundaries, within its own boundaries (collaboration
      rules) and has the ability to adapt to changes in its environment through
      having Manners (knowing which rules to apply in a given context).

      The component thus can monitor its own handling of requests and events by
      accessing its Manners (sometimes stored as meta-data in its Configurable
      Profile).

      Its Manners can be represented as a grammar, in which case we are doing
      grammar-oriented object design.

      Being self-aware does not necessarily imply the "sense of other"; one
      (human) can be self-aware without a sense of other. In fact research in
      consciousness indicates that self-awareness occurs when the mind is in its
      state of least excitation and at this fine level can experience itself
      without reference to anything else. Thus, it is called pure consciousness.

      Regards,

      Ali
      ----- Original Message -----
      From: "David Cymbala" <dcymbala@...>
      To: "David Parker" <david_parker@...>;
      <patterns-discussion@...>
      Sent: Sunday, December 09, 2001 1:05 AM
      Subject: Self-Aware Solutions (was: The moral questions)


      > David Parker wrote:
      > >>Real self-awareness, as understood by humans, in my opinion, requires
      the
      > >>existence of a society around you--only by distinguishing yourself from
      > >>others can you become self-aware.
      > >
      > >Why do you say that? If it is true then it tells us what sorts of
      problems
      > >require self-aware systems as solutions.
      >
      > Federated systems that have hot stand-by
      > capabilities might be an example of this.
      >
      > For example:
      > Each 'node' in the system would remain
      > aware of the other 'nodes' and assume the
      > responsibility of providing services if
      > it doesn't get timely feedback from the
      > current primary provider of the services.
      >
      > There are many other variations of this,
      > such as dynamic scaling of services
      > when demand is high, etc.
      >
      > I suspect that such systems have a
      > bit of artificial self-awareness in that
      > each part bases its activity on the
      > activity of other parts of the system.
      >
      > Is there an implict notion of 'self' when
      > there is also a notion of 'other'?
      >
      > -David
      >
      >
    • Ali Arsanjani
      The meta level is not described by anything metaphysical :-) It is described by Grammar-oriented Object Design in terms of a domain-specific business language.
      Message 2 of 4 , Dec 9, 2001
      • 0 Attachment
        The meta level is not described by anything metaphysical :-) It is described
        by Grammar-oriented Object Design in terms of a domain-specific business
        language.

        Regards,

        Ali
        ----- Original Message -----
        From: "Philip Eskelin" <philip@...>
        To: "Ali Arsanjani" <aarsanjani@...>; <patterns-discussion@...>
        Cc: <feyerabend-project@yahoogroups.com>
        Sent: Sunday, December 09, 2001 10:14 AM
        Subject: RE: Self-Aware Solutions (was: The moral questions)


        > Ali wrote:
        > > To me, "self-awareness" implies the existence of at least two
        > > levels: first a level at which the software ordinarily functions,
        > > through the application of [business] rules and another level
        > > in which we have meta-data, which describes the first layer.
        >
        > What language describes the second level? I think your first problem is
        the
        > "meta" part. The answer probably boils down to metaphysics. One of the
        > classic problems in metaphysical philosophy is that of achieving the
        > "ontological parsimony." Modal Realism and many of its spin-offs (e.g.,
        > Modal Fictionalism, Actualism, Presentism, etc.) end up overcomplicating
        the
        > notion of an ontology. It becomes difficult or impossible to commit to
        one
        > without skepticism. Philosophers continue to strive for a philosophy that
        > everyone feels comfortable committing to.
        >
        > The problem of description that you reference above is the same problem
        > metaphysicists are trying to solve: how do you achieve ontological
        > commitment. The challenge of acknowledging all the objects that exist in
        > our world along with their intrinsic and extrinisic properties is very
        > similar to the challenge you present above. It's a big one. If you can
        > figure out a way to get everyone to commit to the ontology of your meta
        > language, then I think you will be off to a great start.
        >
        > Thanks,
        > Philip
        >
        >
        >
      • Philip Eskelin
        ... This response leads me to believe that you misinterpreted my response to your email earlier today. I see that your background focuses on reuse, patterns,
        Message 3 of 4 , Dec 10, 2001
        • 0 Attachment
          Ali wrote:
          > The meta level is not described by anything metaphysical :-)

          This response leads me to believe that you misinterpreted my response to
          your email earlier today. I see that your background focuses on reuse,
          patterns, rules, and CBD. I have a similar background, but have recently
          really began to learn more about philosophy with a concentration in
          metaphysics. In doing this, I have discovered that the problems they are
          working are very compatible to the one I believe has been discussed on this
          list.

          > It is described by Grammar-oriented Object Design in terms
          > of a domain-specific business language.

          This makes me conclude that you either 1) have a breakthrough solution that
          will change the way we use computers and live our lives, or 2) have an
          approach to the problem (like countless others in this industry) that hasn't
          really gotten us anywhere. Nothing personal. My understanding of the
          Feyerabend Project is that they are motivated by the latter.

          1. What precisely do you mean by "grammar-oriented object design"?

          2. What are the terms in a domain-specific business language
          described in terms of?

          I would also be interested in what some of your experiences at IBM GS are in
          applying this in your engagements, and whether you were able to figure out a
          way to justify its ROI to your sponsors.

          Regards,
          Philip Eskelin
        • Ali Arsanjani
          ... that ... hasn t ... I think there are usually more answers to a question than its extremes. Between the 1) and 2) that you mentioned above are a spectrum
          Message 4 of 4 , Dec 10, 2001
          • 0 Attachment
            Phil Wrote:

            > This makes me conclude that you either 1) have a breakthrough solution
            that
            > will change the way we use computers and live our lives, or 2) have an
            > approach to the problem (like countless others in this industry) that
            hasn't
            > really gotten us anywhere. Nothing personal. My understanding of the
            > Feyerabend Project is that they are motivated by the latter.

            I think there are usually more answers to a question than its extremes.
            Between the 1) and 2) that you mentioned above are a spectrum of
            increasingly
            valuable contributations that we should try and avoid bulk-labeling.

            Instead, consider rhe fact that each solution presents a balancing of some
            forces
            and imbalances others, much like a pattern itself.

            > 1. What precisely do you mean by "grammar-oriented object design"?
            > 2. What are the terms in a domain-specific business language
            > described in terms of?

            Check out some of papers and the link to GOOD on my web page at
            www.arsanjani.org


            > I would also be interested in what some of your experiences at IBM GS are
            in
            > applying this in your engagements, and whether you were able to figure out
            a
            > way to justify its ROI to your sponsors.

            Sure. Application in enagagements has been going on for some while (since
            1998); sometimes
            I don't even mention WHAT we are doing (labeling it) versus "just doing it";
            applying the methods
            and techniques to get results.

            But we can take that offline.

            Regards,

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