Reply #6 to other comments posted about my proposal; February 1998


After having more time to read and consider your (very absorbing) ecology paper and other references, and to better consider the relation between my model and your ideas, I wanted to try to answer your questions better than I did last time.

A model like mine is the only thing I have encountered that has a good chance of meeting your criteria.

It would absolutely have to be able to sense changing environmental factors from a variety of inputs. Otherwise it would soon be as irrelevant as an old newspaper.

It also would have to be able to consider multiple interpretations of the data presented by humans--comparing them to the complex and continuously updated models it has on file, and giving feedback. The feedback would be in the form of showing any conflicts between the interpretations and hard data in the model, or it I suppose could take a variety of other forms.

The worth of such a model in the "wicked environments" of business change would be to facilitate that constant questioning of "standard practices"--it could enable automatic and constant comparisons of the standard practices to the constantly changing world, as reflected in the constantly updated model (which, ideally, would mirror the world in an increasingly accurate way). It could also powerfully facilitate human-initiated questioning/comparison of the standard practices, versus any other ideas the human pay put to it. In other words, in its theoretical fulness, the model is a complete reflection of the real world (including all the nonphysical attributes of the "real world") and can allow comparison of anything with it.

I think besides enabling the "loose-tight" approach to practices, one other need your paper brings to mind is technological--an ability to develop and deploy new software systems in near-real-time, nearly as fast as the needs are perceived and articulated. One (long-term) advantage of a mature version of this model is that it could conceivably supplant other forms of application development for some purposes. Specifically, since by design it contains a comprehensive and constantly updated object model for everything we have conceived, those same objects can be used to throw together new systems on an ad-hoc basis, as simply as you can manipulate digital images with a VR glove. Construction of software becomes more like construction with physical objects. You handle, modify, and assemble representations of software components (which are become nearly innumerable) with a similar natural feel to handling and manipulating physical components in the real world.

So, I don't think the model is constrained, except in finding someone interested enough to support its active development.

I sincerely solicit your candid reactions & thoughts. Thanks.