We consider how we organize work when the product of that work is sustained and relevant computation.
An organization is to deliver computation as a product valued in excess of its cost by a market.
An organization acquires market advantage by collecting and distributing skills more proficiently than competitors.
An organization is limited in its ability to comprehend and then direct its own response to opportunity.
# Network
We envision a network of documents, a hypertext, that provides the persistent, comprehensible, and transformable basis of organization.
We confer onto these documents the ability to describe work and then to do work. Documents model the work they do with high fidelity by being one in the same.
An individual within the organization operates within a limited scope where it is safe to revise documents, assess the impact of their revisions, and then propagate their improvements to the benefit of the whole organization.
An individual cannot be expected to predict with any more certainty the total impact of a revision beyond that provided by the model. Thus the continued refinement of models in response to real events is necessary to maintain useful fidelity.
# Experience
An individual can reasonably ask, what became of my modifications when confronted with reality? Did it deviate from expectation? Did it serve the organization? From this one gains experience.
An individual must transcend the scope within which they work in order to understand its impact in reality. The organisation is advantaged by enhancing this mobility. The diffusion rate of skills depends on individuals finding, comprehending, and then owning the documents that model computation.
We will be helped in understanding the possibilities of this approach if we think of documents as things, mechanisms to be extracted, studied, adjusted and then reinstalled as part of more complex things. When we describe a thing as a document we assert that in operation it will be as observable, comprehensible and as intellectually expanding as what we might write about the thing but without separating the writing from the thing.
See Social Chromatography where we further dissect these document/things.
# Practice
How would federated wiki host such a network of things?
We describe activity in wiki as realizing various workflows. Pages are created, edited, connected and transported in order to effect change. Purpose driven work would follow established workflows. Curiosity driven work would remain free to explore alternatives.
Projects would be chartered to effect change with established workflows. Templates extracted from similar projects would guide the creation of pages.
Projects would be prioritized based on business opportunity, risk and available staff familiar with the necessary workflows. Progress will be manifest within a workflow and doubt assuaged by forking and running models and/or full computations in test environments.
Technology specific markup would describe how computational resources are to be configured in a given environment. Limits, rates and reliability estimates would come from models and inform configurations directly.
Technology in operation would include measured limits, rates and reliability in contrast to prediction. These would be rolled up and recorded consistent with higher level models in the system.