knowquestion :: Stephen Bounds

As a generalisation, computer applications involving user interaction are either intended to augment human intelligence and activity, or to be a service that delivers a pre-determined capability for people. By analogy, it's the difference between an exoskeleton and a car.

For a car, you can catalogue and test an exhaustive list of manufacturer-approved use cases. For an exoskeleton, you can't because the limits of its use are the same as those of human ingenuity. While an exoskeleton will have system limits beyond which it will break, that is different. (And yes, you can do unforeseen things with a car, but no-one will tell you they were a good idea.)

Twitter is an augmentation service, as is email; a travel request system is a service delivering a capability. A case management system is an odd hybrid of the two, which possibly explains why they can never really please everyone.

Limits of augmentation systems are to be worked around; services consciously draw lines which shouldn't be crossed to improve simplicity and repeatability.

Information Technology planning often tries to treat everything into a "service" system. It's an easy trap to fall into; I've done it myself. But taking this approach leaves augmentation systems in an awkward netherworld: designed to operate without constraint, in a world where access to data is corralled through services expecting a non-intermediated human at the other end.

Even the best attempts to bridge this world like IFTTT are very clunky and clearly kludges. I don't have a good solution for the problem but I do think it's becoming bigger.

(NB: This is an edited version of my LinkedIn comment here.)