One of the latest manifestations of this phenomenon is occurring in the development of software products and services. Much like the cotton spinners of old, work that was once done by vast armies of people is being subtly replaced by new innovations that mean gigantic leaps forward in productivity.
The software testing process, in its worst manifestation groups of individuals running through manual test scripts or at its best coded regression scripts taking hours/days/weeks* to complete (*delete as appropriate), is gradually being replaced by lightning fast, change aware, massively paralleled testing tools.
The configuration of test environments, be they one machine, hundreds of machines, a replication of the web, or a replica of a customer, in its worst incarnation an (expensive) in-house army ghosting huge expanses of bare metal to make it suitable, at best a set of script-driven virtual machines, is gradually being replaced by on-demand, template and process driven, pay-as-you-go provided virtual infrastructure. Opex instead of Capex.
Building the software itself, from an individual developer feeding a compiler direct, to a process of continuous delivery that ensures continually executable software. Why do any of this? Self-executing (and self-modifying) code will be the norm for software, with the multitude of ‘evolving’ frameworks becoming the ‘machines’ that process (using free, unlimited processing power) natural and artificially intelligent languages that simply do what you say. Bye-bye abstraction, classes, objects, polymorphism and all of the other structured language paradigms that mystify layman developers of the future. Hello simple sentence coding (Siri anyone?).
The build-deploy-test mantra is “world-class” in the current perceived wisdom. But what we think of as world-class now is phenomenally behind what “future-class” will be. We’ll see macro-delivery of new and innovative (and disruptively competitive) software on the scale we can only dream of now, across all platforms and by all kinds of organisations. In researching what it means to be “world-class” I’ve found that, realistically, nobody is truly world-class. Leading-edge, perhaps, a pioneer, maybe, but definitely not world-class. If fact nobody will truly be world class as, much like during the industrial revolution when the inventor of new technology got ahead of the market for a while, technology is easy to replicate. Once someone delivers something that they believe to be world-class, everyone else will copy it, and “world-class” itself will have moved on to mean something completely different.
Let’s hope change management has improved over the years to handle the inevitable luddites….