Competition and innovation correlate, at the very least it can be said that the former accelerates the latter. Then if the goal is to innovate–a product, a process, an output in general–it is more congenial to have two competing systems than one monolithic one. Now starting from here I generalize, I say that our system S – of people, of teams, of companies – is globally made up of N people clustered in teams of cardinality between 0 and N who intra-team share information and ideas constructively and inter-team act as a blackbox and compete on outputs without making their internal processes accessible therefore. I also say that S is fluid: there is budding i.e. detaching a subteam from the main team, there is the middle ground which is bridging i.e. connecting two teams through a stable channel, there are joint ventures, there are full merges, there are also fully stale teams. Teams so defined can evolve depending on the evolutionary pressure they experience. To grow one needs structuring, normalization of processes and communication channels, specialization and differentiation of tasks, increasing complexity to increase resilience. To innovate, on the other hand, one needs fluidity, constructive chaos, to reverse the structuring process to return to the ideation, prototyping phase, to thus reduce the complexity of interactions to foster confrontation.
This abstract model obviously resembles the typical business organization but there is a huge difference: the rigidity, the non-fluidity, the inertia, the slowness in change. What I want to do in this paper is to hypothesize a (non)corporate structure that tries to approach the abstract model in terms of (non)overall resistance to evolutionary pressure. I reverse the implicit paradigm: I no longer need a given reason to change a structure over time, I need a real reason not to do so, not to explore. The exploration:exploitation ratio can be modulated, one can decide to hold firm on one business segment as within bounds, as effective enough, and at the same time explore the structure of another segment, to refluidify and experiment. That’s the whole point, to refluidify and experiment. I can experiment within constraints of functionality and (non)inertia. Then, for example, I can decree that the components of an entire department are all peers and, restarting from peer-to-peer interactions, observe how over time S self-organizes. Yes this creates a problem in itinere: there is a span of time in which I have no predictability about the level of operation of S. This makes me hypothesize not to do the tests on the production instance of S but to create a staging one with which to experiment.
Why reiterate this over time? Because in different situations, subjecting S to different evolutionary pressures and constraints results in different constructs. Because the appropriate teaming structure to cope with today’s situation is not the same one I would have chosen in the 1950s. There are concurrently trained competing neural networks (ref GANs): if I change the competitor with whom I interact I change the constraints and opportunities for growth and thus create different constructs and structures from time to time, structures that then go into the knowledgebase and are fully exaptable. Here I reintroduce the concept of exaptation, a term from evolutionary biology that stands for a biological trait that has evolved for a certain function (or without a specific function) and is subsequently co-opted for a different function. What I experienced E induced my structure but this is not equivalent to saying that E limits my potential.
The title of this article comes from purple team, a cybersecurity construct that involves, like the GANs I mentioned earlier, an attacking team and a defending team. An exploit that gets a reverse shell goes to induce an improvement in defenses, and a network probe that worked goes to improve a malware. Mutual elicitation of ideas, that is also what fictitious segmentation of a team is for, but that precisely involves experimenting, and having space and time to do so. In conclusion, reproducible techniques for systematically becoming competitive again are there, but as presumable they are complex and largely abstract and expensive. This one I mentioned is an example: one can modulate teaming structures as a form of elicitation.