This aphorism by Nassim Taleb condenses everything one needs to know about the dynamics of inertia, complexity, and thinking in linear trajectories.
"People bred, selected, and compensated to find complicated solutions do not have an incentive to implement simplified ones." — @nntaleb
— Data Science Fact (@DataSciFact) March 5, 2017
Inertia, also known as path dependency, stands for a dynamic in which a network has so much invested in its perpetuation along a set of linear parameters that any deviation from those parameters [the path] will increase its operational costs to catastrophic proportions. The network – usually a vertically integrated complex such as a large bureaucratic organization – views path deviation as a direct challenge to its performativity, and aims to preempt deviations by continuously generating a wall of reinforcement noise RN [RN = organizational culture]. RN increases complexity along the entire structure of the network by adding an additional layer of agency to be performed along with the rest of its functions. Over time, as inertia is dynamically reinforced, such networks start resonating with RN to a threshold after which a process of dynamic self-selection starts eliminating network nodes not resonating at RN frequencies. From the perspective of RN [and senior management] the elimination of these nodes is intended to increase coherence and structural integrity. Paradoxically however, this process increases complexity even further, as it prunes any nodes capable of performing non-linearly and pulling the network away from inertia.
The RN driven self-selection process is an emergent quality of networks stuck in inertia. Crucially, while such networks generate RN as a defensive mechanism, it is also their biggest weakness, as all it takes to completely disrupt the network is to hack/jam/modify the RN, or for reality to intervene in its usual nonlinear way. The aphorism above captures this scenario.
Inertia rule of thumb: the moment an external observer can detect identical RN signals emanating unprompted from at least three structurally distinct nodes of a network [i.e. front-side, admin, senior management], one can deduce the network has flatlined into path dependency.
At that moment, when viewed from the inside, the network appears stable and full of robust momentum [‘we are all on the same page’]. However, RN driven inertia is a death sentence to any network as it leaves it completely exposed to, and at the mercy of, other networks capable of attenuating its RN and/or using the predictability of its path dependency to their advantage. Most importantly, even if no such outside networks exist [maybe they all are stuck in path dependencies], the costs of performativity of inertia rise exponentially with changes to the environment in which the network operates. While the network’s path is by necessity linear, the changes outside of it are always nonlinear. This divergence generates entropy and inevitably leads to collapse.