Looking at any system, any process, or anything that can be described as a black box with inputs and outputs, I can’t help but notice a distinct, linear progression, directed with a profound and powerful underlying idea that systems tend to increase in complexity naturally.

When a system is born, it is small and vulnerable. It makes many mistakes and may fail easily, but if it doesn’t, it adapts and quickly becomes better and grows in its capabilities. It remains easy to grasp mostly because of its size.

Over time it increases the number of degrees of freedom it can handle.

At some point the mechanics of the system become more like an art — it has enough degrees of freedom, and is stable enough to experiment with its controls. In this highly creative stage, it truly defines itself.

Size is the primary enemy of art so as the system becomes more complex and bigger, its systemization begins. The experimentation gives way to proceduralization, and as some of its outputs are deemed more valuable than others, they are commoditized.

After systemization, these systems focus narrowly on maximizing efficiency of these designated outputs.

The final stage is a natural consequence of specialization and optimization — the system begins to rely on its optimizations. Small deviations in output become costly, as are small deviations in input. Since no system exists in a vacuum, eventually every system becomes irrelevant as the world around it changes. The system dies (a death of explosion if its construction or attrition).

You have seen this evolution everywhere — companies begin their life as small startups that can likely fail but are also agile and productive. As they gain confidence in their status and stability, they begin their creative phase — a killer feature, or a risky but profitable expansion into an unlikely market. Once they turn into corporations, ad hoc work becomes proceduralized, the company is too large to quickly adapt so it focuses on what it does best. Once that’s defined, it minimizes costs. But the industry changes and the corporation, too large to change its operating models, becomes irrelevant. Just think of what happened to Blockbuster’s.

We can expand this to TV shows. A new show much catch the eye. It’s simple and has a small base of supporters. It can change rapidly based on early feedback, but it also plays with its characters to gauge viewers’ reactions. As it gets big, it is doomed to repeat the same tricks, the same lines, the same plot twists, because that’s what the viewers are used to, and it’s difficult to change in a way that doesn’t turn a large portion of the audience off. It becomes formulaic — it has its distinctive style, and it’s no longer creative. As the viewers change their tastes (or as one generation is replaced by another), the show becomes irrelevant and eventually gets cancelled.

This is also true with people, though with some parallels (for example, death means irrelevance to society; systemization means having a daily routine, having a rigid set of preferences).

The best systems can resist this progression for a long time — by remaining agile, maintaining its growth through compartmentalization and appropriate scaling, and maintaining a careful equilibrium between death of attrition (irrelevance) and death of destruction.