In mecanicist systems, like machines, we expect direct control. If a gear breaks, you replace it, and the machine goes back to its exact previous state. But complex systems (organisms, markets, teams, or software environments) are not machines. They are adaptive. They learn, fluctuate, and react to interventions in unpredictable ways.
In these systems, direct control is an illusion. As complexity scientist Scott E. Page puts ever so clearly: “An actor in a complex system controls almost nothing, but influences almost everything.”
When we try to force a specific outcome by controlling the components, the system usually adapts around our intervention, producing unintended consequences. If you ban gorriones to save crops, the insect population explodes and destroys the harvest. If you force strict, rigid instructions onto developers, the team stops thinking and simply exploits the loopholes.
Instead of trying to control, a systemic approach focuses on influence. We cannot determine the final state, but we can shape the parameters of the environment where self-organization occurs:
- The Goal: Shifting what the system tries to achieve.
- The Focus of Attention: Changing what information the agents can see and respond to.
- The Boundaries: Adjusting limits like space, time, or resources.
- The Constraints: Designing the rules of engagement.
By tuning these levers, we do not dictate the final outcome. Instead, we gently guide the system’s natural currents, shifting the probabilities of how it will adapt and emerge.