Meadows's late-career formulation for the practice of engaging with complex systems. Complex systems cannot be controlled — the attempt produces the opposite of what the controller intends, because the system generates responses the controller did not anticipate, across timescales the controller did not consider. What complex systems permit is influence, and influence requires a specific posture: expect to be surprised; protect information flows; use language with care; pay attention to what is important rather than only what is quantifiable; hold multiple perspectives simultaneously; stay humble. The posture converges on a practice rather than a position — continuous, adaptive, humble engagement with a system that always demands fresh response.
The principles apply to the AI transition with the specificity of a diagnosis. Expect to be surprised. The ecosystem will produce behaviors no participant predicted — the productive addiction was such a surprise; the dissolution of specialist boundaries was another; the speed of adoption yet another. The appropriate response is neither to pretend the surprise was predictable nor to treat it as evidence the system is beyond influence. It is to learn and adapt.
Protect information flows. A complex system can be influenced only by participants who can see its actual behavior. When dashboards show output but not the erosion of depth, when metrics show productivity but not depletion of cognitive reserves, participants cannot see clearly and cannot intervene effectively. Use language with care. In a complex system, language is not merely descriptive but constitutive; 'disruption' implies a temporary event, while 'transformation' implies an ongoing process requiring continuous adaptation.
The practice Meadows prescribed converges on what her final years embodied: she spent them on a Vermont farm, tending soil and animals, maintaining the small complex system that depended on her daily attention. The return from global modeling to local stewardship was not retreat but application — the skills that engage productively with any complex system are the same skills whether the system is a global economy or a patch of ground. The AI transition will be navigated at every scale, but the daily practice of engagement happens in specific organizations, classrooms, and families, in individual decisions made each morning about how to engage with tools that are always available.
Meadows articulated the principles most fully in her essay "Dancing with Systems" (1997) and in the closing chapters of the posthumously edited Thinking in Systems (2008). The phrase became her signature orientation — not optimism, not pessimism, but a disciplined humility that treats uncertainty as permanent rather than provisional.
Influence, not control. Complex systems do not yield to command; they respond to continuous, adaptive engagement.
Surprise is data. Unexpected behavior is information about the system, not evidence of failure.
Language is constitutive. Words shape what interventions are visible; word choice is itself an intervention.
Multiple perspectives simultaneously. No vantage point captures the whole; productive tension between perspectives generates interventions no single view would produce.
Humility as posture. The system is more complex than any model of it; humility is the only stance complexity does not eventually punish.