Every institution starts with a mission: educate children, heal the sick, represent citizens, advance knowledge. Over time, a silent inversion occurs. The institution begins optimizing for its own continuation rather than its original purpose.
Schools optimize for test scores instead of understanding. Hospitals optimize for throughput instead of health. Governments optimize for re-election instead of governance. Corporations optimize for quarterly returns instead of value creation.
The problem is structural. Institutions measure proxies: compliance metrics, satisfaction surveys, revenue growth, publication counts. These proxies are always gameable because they are socially constructed rather than physically grounded.
Once the proxy becomes the target, the institution drifts. Slowly at first, then completely. The people inside the institution are often aware of the drift but have no mechanism to reverse it. The incentive structure rewards drift.
The Extropy Engine proposes a different governance architecture: the Decentralized Feedback-Aware Organization (DFAO). Instead of static hierarchies with proxy metrics, DFAOs use:
1. Entropy auditing — continuous measurement of whether the organization is actually reducing entropy in its domain.
2. Adaptive feedback loops — governance rules that automatically adjust when entropy metrics diverge from mission targets.
3. Role fluidity — authority flows to whoever is demonstrably reducing entropy, not whoever holds a title.
4. Falsifiable mandates — every policy has explicit failure conditions. If the conditions are met, the policy auto-sunsets.
Entropy auditing at institutional scale is computationally expensive and politically contentious. Defining entropy reduction for a school versus a hospital versus a legislature requires domain-specific frameworks that don't yet exist at production quality. The governance layer also faces a bootstrap problem: who governs the governors before the system is self-sustaining? See open problems.
If you work inside an institution and feel the drift, this framework gives you language and architecture to name it and propose structural alternatives. You don't need to adopt the whole system. Start by asking: what would our organization look like if we measured actual entropy reduction instead of proxy metrics?