Decontrol vs Control Problems
Discussion
David Ronfeldt:
"Today’s world abounds with control problems: e.g., autocrats intent on imposing control, lest they lose control; wars in Ukraine, Gaza, and elsewhere for geopolitical and noöpolitical control; AIs being designed to exercise control over one process or another — with people everywhere feeling one matter or another is out of control. All as though there is a spectrum with control at one end, loss of control at the other, and only control is worth valuing.
Yet many people know better. They often want to benefit from decontrol. But since it’s not an accepted word, they use concepts from spectrums that are off to the side of the control spectrum — freedom, liberty, democracy, liberalization, decentralization, deregulation, delegation, etc. I have long argued (Ronfeldt, 2009, 2023e, 2024, 2025) that decontrol processes are as important as control processes — for example, when parents let children go their own way, or when officious aristocratic clans accept that government bureaucracies should move from patrimonial to professional staffing criteria, and when governments learn to cease dominating their economies so market systems can develop. Knowing how, when, where, and why to decontrol an entity, system, or sub-system is periodically more important than trying to exert control over it. Without decontrol, there can be no freedom, no markets, no democracy, no commons.
Unfortunately, cyberneticists, complexity theorists, and other evolutionary theorists have not developed a science around decontrol the way they have around control. Cybernetics —a founding sciences behind AI — appears to be on reason why AI is so focused on maximizing control. For the field defined as a science of “communication and control” that focuses on “control hierarchies” and “higher levels of complexity and control” (Umpleby, 2000). Cyberneticists have yet to regard decontrol as a concern for cybernetic theory (unless there’s a literature I’ve missed). American cyberneticists have long favored freedom and democracy, but have viewed them too as control problems. Some proponents of AI may want to decontrol the environment in which AIs are being developed (e.g., Gourley, 2023). But designing AIs themselves for decontrol purposes is not on any policy or technology agenda, though adjunct issues slip into some debates — e.g., how AI could help assure democratic voting or decentralized decision-making.
Yet some kind of decontrol challenge figures in every “major evolutionary transition” (MET) where societies advance phase-by-phase from simpler to more complex designs — as in progressions from early tribal societies to modern nation-states. The key step that defines an MET is the differentiation and addition of a new system (sub-system) — e.g., a market system — that transforms a society’s structural-functional design, improving its performance and fitness. This process must accommodate to decontrol at some point, as people struggle over how a newly-added sub-system works with a society’s existing systems and sub-systems.
The growth of the noosphere is bound to generate decontrol as well as control problems. It happens all the time, across all ages, as different ideas arise and come into contact. Unfortunately, past theorists have not analyzed geological, biological, and/or societal evolution in control-decontrol terms that identify precedents from which to extrapolate regarding the noosphere’s evolution.
Theorists often view evolution toward higher levels of complexity as a process that requires constantly coming up with new forms and levels of control, say to be rid of free riders and “bad actors,” or to assure that sub-systems coordinate better. Theorists who work on METs of all kinds seem especially preoccupied with control problems."
(https://davidronfeldt.substack.com/p/our-worlds-ballooning-noosphere-3)