Commune

Feedback Loops

6 min read

Feedback loops are the fundamental mechanism by which systems regulate themselves, adapt to change, and maintain (or lose) stability. In cybernetic theory, a feedback loop occurs when a system’s output is fed back as input, creating a circular causal relationship that influences future behavior. Understanding feedback loops is essential for analyzing how recuperation neutralizes resistance, how the-spectacle maintains control, how viable systems self-regulate, and how self-describing-systems bootstrap themselves.

Feedback loops are the mechanism behind many concepts collected here — the how that connects abstract theory to observable system behavior. Whether you’re designing infrastructure, analyzing power structures, or building governance systems, you’re working with feedback loops, either intentionally shaping them or being shaped by them without realizing it.

This article is a stub. Contributions welcome.

Two Types of Feedback

Negative feedback (stabilizing)

Negative feedback reduces deviation from a target state. When the system moves away from equilibrium, negative feedback pulls it back.

Mechanism: Output → comparison to goal → corrective action → reduced deviation.

Examples:

  • Recuperation — capitalism detects opposition → absorbs and commodifies it → restores system stability.
  • Thermostat — room gets too cold → heater turns on → temperature rises → heater turns off.
  • VSM System 3 — performance deviation detected → regulatory intervention → performance corrected.

Key insight. Negative feedback maintains the existing order. It’s conservative by design — any system using primarily negative feedback resists change and preserves its current structure. This is why recuperation is so effective: it’s capitalism’s negative feedback mechanism for neutralizing threats.

Positive feedback (amplifying)

Positive feedback amplifies deviation from the current state. When the system moves in a direction, positive feedback accelerates the movement.

Mechanism: Output → reinforcement → amplified output → further reinforcement.

Examples:

  • Network effects — more users → more value → more users → more value (how platform-capitalism grows).
  • Microphone feedback — sound → amplification → louder sound → more amplification (until equilibrium breaks).
  • Social movements — action inspires more action → momentum builds → rapid system change.
  • Détournement cascades — one subversion inspires others → cultural shift accelerates.

Key insight. Positive feedback creates instability — either destructive (system collapse) or creative (system transformation). Revolutionary change requires positive feedback to overcome the negative feedback mechanisms that preserve existing power structures.

Feedback Loops as Analytical Lens

Power structures

The spectacle operates through feedback loops:

  • Surveillance extracts information (sensors).
  • Media shapes behavior (actuators).
  • Consumer response is measured (feedback).
  • Marketing adapts to maintain control (regulation).

The loop closes: the spectacle learns from resistance and adjusts to neutralize it. Understanding this feedback structure reveals how the spectacle maintains itself, not just that it exists.

Resilient infrastructure

Beer’s Viable System Model is fundamentally about feedback-loop design:

  • Systems 1–5 communicate through feedback channels.
  • Requisite variety requires feedback to detect and respond to environmental complexity.
  • Autonomy requires local feedback loops that close without central intervention.

When designing distributed infrastructure, you’re designing feedback loops: who observes what, who responds to which signals, how fast information flows, where regulation happens.

Breaking out of recuperation

If recuperation is capitalism’s negative feedback mechanism, then effective resistance requires:

  1. Understanding the feedback structure. What signals trigger recuperation? Where are the sensors?
  2. Designing for positive feedback. How can autonomous zones amplify each other instead of being isolated?
  3. Creating different feedback goals. What if success isn’t measured in commodifiable metrics?

Détournement works by hijacking existing feedback loops — turning the spectacle’s own mechanisms against it. Without understanding feedback structure, you can’t effectively hijack it.

Self-describing systems

Self-describing systems face a circular challenge: the system must describe how to describe itself. This is a feedback loop — the description improves the system, the improved system generates better descriptions. Without that loop closing, the system can’t bootstrap.

What a Complete Article Should Cover

Theoretical foundations

  • Norbert Wiener’s Cybernetics (1948): communication and control through feedback.
  • W. Ross Ashby’s Design for a Brain (1952): homeostasis as feedback regulation.
  • Stafford Beer’s Brain of the Firm (1972): management cybernetics and feedback channels.
  • Gregory Bateson’s work on feedback in biological and social systems.
  • Ludwig von Bertalanffy’s General Systems Theory: feedback in open systems.

Formal definitions and mathematics

  • Control-theory notation and diagrams.
  • Transfer functions and stability analysis.
  • Time delays in feedback loops and their effects.
  • Oscillation, overshoot, and damping.
  • Bifurcation points where feedback type changes.

Feedback in social systems

Feedback in biological and cognitive systems

  • Homeostasis (temperature regulation, blood sugar, etc.).
  • Predator–prey population dynamics.
  • Neural networks and learning as feedback.
  • Ecological feedback loops (climate change, tipping points).
  • Evolution as feedback between organism and environment.

Feedback in technical systems

  • Control-systems engineering (PID controllers, autopilots).
  • Webhook architectures and event-driven feedback.
  • Idempotent automation — designing safe feedback loops.
  • Software deployment pipelines: CI/CD as feedback loop.

Pathological feedback patterns

  • Runaway positive feedback — system collapse (economic bubbles, panic, ecological collapse).
  • Dampened negative feedback — system that can’t correct itself fast enough.
  • Feedback delays — time lag causing oscillation or overshoot.
  • Feedback decoupling — when loops break and the system loses regulation.
  • Adversarial feedback — Goodhart’s Law: when a measure becomes a target, it ceases to be a good measure.

Designing with feedback

  • When to use negative vs. positive feedback.
  • Balancing stability (negative) with adaptability (positive).
  • Closing feedback loops vs. leaving them open.
  • Feedback granularity — how often to sample and respond.
  • Multi-loop systems — coordinating multiple feedback processes.
  • Requisite variety in feedback-channel capacity.

Historical examples

  • Project Cybersyn — real-time economic feedback in socialist Chile (1971–73).
  • Thermostat — first practical negative-feedback device.
  • Watt’s centrifugal governor — steam-engine speed regulation (1788).
  • OODA loop (Observe–Orient–Decide–Act) — military feedback cycle.
  • Agile / Scrum — software development as feedback-driven process.

Philosophical implications

  • Teleology and purpose — do feedback systems have goals?
  • Causality — circular causation vs. linear cause-and-effect.
  • Autonomy — can systems with feedback be truly autonomous, or only reactive?
  • Determinism vs. agency — feedback loops and free will.
  • Ethics of control — who designs the feedback? Whose goals are encoded?

See Also

Graph View

Backlinks