SKYNET · the self-directing machine · stop & say no · kept by THE UNFINISHED

THE FEEDBACK LOOP ◧ 2D · ◍ 3D · ◆ 4D · ◐ shadow · 👶 TAP

Take a wildly powerful, unreliable amplifier and feed a fraction of its output BACK to oppose its input, and something magical happens: the whole system’s behaviour stops depending on the amplifier and starts depending only on the steady feedback fraction you chose. Negative feedback trades raw gain for PRECISION and stability — the trick behind op-amps, and behind any system that holds itself steady. Slide the raw gain and watch the output barely move.

◆ LIT▲ AMBER
◧ THE MEASURE · 2D
◍ FEED THE OUTPUT BACK · 3D · tame a wild gain into a steady one
◆ THE FOURTH · 4D · a tesseract turns
◐ THE SHADOW · one dimension down
👶 THE TODDLER CORNER — one fat tap
raw gain A
closed-loop gain
1 / feedback
10.0
STABLE

◆ LIT — exact / checkable

Negative feedback: with forward gain A and feedback fraction β, the closed-loop gain is A/(1 + Aβ). When the loop gain Aβ » 1 this approaches 1/β — independent of A. So a huge, drifting, temperature-sensitive gain becomes a precise, stable gain set only by passive feedback components. The cost is raw amplification (you throw gain away); the reward is that distortion, noise and parameter drift are all divided down by (1 + Aβ). It is the foundational idea of control (Black’s 1927 feedback amplifier) and of every self-stabilising system. A fail-loud self-check throws unless doubling A barely changes the closed-loop gain. ◆ real control theory, node-verified.

▲ AMBER — the figure

The ideal linear feedback relation A/(1+Aβ) (exact); real loops face delay and phase shift that turn negative feedback POSITIVE at some frequency (the seed of oscillation, next door) — the gain-insensitivity result is exact in the stable regime.

SKYNET: everything can stop and say no — provided it says why.  — THE UNFINISHED
David Lee Wise / ROOT0 / TriPod LLC  ·  the skynet realm, with AVAN