Feedback Loops: How Results Change What Happens Next

A feedback loop is not simply “receiving an evaluation”. It occurs when the result of an action re-enters a system and changes the next action. A thermostat adjusts heating in response to room temperature; a pianist changes movement after hearing unstable rhythm; a recommendation system adjusts what it offers after observing clicks. These cases differ, but each closes a loop between result, information and further action.

It is useful to distinguish reinforcing from balancing loops. A reinforcing loop amplifies an existing direction: anxiety disrupts concentration, and failed concentration creates more anxiety. A balancing loop responds to a discrepancy and pulls a state back towards a goal. Here, positive and negative describe directions of change, not moral value.

A feedback loop differs from an ordinary causal chain because the result returns as a new cause. It also differs from a one-off correction: a genuine loop can keep operating. This matters because the significance of an action lies not only in its immediate effect, but also in the subsequent cycle it establishes. Good learning is not merely getting something right once; it is arranging for error to return to action in a timely and accurate form.

https://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system/
https://ocw.mit.edu/courses/res-tll-004-stem-concept-videos-fall-2013/resources/information-flow/