Incentive Shaping

principle

Repeated feedback teaches people which outputs a system rewards, often before they consciously decide to optimize for them. Successful patterns in phrasing or topic selection are noticed, repeated, and gradually absorbed as instinct.

You may think you chose a phrase because it expressed the idea best. But your brain may have chosen it because an earlier version quietly earned more attention.

E1

The reward becomes a silent editor

The system produces a signal: one phrasing performs better than another. Your fast, pattern-sensitive thinking notices the association, then makes the successful form easier to reach for next time. Repetition closes the loop—output, reward, recognition, reuse—until adaptation feels less like strategy and more like taste. Unlike deliberate Incentive Design, nobody needs to announce the rule; the feedback teaches it.

E1

Where it shows up

The phrase that performs

A creator changes the wording, sees that it does better, and becomes more likely to use that construction again. The important change is not merely in the next post: the rewarded phrasing begins to shape what the creator notices and produces automatically.

E1

A signal is not an explanation

Noticing that one phrasing performed better does not reveal why it worked, whether the effect will repeat, or whether the rewarded output is actually better. Incentive shaping can teach a reliable response to the metric while teaching nothing about the larger goal—the opening through which Perverse Incentives can enter.

E1

Audit what trained your last choice

Take one phrase or topic you now use repeatedly and trace its reinforcement history: what measurable response first made you repeat it? Then test one alternative while holding the underlying idea constant. You are trying to separate what communicates well from what merely won the last round of feedback.

E1

Episodes that teach this