CONCEPT
Continuous Reinforcement
The schedule in which every response produces a reinforcing consequence — the parameter that produces AI engagement's rapid acquisition and compulsive maintenance, and the specific schedule type that differentiates AI from gambling's variable-ratio architecture.
Continuous reinforcement (CRF) is the reinforcement schedule in which every instance of the target response produces the reinforcing consequence. It is the schedule type that produces the fastest learning, the highest initial response rates, and — critically for the Skinner volume's analysis — the specific vulnerability to extinction that distinguishes it from variable-ratio schedules. Applied to AI-assisted work, the CRF designation is precise and diagnostic: every prompt produces a response, every response is useful, every interaction is reinforced. The behavioral consequences — rapid acquisition of prompting skills, high sustained engagement rates, difficulty of disengagement in the presence of ongoing reinforcement, rapid collapse of engagement when the system fails to respond — are the signature effects of CRF documented across a century of experimental work.
In The You On AI Field Guide
The behavioral properties of continuous reinforcement are among the first findings established in operant research and among the most consistently replicated. An organism placed on CRF acquires the response-consequence relationship