During development, the complex neuronal circuitry of the brain emerges from limited information contained in the genome. After the genetic code instructs the birth of neurons, the emergence of brain regions, and the formation of axon tracts, it is believed that neural activity plays a critical role in shaping neural circuits for behavior. Current AI techniques are modeled after the same principle: connections in an initial weight matrix are pruned and strengthened by activity-dependent signals until the network can sufficiently generalize a set of inputs into outputs. Here, we challenge these learning-dominated assumptions by quantifying neuronal activity contribution to the emergence of the OptoMotor Response (OMR), a complex, visually guided swimming behavior in larval zebrafish that involves the whole nervous system, from sensory input to motor output. Dark-rearing zebrafish revealed that visual experience plays no effect on the emergence of the OMR behavior. We then raised zebrafish under continuous anesthesia in complete darkness, such that from fertilization onward, they were deprived entirely of both sensory experience and neuronal activity. Strikingly, these animals could immediately perform swim bouts and respond to visual stimuli with 75% accuracy in the OMR paradigm. Shorter periods of anesthesia and dark-rearing during development did not reduce OMR performance below 90% accuracy, calling into question classical critical periods for visual development. Thus, contrary to what you learn on your mother’s knee, the OMR in larval zebrafish is wired up by largely activity-independent developmental mechanisms, whereas neuronal activity plays only a minor role in refining the circuit.