Co-regulation
In the learning sciences we distinguish self-regulation from co-regulation — the regulation of a task shared between a learner and a more- or less-capable partner who lends structure until the learner can hold it alone. When that partner is a person, the arrangement is stable; people change slowly, and they tell you. When the partner is a machine, the arrangement is only as durable as a version number.
This matters for distributed learning: as more learning happens alongside AI, the relationships that carry it become co-regulated, and therefore fragile — contingent on a partner that can be replaced overnight. The post “Taste Collapse” works one corner of this: what happens to a writer’s voice when the machine does too much of the regulating.