Information Cost of Exit

The Information Cost of Exit is the systemic price paid by an institution when its members depart: not merely the loss of the members themselves but the loss of the specific diagnostic knowledge they carried, knowledge that would have enabled the system to correct what prompted their departure. Because exit is information-poor — it communicates dissatisfaction without communicating its content — the system learns that something is wrong but not what is wrong. The practitioners who possessed the specific diagnosis have removed themselves from the conversation, and with them has gone the knowledge that would have made correction possible. In the AI transition, this cost is unusually high because the departing senior practitioners possessed embodied knowledge that cannot be reconstructed from their output.