In the project for a fully digital control system at an American nuclear power plant, manual steps were intentionally preserved within sequences that the automation could have performed independently. This is recounted by the head of the system's development, describing a decision made more than ten years ago.
According to his explanation, the goal was not the maximum reduction of manual work, but the preservation of operator skills. An operator who merely observes automation gradually ceases to act as an operator.
This example is significant for discussions on AI efficiency: automation can accelerate processes but simultaneously reduce the regular practice necessary for experts. The source does not report how widespread the problem is or which prevention methods have proven effective.
editorial commentary
Why it matters
Likely consequence: AI developers and those working on critical systems will increasingly evaluate not only the speed of automation but also the retention of regular practitioner experience. The next observable signal will be formal requirements for manual training, backup procedures, or operator skill verification. Significant uncertainty remains: the provided source describes a single project and does not show how applicable the finding is to other systems.