Becoming the loop
When a human in an AI workflow is so consumed by supplying context, checking and repairing output that they effectively perform the work through the machine, rather than exercising judgment at key moments.
There is a difference between a human staying in the loop and a human becoming the loop. The first applies judgment at defined moments: approving a consequential action, resolving an ambiguous case, setting a policy. The second is present at every step, feeding context, verifying each output and fixing mistakes, so the AI is accelerating a task that a person is still effectively carrying.
Both are described as "human in the loop", which is why the label hides so much. The difference shows up in the numbers: a person who has become the loop has a high supervision load per dependable outcome, even if the automation rate looks high.
How to avoid it
Separate mandatory judgment from preventable repair, fix recurring errors at their source and measure whether human attention grows with volume. Use abstention so people see only the cases that need them. Aim for staying in the loop, not becoming it.
Read more in The hidden metric in AI automation is supervision.
Related terms
Human in the loop
A design where a person reviews, approves or corrects an AI system's work at defined points. In practice it can mean a person with real judgment, or a person who has become the loop.
Supervision load
The human attention an AI system still consumes — context supply, review, approval, correction and exception handling — measured per dependable, completed unit of work.
Automation rate
The share of work an AI system completes without a person doing it. A useful headline number that says nothing on its own about the human attention left behind.