Definition
Plain language
Researchers known for showing that interrupting someone at the wrong moment is far more costly than the interruption itself.
As stated in the literature
HCI researchers whose 2004 work on interruption timing established that recovery costs from poorly-timed interruptions far exceed the interruption itself; cited as prior art for clarification-timing in long-horizon agents.
Why it matters: Their findings underpin modern thinking about when AI agents should ask clarifying questions — wait for a natural break, not the middle of a step.
For example, their work showed that interrupting someone mid-task costs far more time and accuracy than interrupting them between tasks.
Heard on the show
“Adamczyk and Bailey in 2004.”Episode 035 — Why Frontier Agents Ask for Clarification at Exactly the Wrong Moment