60% faster, 32% drop in cognitive load. Everyone’s quoting the top line and stopping there.
The interesting bit is buried in the methodology: the effect inverts almost completely for users who annotate AI output before accepting it. The act of marking up — challenging, correcting, questioning — reinstates the cognitive engagement that passive acceptance strips out.
There’s a “pilot vs. passenger” angle here that lines up exactly with the HBR workslop study. The tool isn’t the variable. The posture toward the tool is. Which is good news for leaders: posture is something you can train.