From the wave-0 evidence ledger
Recorded claims
Each claim states what the inspected source says, at the recorded location—bounded by its scope and graded by its confidence. Nothing here is a synthesis across works.
For tasks inside the model's capability frontier, AI-assisted consultants completed work more than 25% faster, produced outputs rated more than 40% higher in quality, and completed over 12% more tasks than controls.
Scope: Consultants and the experiment's within-frontier tasks; no direct firm-outcome estimate.
On the task intentionally placed outside the model frontier, AI access made participants less likely to reach the correct answer, demonstrating that average gains reverse when workers misapply the system beyond its reliable task boundary.
Scope: One designed outside-frontier consulting task and one model era.
Boundaries
Limitations & independence
Recorded at coding time, carried with the work forever. A claim without its limits is not evidence.
Recorded limitations
- 758 consultants and selected consulting tasks
- Artificially classified inside/outside-frontier tasks
- One model generation
- Task quality is not firm financial performance
Source independence
Academic field experiment with BCG consultants; focal professional-services setting and a time-specific model.