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.
The survey collected 936 examples from 319 knowledge workers and found that higher confidence in generative AI was associated with less reported critical-thinking effort, whereas higher self-confidence was associated with more critical thinking.
Scope: Knowledge workers' recalled use examples; association, not causal deskilling.
Respondents described work shifting toward goal-setting, response verification, and integrating outputs, indicating that AI use can relocate cognitive effort into oversight rather than simply remove it.
Scope: Reported practices, not measured organizational performance or causal skill change.
Boundaries
Limitations & independence
Recorded at coding time, carried with the work forever. A claim without its limits is not evidence.
Recorded limitations
- 319 self-selected knowledge workers and 936 recalled examples
- Cross-sectional self-report
- No direct cognitive-skill test or long-run deskilling measure
- Confidence and task selection may confound reported effort
Source independence
Microsoft Research-affiliated survey study of knowledge workers; product/vendor affiliation and self-report limit independence.