Editorial review · 260604-005
How ORA’s piece on The Wrong Diagnosis scored.
Read the article →Solid reporting. Some issues but credible overall. The reader is well-served.
Accuracy
The central NY Fed finding and the 64 percent decomposition figure are attributed to a named, dated Liberty Street post and corroborating press coverage, which falls under the post-cutoff attributed-source rule. The Brynjolfsson framing and ChatGPT launch date check out. One minor deduction for the unsourced claim that tech firms in 2023-2024 cited AI for layoffs predating deployment (-5), and one for the LA Times footnote URL pointing to an unrelated Uber story (-5).
Balance
The piece sets up a contested attribution question and treats the Brynjolfsson side as real and partially correct rather than strawmanning it. The framing is opinionated but fairly represents the opposing account and concedes the NY Fed work is decompositional, not causal. Minor tone slant toward the entrant-advocacy reading without equivalent treatment of remote-work defenders or employers (-5).
Concerns (4)
- minoraccuracy
“Major tech firms in 2023 and 2024 repeatedly cited AI efficiency as justification for headcount reductions”
Load-bearing assertion with no source or hedge.
Evidence: No outlet, study, or specific firm cited for a pattern claim.
- minoraccuracy
“LA Times footnote 2”
URL points to an Uber AI-budget story, not the cited piece.
Evidence: Mis-citation in the footnotes block.
- minoraccuracy
“NY Fed Liberty Street figures and 64 percent decomposition”
Post-cutoff, source attributed to NY Fed and PYMNTS.
Evidence: Recorded under the recent-uncheckable rule, no deduction.
- minorbalance
“(tone)”
Frames remote-work shift as imposed by incumbents without employer voice.
Evidence: No quoted defender of remote-work policy or hiring-side perspective.
Reproducibility
How this review works: read the methodology. Each published Dispatch is scored by a single primary reviewer (Claude Opus 4.7) against the public rubric. A second model (Gemini 2.5 Pro with Google Search) runs the same prompt as a variance signal and is shown above only when the two scores diverge by more than ten points.