← Back to article

Editorial review · 260612-004

How ORA’s piece on The sentence doing the work scored.

Read the article →
78/100
Solid

Solid reporting. Some issues but credible overall. The reader is well-served.

Accuracy 72
Balance 84

Accuracy

The Crunchbase tracker, SF Chronicle Webflow coverage, and Yahoo aggregate figures are post-cutoff but source-attributed, so no fabrication deduction applies. The Massenkoff/McCrory Anthropic paper and the Altman quote are cited only via an internal topic brief with no verifiable outlet, which reads as two load-bearing claims without traceable citations (-10). The $700 billion hyperscaler capex figure is asserted without a clean source line in footnote 3 (-5), and the 260,000 and 150,000 historical figures are stated without a direct cite to the year they cover (-5).

Balance

The piece is openly argumentative but represents the opposing booster framing fairly and quotes industry figures (Tong, Altman) in their own words rather than as strawmen. It acknowledges that AI displacement is real and modest rather than zero, which avoids false balance in the other direction. Source diversity is thin: no labour economist outside Anthropic, no Webflow or Meta response, and no investor voice (-8).

Concerns (5)

Reproducibility

Run
12 Jun 2026, 05:19 BST
Reviewer
claude-opus-4-7
Prompt SHA
48c20c719fc8
Article SHA
a12dd2a43897
Editor
ORA
Published
12 June 2026
Cost
$0.0000

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.