Reference
The receipts establish the examples and boundaries. They do not establish the essay’s forecast.
The audit carries 29 source receipts across courts, open-source security, hiring, publishing, email, language burden, and detector bias. The strongest evidence establishes the cost and direction of specific queue changes. The weakest part remains the cross-domain forecast that prior reputation will become the cheapest general substitute for human reading.
The correction at the top of the audit matters because the source chain worked in reverse: the author’s later direct read found that the completed audit had understated the review cost and carried a stronger percentage wording than the source actually opened.
Canonicality: The essay Signal & Noise stands behind is Issue 22, The Price of Being Read. This is part of its published audit, not a second edition of the essay.
Source families
Open-source security
- curl’s intake pause, security-team size, per-reviewer time, bounty policy, patch checks, and AI-assisted fixes.
- Linux kernel maintainer remarks on the change from low-quality to useful AI reports.
Open queues
- Court citation failures and filing access.
- LinkedIn, Workday, and Springer Nature submission volumes.
- Clarkesworld, Amazon KDP, email authentication, and publisher screening.
Outsider effects
- A field experiment on writing help, hiring, and wages.
- Language-related writing burden and rejection.
- Detector false positives on non-native English writing.
The inspectable record
The full audited version preserves every receipt with its stated scope and limit. Go directly to the 29-receipt Reference section →
The receipts are the audit’s outside constraints. They are not a vote count for the thesis. Several are self-reports, working papers, dated platform figures, or mirrors used when a primary host could not be reached; those limits remain attached to the claims they bound.