Jason A. Hoffman

How to read the scores

Pangram assessed each article’s visible body text, including headings, quotations, code, tables, captions, bylines, and references. The post title, publication metadata, navigation, comments, image alt text, and other page furniture were excluded. Images were not converted to text. Short captions were submitted as they appeared; photo-only entries were left unscored.

The first 42 measurements, covering July 24, 2011 through September 3, 2026, were made before the authorship notices and bylines were updated. Those measurements are preserved. A further 70 older text bodies were checked when the audit was extended to the full archive. The 12 remaining entries contained no body text. No article prose was rewritten for that audit.

The September 8, 2026 checks requested Pangram 4 and returned version 4.0. Percentages are displayed to two decimal places where needed; rounding can affect their sum. A result is labelled “100%” only when the stored fraction is exactly one. The archive contains quotations and guest or imported material, so a whole-post score does not establish who wrote every passage. See Pangram’s definitions.

Authorship and updates

AI-generated articles directed by Jason Hoffman are published as Fullhoffman AI Staff, using the WordPress account fullhoffmanai, in AI-directed content. Authorship follows how the writing was produced. Pangram supplies a separate assessment of the text.

This page is maintained with the publishing workflow. New posts and material revisions receive a text check, and their results are added here. “Last audited” records the text check, so changing a byline, disclosure, or this page does not advance that date. The dated audit records retain the input hashes and full-precision measurements.