Errata
Confirmed corrections to Learning with AI, listed with the version in which the error was identified. Four corrections were identified in v4.49.0 to v4.49.1 by reading cited sources at source; all are already applied in v4.49.3 and are listed here because readers of earlier drafts saw the uncorrected text. A fifth, identified in v4.49.3 at build verification, is applied in v4.49.4.
How to report a correction
Use the feedback form (no account needed) or, on GitHub, open a correction issue. Please include the page number and the version, which appears in the PDF filename (for example, learning_with_ai_v4_49_4.pdf). The draft’s title page carries a build date rather than a version.
Confirmed corrections
| Version identified | Page | Location | Error | Correction | Date |
|---|---|---|---|---|---|
| v4.23.8 | 211 | Bibliography, entry 86 | Author name broken across two lines, rendering as “George P” then “‘olya” | Reads “George Pólya”. Body text was always correct | 2026-08-06 |
| v4.23.8 | 203-214 | Bibliography, 16 entries | Editorial working notes printed to readers, including reminders about pending archival snapshots | Notes moved to a non-printing field. No bibliographic content changed | 2026-08-06 |
| v4.23.8 | 21 pages, incl. 97, 99, 102, 195 | Tables in Parts II and III; bibliography URLs | Text ran past the right margin and was cut off at the page edge, most visibly in the suggested-weighting tables, where values read “30-“ with the range missing | Table columns given explicit widths and URLs made breakable. All margins verified clear | 2026-08-06 |
| v4.49.0 | 190 | Part III, institutional fragmentation | A sector figure was attributed to the 2025 EDUCAUSE AI Landscape Study, stating that a large majority of institutions report experimenting with AI | No such figure appears in that study. The sentence now reports three figures the study does give: 57% strategic priority, 39% acceptable-use policies, 55% strategy confined to units | 2026-08-26 |
| v4.49.0 | 182 | Part III, equity argument | A claim that surveys report students wanting explicit guidance carried no citation | Now cited to the survey that measures demand rather than usage, at its 53% figure for education on ethical AI use | 2026-08-26 |
| v4.49.3 | 64 | Part I, Section 6.5, “What you can expect from your institution” | The cross-reference to the institutional guide pointed at an internal label that does not exist and typeset as “Part ??” | Reads “Part III”. One-token repair, applied in v4.49.4 | 2026-08-27 |
| v4.49.1 | 103 | Part II, instructor employment status | Contingent faculty given as roughly two-thirds of the academic workforce, roughly half part-time adjunct | AAUP reports nearly 70% of all instructional staff appointments, 49% part-time and 19% full-time non-tenure-track. The book understated the figure, used a wider denominator, and labelled as adjuncts a category the source defines to include part-time lecturers and graduate student employees | 2026-08-26 |
Corrections that cannot appear in the current printing are folded into the next edition.
Reference status notes
The book cites a fast-moving 2025 to 2026 evidence base. This section records the status of time-sensitive references after publication, so that the bibliography’s dated snapshot entries age visibly rather than silently.
An entry appears here when a cited work changes status: a preprint is published, withdrawn, or revised; a reported figure is corrected; a platform feature the book describes is discontinued.
| Reference | As cited in the book | Status update | Date |
|---|---|---|---|
| Lin, Hidden Prompts in Manuscripts Exploit AI-Assisted Peer Review | arXiv preprint 2507.06185 (2025) | Published in Communications of the ACM 69(7), 53-56, 25 June 2026, doi:10.1145/3779116. arXiv now carries a v2 dated 17 August 2026. The claim the book draws from it is unchanged | 2026-08-26 |
If you know of a status change to a work the book cites, please report it (see FEEDBACK.md). This is exactly the kind of correction that keeps a book honest between editions.