Late Lessons, Jensen Huang and AI
Markdown version

Review: public framing of 06 and 05#

Working file, 26 September 2026. Scope: whether 06-huang-and-late-lessons-through-maynard.md (890 lines) and 05-maynard-risk-and-ai-map.md (1,192 lines) read as standalone public documents for a web knowledge base. The review checks for project-internal commentary, direct address to Maynard, project mechanics, and whether the provenance notes are accurate and neutral. Method: I read 06 in full. For 05 I read §1–3, §7–9 and Appendices A–C in full and searched the rest by pattern (second person, first person outside quotations, “request”, “asked”, “batch”, “working/”, “INTERNAL”, “clarification”, the excluded book, Garbee). I checked the quotations of the series introduction against Maynard stuff/substack article 1 draft 3.md, and checked the supporting files that both appendices point readers to.

Overall. Neither document addresses Maynard directly. Neither contains questions for him or notes for the future essay, and there is no “Andrew asked” or “as requested” language in running text. AI and the Art of Being Human does not appear in either document. The Garbee ruling is reflected in both. The Anthropic disclosure in 06 §1.4 is good practice. The problems are at the edges: the appendices, the companion-file references, a pre-publication source, identifiers a public reader cannot resolve, and a handful of phrases that expose the process. Issues are ranked by the harm they would cause if the documents went live as they are.

Ranked issues#

1. The appendices send public readers to working files that are unpublished or not fit to publish (HIGH)#

If the files are not published, every path in the appendices is a dead reference. - Fix. In both documents, replace the path lists with a neutral note: “The working notes, dimension analyses and source copies behind this report are not published. Posts are cited from the public text mirror (https://text.futureofbeinghuman.com/substack/SLUG.html); other works are cited by the full references in 05, Appendix C.” If any of the working layer is to go online, publish a curated subset only after removing every INTERNAL section, the “ask Andrew” line, all references to the excluded book, and the Garbee co-written tag, and without any NYT or journal copies. In 06 §1.3, also remove “(Appendix B)” as the place to find the dimension analyses, or say they are not published.

2. A draft is quoted as if it were the published series introduction (HIGH)#

3. Companion document 04 is identified by a draft filename that still contains notes addressed to Maynard (HIGH for the set; MEDIUM for 06)#

4. The provenance note in 06 is neutral but incomplete, and it is not aligned with how the series describes the work (MEDIUM)#

5. 05’s provenance note omits the model, the vendor disclosure and a relevant co-authorship (MEDIUM)#

6. The “clarification” wording signals an off-page correction and is repeated mechanically (MEDIUM)#

7. Residual process language in 05 (MEDIUM)#

8. The “In brief” paragraph on the article is unlabelled and reads as an endorsement (MEDIUM)#

9. Quotations from the series introduction are attributed a little beyond what the text says (MEDIUM)#

10. Identifiers a standalone reader cannot resolve, or can misread (MEDIUM)#

Always prefix foreign codes with the document number (“01 T4”, “01 M1”, “02 T9”). Consider renaming the key “S3 2026” to “SCURVES 2026”. - Undefined terms. - “The Mirror” (line 849): add “01’s rule that every lens question is also asked of critics”. - “Maynard’s Garbee lesson” (562) and “His Garbee lesson” (694): use “the lesson of his 2019 work on responsible innovation in entrepreneurial cultures (2019-08-13)”; see also issue 12. - “the Raine case” (568): add a gloss. - “Centaur” (352): add “a model of human cognition” or similar. - “the lens” is used both for 01’s 72-entry diagnostic lens and for Maynard’s lens (for example line 728, “the lens’s application” against “his lens would raise…”). Call 01’s “the Late Lessons diagnostic lens” throughout.

11. 05’s authorship sentence could imply undisclosed ghost-authorship (LOW–MEDIUM; fairness to Maynard)#

12. The public wording of the Garbee ruling may read as a judgement on the co-author (LOW–MEDIUM)#

13. Residue of the excluded book in 05’s lens D3 (LOW–MEDIUM; ruling 2)#

14. Timestamp and source conventions in 06 §1.5 and §1.2 (LOW)#

15. “Post-recording” is dated from publication (LOW)#

16. A small inconsistency about whether Maynard has published a view of Huang (LOW)#

17. “A writing skill trained on Maynard’s prose” (LOW)#

18. Minor accuracy and tone points in 06 §1.1 (LOW)#

19. The map’s own voice slips into “we” and “our” (LOW)#

20. Loose wording in 06 §1.4 about mixed provenance (LOW)#

21. An unspecified conflict of interest in 06 §10.4 (LOW)#

INTERNAL (not for publication)#

Gates before 06 and 05 go live (overlapping with priority 1 in INTERNAL-notes-for-essay.md) 1. Decide whether any of the working/ layer will be published. If it will, it needs a cleaning pass: INTERNAL sections in M1–M9; INTERNAL-notes-for-essay.md; publications.md:233 (“ask Andrew for the file”); references to the excluded book in timeline.md, concept-index.md, the themes and supplement/S6–S7; the Garbee “co-written” tag in timeline.md:14; and no NYT or journal copies. 2. Re-check the series-introduction quotations against the published text on 27 September. 3. Confirm with Maynard whether he will edit 06 or only review it. Then set “reviewed by him” or “reviewed and edited by him” consistently in 05, 06 and the M-file headers. 4. Confirm which 04 text is final, and rename it.

Questions for Maynard - Why were the Modem Futura posts excluded (co-hosted? spoken? promotional?), so that 05 line 57 can give a neutral reason? - Is he content for the two September 2026 points to be cited as “personal communication”, or would he rather publish them (for example in the Tuesday piece) so that they can be cited as published statements? - The series introduction describes a “second article … through the lens of my own work — also authored by Claude” for Tuesday. Is that article 06, or a shorter piece derived from it? If it is derived, the introduction should point readers to 06 on the website, and 06 §1.2 might mention the article. - Is he comfortable with the published wording of the Garbee ruling (issue 12) and of the authorship sentence (issue 11)? - Does he want the “trained” wording (issue 17) kept to match his introduction, or made technically precise?