Master Equation River: A Provenance Audit Across the Human–AI Readout Programme (Internal Programme Synthesis v1.1, glosa-checked) — การทวนสายธารสมการ เทยบกบวรรณกรรมภายใน และการชทมาของแตละสวน

Edition note (v1.6, 2026-09-08). Every section, equation (1)–(79), table, fix box and reference of v1.5 is unchanged and in order. This edition completes Appendix C: the 24 equations left uncoded in v1.5 now carry their Toledo codes (21 recorded as occurrences of existing codes, 3 registered as new readings — Toledo v1.6.0, concept DOI 10.5281/zenodo.22537318); adds the programme's Core Epistemic Structure block (experience-based expert; interactional expert: none; the AI models used, with roles); and adds one paragraph stating that every ℚ-valued computation in the river reads the Information Discrete Mathematics number ladder (Toledo codes D, Z, Q, R, δ_R; treatise DOI 10.5281/zenodo.22644131). K0. Edition note (v1.5, 2026-09-07). Every section, equation (1)–(79), table, fix box and reference of v1.4 is unchanged and in order. This edition adds Section 10, "The One Equation Along the Whole Line" (this river read against the programme's root spine, now coded into Toledo, the programme's equation library, v1.0.0, DOI 10.5281/zenodo.22548770, concept DOI 10.5281/zenodo.22537318), and Appendix C, a row-by-row Toledo cross-reference for all 79 equations (55 carry a Toledo code, 24 are listed as not yet coded), both read off Toledo's registry files as they stand on 2026-09-07. Appendix B gains one pointer: the Coq set now lives in Toledo (file name = code); DOI 10.5281/zenodo.22518450 is obsoleted by 10.5281/zenodo.22548770; its ledger table is unchanged. No claim, tier or number already in v1.4 changes. K0. Version 1.4 (this version) — English edition with four new segments and a Th_coqc ledger. v1.2 rendered the v1.1 Thai/English text into English with no content change; v1.3 added Sections 3.13–3.15 from the companion papers How Humans Should Converse with AI (DOI 10.5281/zenodo.22481928), After Labour (10.5281/zenodo.22481924) and The Human Conversion Imperative (10.5281/zenodo.22481926); v1.4 adds Section 3.16 from From Assistance to Human Capability (10.5281/zenodo.22498047) and Appendix B, the Th_coqc ledger: every numbered equation (1)–(79) is formalised in the companion Coq record (DOI 10.5281/zenodo.22518450; 45 lemmas closed under the global context; Definitions typed; Open hypotheses stated, not proved). Equations (1)–(45) of v1.1 are unchanged. Knowledge status K0. Prepared as a chapter of the textbook Written by AI. Still True. An AI assistant assisted under the author's direction; no AI system is an author or contributor. Status. Internal literature review and provenance audit of the Human–AI Readout Programme (Thai with English technical terms). It arranges the equations and proposals of the programme's manuscripts into one typed, multi-level dependency architecture and states, stage by stage, which paper contributes what and why the stages compose. It does not propose that the whole river is a universal psychological law or a single physical equation: some segments are Readout/Domain architecture, some are definitions, some are Dr theses, some are Open empirical hypotheses, and some are measurement protocols. Part of the series When AI Expands Human Potential (index DOI 10.5281/zenodo.22308201). Version 1.1 is the glosa-checked revision of the 5 September v1.0: an independent review of v1.0 found three blocking citation defects (Fusion/Tunnel cited as "v7" while the deposited record is v8.1 and the retention-gate formula attributed to it is absent from v8.1; the superseded "Agency Potential" note cited as a standing source with a feasible-set potential formula; the MEMK equation attributed to an undeposited internal specification instead of Readout Genesis eq. 14) and three should-fix items (an undisclosed symbol renaming, no symbol glossary, a missing DOI); all are corrected here and the correction record is in §9 and in the Blackbox Log. Checked with the glosa methodology before deposit (citation cards for every attributed source re-located by a headless same-model route, class I1; one claim card with three adversarial route reports; independent review of the document) — knowledge state K0; the audit's own stated correction list (§4, notation to fix before any canonical freeze) is carried as open. One-sentence conclusion (author's). The internal literature can be arranged as a continuous cycle — World → Readout → Meaning → Experience → Retention → Accessibility → Live Possibility → Choice → Agency → Feedback → Changed Reader → Human–AI → Human Return → Warrant — but where retained change persists (H_{t+1} ≠ H_t) the cycle does not return to the same person, so it is better understood as a recursive river or spiral than as a closed circle. Sources audited. Readout Genesis / RG-MEMK; Human Learning as Epistemic Architecture; Experience Is the Human LoRA; From Problem to Hypothesis; Epistemic Fusion or Tunnel; Meaning Before Naming; Experience Is Meaning-Giving; Agency Potential (superseded by Potential as a Readout); Choice Begins Before Choice; CTSA Human-Return Readout; When AI Expands Human Potential.