The DAX Foundation: Writing https://daxfoundation.org/writing/ Pieces written in the Foundation’s voice, around its definitions: the reasoning, the measures, the consequences. Versioned, in public. A record, not a wall https://daxfoundation.org/writing/a-record-not-a-wall/ https://daxfoundation.org/writing/a-record-not-a-wall/ Sun, 18 Oct 2026 00:00:00 +0000 How aviation and nuclear power built their safety practice around kept records, and where AI's record is still missing. Fourteen small experiments, and why small first https://daxfoundation.org/writing/fourteen-small-experiments/ https://daxfoundation.org/writing/fourteen-small-experiments/ Sun, 18 Oct 2026 00:00:00 +0000 What E01 to E14 each test, which have run, and the bottlenecks found before anyone was hurt by them. How a showcase is made https://daxfoundation.org/writing/how-a-showcase-is-made/ https://daxfoundation.org/writing/how-a-showcase-is-made/ Sun, 18 Oct 2026 00:00:00 +0000 Brief, package, scenes, reader: the method behind the curry, the Stade and Words into Algebra, and what the honesty costs. How knowledge compounds under a constraint https://daxfoundation.org/writing/how-knowledge-compounds-under-a-constraint/ https://daxfoundation.org/writing/how-knowledge-compounds-under-a-constraint/ Sun, 18 Oct 2026 00:00:00 +0000 Technique reuse, a federated registry, and fetching the spot instead of the course: what compounds is the record of what was tried. How this is paid for, and what we will measure https://daxfoundation.org/writing/how-this-is-paid-for-and-what-we-will-measure/ https://daxfoundation.org/writing/how-this-is-paid-for-and-what-we-will-measure/ Sun, 18 Oct 2026 00:00:00 +0000 Compute, coffee and sponsors; no Foundation money; and the measurement that has not happened yet. If you are the tutor in the group https://daxfoundation.org/writing/if-you-are-the-tutor-in-the-group/ https://daxfoundation.org/writing/if-you-are-the-tutor-in-the-group/ Sun, 18 Oct 2026 00:00:00 +0000 A kit for your hardest topic, an Edition for each learner, and the report a parent can read. Read your page before your child reads theirs https://daxfoundation.org/writing/read-your-page-first/ https://daxfoundation.org/writing/read-your-page-first/ Sun, 18 Oct 2026 00:00:00 +0000 What the guide's page holds, section by section, and why the one line of "why" under each step matters. Refusing a bypass, and keeping the receipt https://daxfoundation.org/writing/refusing-a-bypass-and-keeping-the-receipt/ https://daxfoundation.org/writing/refusing-a-bypass-and-keeping-the-receipt/ Sun, 18 Oct 2026 00:00:00 +0000 What the one run of the Constraint Protocol recorded on 4 May 2026, as far as the public record goes. Run it with us https://daxfoundation.org/writing/run-it-with-us/ https://daxfoundation.org/writing/run-it-with-us/ Sun, 18 Oct 2026 00:00:00 +0000 What the first run did not test, what a second run needs, and how to break the protocol on purpose. Session two https://daxfoundation.org/writing/session-two/ https://daxfoundation.org/writing/session-two/ Sun, 18 Oct 2026 00:00:00 +0000 The next-time note, the learning record your family keeps, and why the second session is where it starts to compound. The first run: a first page in eight minutes, and what it got wrong https://daxfoundation.org/writing/the-first-run/ https://daxfoundation.org/writing/the-first-run/ Sun, 18 Oct 2026 00:00:00 +0000 What the recipe built on 27 September, who played the learner, and the eight things we found. The game has no AI in it https://daxfoundation.org/writing/the-game-has-no-ai-in-it/ https://daxfoundation.org/writing/the-game-has-no-ai-in-it/ Sun, 18 Oct 2026 00:00:00 +0000 What the learner's side is, what it never does, and why the answers wait until the child has tried. The slower model was not the safer model https://daxfoundation.org/writing/the-slower-model-was-not-the-safer-model/ https://daxfoundation.org/writing/the-slower-model-was-not-the-safer-model/ Sun, 18 Oct 2026 00:00:00 +0000 Three Meta DAX experiments from 10 October 2026: fetching one piece instead of a whole course held quality, a small model matched a large one, and on the one acute safety case the large model was the one that got it wrong. The stranger test https://daxfoundation.org/writing/the-stranger-test/ https://daxfoundation.org/writing/the-stranger-test/ Sun, 18 Oct 2026 00:00:00 +0000 Before anyone else tried the public page, we had it followed by stand-ins with no context at all. Here is what broke, what was fixed by when, and what has still not been tested by a person. Twelve scenes, 1: The nurse's walkthrough https://daxfoundation.org/writing/twelve-scenes-01-the-nurses-walkthrough/ https://daxfoundation.org/writing/twelve-scenes-01-the-nurses-walkthrough/ Sun, 18 Oct 2026 00:00:00 +0000 A charge nurse in Enugu has forty minutes, and the first thing the walkthrough does is ask what she brings. What a declared constraint is https://daxfoundation.org/writing/what-a-declared-constraint-is/ https://daxfoundation.org/writing/what-a-declared-constraint-is/ Sun, 18 Oct 2026 00:00:00 +0000 Declaring in advance, keeping author, witness and validator apart, and where this sits among the records that already work. Where your child is stuck is the whole brief https://daxfoundation.org/writing/where-your-child-is-stuck-is-the-whole-brief/ https://daxfoundation.org/writing/where-your-child-is-stuck-is-the-whole-brief/ Sun, 18 Oct 2026 00:00:00 +0000 Five lines, no names, and the recipe has everything it needs. Human delta value: the measure https://daxfoundation.org/writing/human-delta-value/ https://daxfoundation.org/writing/human-delta-value/ Thu, 01 Oct 2026 00:00:00 +0000 We intend to measure what a person brings; this version says how, and what has changed since 1 October.