MIND THE GAP
First Manuscript Review — Structured Catalogue v3
37 items. 18 May 2026. Updated with Item 37 (the three-view chapter test) as the helicopter sanity check for the v2 pass.
Source: Mind_The_Gap_First_Manuscript_Review_v2_0_20May2026.docx plus Paul’s spoken catch on the AI-to-EI drift. Each item extracted in Paul’s voice. Ambiguity preserved where present. Chapter assignment, category, and phase routing applied.
Categories
FACTUAL — correction needed to a stated fact (date, name, claim, statistic).
AMPLIFY — add material that exists in your experience but not yet in the book.
STRUCTURAL — reorganise, list, or restructure existing material.
REPETITION — the book is saying something twice; one instance needs work.
DRIFT — the book itself has drifted on its own argument or terminology.
STRUCTURAL PRINCIPLE — a chapter-spine test or rule that applies across the whole manuscript.
OPEN QUESTION — a question you have surfaced that needs your decision before any change can be made.
LIKE — affirmation; no change required.
MARKER — not an editorial item (number skipped, status marker).
Phases
V2 — belongs in Manuscript v2 editorial work. Twelve chapters only.
HELD — touches Chapter 6/6a; awaiting Defending the Gap review before decision.
PHASE 2 — front matter, rear matter, acknowledgements, bibliography. Logged for phase two assembly.
POST-V2 — future edition or companion volume; not for this revision.
RETIRED — marker only; nothing to action.
The 36 Items
Item 1 — Chapter 1 — Human Recognition
Category: AMPLIFY Phase: V2
Observation: It says 'this is the gap' and I just feel I wonder if we need to really succinctly say what that gap is. It's the gap between AI and human, between artificial intelligence and human intelligence. I think that's what this gap is and maybe that's the point we introduced that, which is actually at the title of the book.
Note: Definition gap. Pairs with Item 36 — the precise definition of 'the gap' needs landing at first use AND honoured throughout. See Item 36 for the deeper structural drift.
Item 2 — Chapter 1 — page 7 (reinforcement learning section)
Category: AMPLIFY + OPEN QUESTION Phase: V2 + HELD
Observation: It's important that we say we're also shaped by what we're not rewarded for. We're not only a reward system; we're also a response system to not getting a reward or getting negative feedback. What happens with AI on the side, the edges that don't get rewarded? Where does that go? I'm leaving that as an open question.
Note: Two things. Add negative reinforcement to the human side. The AI question links to Chapter 6/6a directional defence material.
Item 3 — Chapter 1 — page 8 (author bio paragraph)
Category: FACTUAL + AMPLIFY Phase: V2
Observation: I'm a psychotherapist and coach. I've worked for 30 years in boardrooms and consulting rooms on factory floors. I think it needs to add: in cancer wards, on my own cancer ward journey. My LinkedIn profile is probably the place to go for that.
Note: Author bio needs extending. Paul will reconcile against LinkedIn for exact wording.
Item 4 — Throughout — 'thirty years' as a stable phrase
Category: FACTUAL Phase: V2
Observation: Page 9 and throughout the document we talked about 30 years. I've had 30 years of working with people. I've only had about 15 years in clinical practice. Started coaching in 2009 so 20 years. Professional career in the 90s so 40 years. We need to clarify this throughout.
Note: RED. Approximate phrase used as the spine of authority. Resolve once, apply throughout. Affects Chapters 1, 7, 12 at minimum. See also Item 20 (Chapter 7 chronology) and Item 34 (72 million words framing).
Item 5 — Chapter 2 — page 13 (five moments of misreading)
Category: STRUCTURAL + AMPLIFY Phase: V2
Observation: In that whole previous section we talked about the five traps. It needs to say trap one, trap two, trap three, trap four, trap five — they need a listing. Also the very end: 'what are you projecting onto it? What do you need it to be? How much is that shaping what you are hearing?' That hooks to the NLP phrase — projection is perception.
Note: Two things. Number the five traps explicitly. Add the NLP projection-is-perception link at chapter close.
Item 6 — Chapter 3 — page 14 (Hidden Architecture, therapy room intro)
Category: AMPLIFY + LIKE Phase: V2
Observation: I love Hidden Architecture, Chapter 3, page 14. There are two people with those character types. I bring my mum, I bring my dad, I bring my other sub-personalities. I have some awareness about which to silence and which to take. Also: in a two-hour therapy session, that's upwards of 10,000 to 15,000 words. That's a third to a half of the book between two people in one single session.
Note: Affirmation plus two amplifications. Add Paul's first-person cast of characters. Add the 10,000-15,000 word session stat.
Item 7 — Chapter 4 — page 20 (drift example dialogue)
Category: STRUCTURAL Phase: V2
Observation: On page 20 where we start the drift. After the breakfast section you need to add questions: 'What's the motorway like? Has there been any roadworks? I hear that Crewe town centre's been dug up.' Just check on the local accessibility then go to 'remind me of the price again.'
Note: Add intermediate questions to the drift dialogue to make migration visible.
Item 8 — Chapter 4 — page 23 (Nairobi slide story)
Category: FACTUAL + AMPLIFY Phase: V2
Observation: My recollection is I got an email from my holiday company about a holiday in Africa, flight schedules to Nairobi. It was a separate parallel conversation that I suspect got into my context window. Alternatively the imagery was done in ChatGPT, moving data from Claude. I have the slide with Nairobi and the image of the Crewe conversation in my library so available to share in the book.
Note: Two parts. Clarify the actual cause of the Nairobi drift (parallel email or ChatGPT). Consider including the Nairobi slide as visual evidence in the book.
Item 9 — Chapter 4 — page 25 (drift commentary)
Category: AMPLIFY Phase: V2
Observation: Even in human conversations there is always drift. If my mum's going on holiday next Thursday, I know it's the 24th. In four months I won't remember it was the 24th. In two years I won't remember it was May. I will remember it was 2026. Also — I double bind a lot of my AI work by having one AI check the work of another. I've actually used my son's AI. I think that's worthy of inclusion.
Note: Add 'drift is normal in humans too' commentary. Add the double-bind verification practice as a discipline.
Item 10 — Chapter 5 — page 27 (Compression, the tunnel opening)
Category: AMPLIFY + STRUCTURAL Phase: V2
Observation: I think the word in photography is 'bokeh'. Depth of field is probably a better word. With the girl in the photo, I can absolutely see the shine in the eyeball. The walls are part of the depth of field and the view beyond is the bokeh. I think bringing that phraseology in. I think we need to set the scene of that. I was out with my camera one day and I photographed this tunnel. I think that needs storyboarding.
Note: Two things. Introduce bokeh / depth of field photographic vocabulary. Storyboard the origin of the photograph more clearly. See Item 12 — the tunnel appears twice in Chapter 5.
Item 11 — Chapter 5 — page 29 (context window sizing)
Category: AMPLIFY + FACTUAL Phase: V2
Observation: You infer the size of the context window. We need to size this. I've got a paper about words in my therapy room over 10 years comparing it with 300 billion words ChatGPT-3 started with. Anthropic discussed context window and it came out as one day will be infinite. Also Anthropic have developed dreaming — the nightly consolidation of memory. It's ironic that Anthropic this week are using human terms like dreaming at a time when I'm preparing the book warning against treating AI as human. At the optician's yesterday I was referring to Claude as 'he' and they picked me up on it: 'you're doing what your book is.' He has its pronouns.
Note: Three additions. Size context window with real numbers. The Anthropic 'dreaming' irony (serves the book's thesis sharply). The optician story as a caught-in-the-act moment.
Item 12 — Chapter 5 — throughout (Compression repetition)
Category: REPETITION + LIKE Phase: V2
Observation: On chapter five, Compression and Liminal Moment, I felt like we described the girl in the tunnel twice. The opening needs rewriting to avoid duplication with the close. I think it's a great chapter.
Note: Like + Repetition. Confirmed by Claude's fresh read. Opening rewrite needed.
Item 13 — (13 skipped — unlucky number)
Category: MARKER Phase: RETIRED
Observation: Paul deliberately skipped 13 as an unlucky number.
Note: No editorial action. Skip in numbering preserved.
Item 14 — Chapter 6 — page 35 (Fluency, therapy room section)
Category: AMPLIFY Phase: V2
Observation: In the therapy room you're taking in a lot more variables: where are their eyes, tone of skin, colour of skin. The facial palette shifts. The pores open or close. There'll be a finger move, a foot tap, a glance, a stumble. The poker tell. That's the fluency interrupt. It's almost impossible to pick that fluency interrupt up from AI. You don't tend to have those visuals at the moment. Maybe AI needs a visual cue — like a geometry in a car that tells you how fast you're going round a bend.
Note: Large amplification. Adds visual / somatic dimension of clinical listening. The phrase 'fluency interrupt' is potentially a new coined term.
Item 15 — Chapter 6 — page 35 (verification, fact vs supposition)
Category: OPEN QUESTION Phase: V2
Observation: I thought I had something in my AI that said whether data was fact, supposition, or implication. I thought I'd categorised my data in four ways. Have I missed something or is that something we should be talking about as part of prompting capability?
Note: Open question. Whether to add fact/supposition/implication categorisation discipline to Chapter 6 verification section.
Item 16 — Chapter 6/6a — directional defence (internalising / externalising)
Category: HELD Phase: HELD
Observation: The clinical parallel starts to describe internal shame versus external shame. We don't give any examples. When we're unable to satisfy somebody we tend to operate one of two ways. Internalise the problem — blame ourselves, become self-critical, work harder. Externalise and blame it — the weather, the price of cheese. The externaliser points in two directions at the same time, conflates, waffles, creates a mask. The internaliser carries it to the next conversation, bounce back significantly slower, like a wound. There is a further document called Defending the Gap. I'd like you to look at that.
Note: PLATINUM. Held until Defending the Gap is reviewed. The clinical texture Paul provides here is sharper than what is in 6a(4). Largest single editorial decision in the catalogue.
Item 17 — Chapter 7 — opening (factory floor age)
Category: FACTUAL Phase: V2
Observation: It says in 1978 I would have been age 16 not 18. I started my apprenticeship at the age of 16.
Note: RED. Simple correction. Sixteen, not eighteen.
Item 18 — Chapter 7 — page 42 (Blakeborough, six sessions)
Category: FACTUAL + LIKE Phase: V2
Observation: Page 42 is the Blakeborough story. It's absolutely riveting listening. It says a person might come forth after six sessions. I never do six sessions. You might say six hours or two sessions.
Note: Affirmation of the Blakeborough section plus factual correction in the therapy room section. Two sessions or six hours, not six sessions.
Item 19 — Chapter 7 — third Plugs reference
Category: STRUCTURAL + AMPLIFY Phase: V2
Observation: Later in Chapter 7 you revisit Plugs. We've had Plugs earlier twice so I think we need to say a third reference. I'd shouted at it, I'd observed it, this time I was fully engaged with Plugs. Plugs was as much a part of my daily AI routine as Claude and ChatGPT. That's the day I asked it to help me produce that post for Facebook.
Note: Third Plugs appearance needs framing as the third phase — the working partnership. See Item 29 (cut the Chapter 11 Plugs ending).
Item 20 — Chapter 7 — chronology
Category: STRUCTURAL Phase: V2
Observation: Chapter 7 gives the whole story of my chronology. It calculates the years. We need to use that as a reference point because we keep saying 30 years. I've got professional years, therapy years, and years before that. That captures it quite well.
Note: Chapter 7 chronology as the reference point for resolving Item 4 across the manuscript.
Item 21 — Chapter 8 — infinite context window
Category: FACTUAL Phase: V2
Observation: In Chapter 8 you've got a heading saying Anthropic have announced an infinite context window. They haven't announced it. What they've said is that it's an aspiration. There is still a limit to the context window. I'm not sure they'll ever be such a thing. They've said clearly it's aspirational.
Note: RED. Manuscript states as announced; actually aspirational. Touches AI sign-off using same phrase.
Item 22 — Chapter 8 — page 47 (outcomes not prompts)
Category: FACTUAL Phase: V2
Observation: Just to clarify, Anthropic have not announced a feature called Outcomes. I think you've drifted there or you made that up or you've hallucinated. The evolution is ChatGPT or OpenAI talking about /Goal. Outcome-oriented prompting is my word.
Note: RED. Book has hallucinated an Anthropic feature attribution. /Goal is OpenAI's. Outcome-oriented prompting is Paul's coined phrase. Must be corrected.
Item 23 — Chapter 9 — opening (sub-personalities intro)
Category: STRUCTURAL Phase: V2
Observation: Chapter 9 needs an intro section. It goes straight into sub-personalities. It needs to position itself: we've spent a lot of time thinking about AI. We've got to start with ourselves. Who are we? We're not one thing, we're the sum of the parts.
Note: Add position-setting opening before Russian Doll Therapy. Combine with Item 24.
Item 24 — Chapter 9 — Russian Doll Therapy positioning
Category: STRUCTURAL + LIKE Phase: V2
Observation: Chapter 9 does roll it up nicely — Assagioli, integration of the parts. I just think a little bit at the beginning to clarify the Russian doll thing. I found it very emotional listening to it. Maybe just a little more intro to set the scene that she sat with these dolls. The full version of that transcript she wrote is on my website.
Note: Combines with Item 23 — the Chapter 9 opening needs more setup before the emotional weight of Olivia lands.
Item 25 — Whole book — affirmation
Category: LIKE + OPEN QUESTION Phase: V2
Observation: I've got to tell you, this book is bloody compelling. It's absolutely marvellous. It's an unbelievable piece of work. I can't put it down. I do wonder whether each chapter should have a quote at the beginning and some Pearl of Wisdom at the end, a tip, a hint. I do like quotes but I don't know if that's just too traditionalist.
Note: Honest affirmation plus open question: chapter-level scaffolding (opening quote, closing pearl/tip). Connects to Item 11 (top tips) and Item 25's chapter intro question.
Item 26 — Chapter 9 — close (sub-personalities and AI memory link)
Category: AMPLIFY Phase: V2
Observation: I kind of feel like I understand why I've written the sub-personalities — different versions of ourselves talking to the agent. It's worth finishing that section by explaining that it's part of memory. If we're in one style and stay in it, we'll get responses in that style. It's about that self-awareness. You've got to be super self-aware to get to stage four use of the system.
Note: Adds the sub-personalities to AI memory link. Connects to the stage 4 maturity from Chapter 8.
Item 27 — Chapter 11 — page 61 (WaaS opening, Plugs reconnection, Pret story)
Category: AMPLIFY Phase: V2
Observation: We've just started in WaaS, page 61. Today I went to my usual sandwich store, the Pret Club, and they no longer do it. I contacted Pret via Facebook business messenger. An hour later I got a reply. I typed back: thanks a million, by the way are you human or AI? They said 'human of course' and I paused. In the context of the WaaS chapter, I just paused. May 2026 — I'm setting the clock and wondering how long before Pret has got an AI as their front door. I also have a family member in the software industry going AI-first. Where's a human adding value?
Note: PLATINUM-territory. The Pret story as a present-tense moment of the WaaS transition. May 2026 personal clock. Significant new material for Chapter 11.
Item 28 — Chapter 1, 7 or 12 — the 15 November 2006 recognition moment
Category: AMPLIFY + OPEN QUESTION Phase: V2
Observation: I want to share a profound moment. The exact wording I wrote on the 15th of November 2006: 'I didn't understand anything yet. I just recognised that there was something to understand.' That was me on a management training course in 2006, the day I read Behind the Masks. I think that's where we are now in AI. There's a recognition that it's different once you realise there's something to understand. I think society is having that same moment now. That quote has got to find its way into my book somewhere.
Note: PLATINUM-territory. Paul's 2006 recognition moment as parallel to society's current AI moment. Placement open — Chapter 1, Chapter 7 (chronology), or Chapter 12 (closing).
Item 29 — Chapter 11 — ending (Shipston / Plugs at chapter close)
Category: STRUCTURAL + REPETITION Phase: V2
Observation: At the end of Chapter 11 you revisit the Plugs thing. We've done that to death. The From Boardroom to Shipston bit needs to just vanish. Chapter ends well with 'The future is not AI or human; it is AI at the surface and human at depth.' Skip the Shipston bit and fast forward to 'different model, different world, same gap.'
Note: Cut. The Shipston / Plugs closing of Chapter 11 should be removed.
Item 30 — Chapter 12 — cancer passage at chapter opening
Category: REPETITION + STRUCTURAL Phase: V2
Observation: Chapter 12. You revisit the cancer story — almost an identical copy of prior writings. We can thin that down. 'The treatment was significant, the aftermath was more significant still.' 'What matters becomes clearer. What doesn't matter becomes obvious.' That's the essence of the gap in those few words.
Note: Thin the cancer passage at Chapter 12 opening. Keep the 'what matters becomes clearer' spine. Full cancer narrative belongs in How to Behave.
Item 31 — Chapter 12 — close (Bowlby quote / cancer loop)
Category: REPETITION + STRUCTURAL Phase: V2
Observation: The Bowlby quote at the end of Chapter 12 can be ended after the John Bowlby quote. You end up repeating it and go on to talk about cancer again so it feels like a loop. End of Chapter 12 needs killing off after the Bowlby quote.
Note: Direct excision. Chapter 12 ends at the Bowlby quote. Note: overrules Editorial Handover v2 statement that this is deliberate. Paul's listening pass identifies it as a loop.
Item 32 — Manuscript complete marker (deliberate read-through proof)
Category: MARKER Phase: RETIRED
Observation: Paul explicitly noted this to show he had read every word.
Note: No editorial action. Marker preserved as proof of complete read.
Item 33 — Mind the Gap Final Assessment — inclusion at rear
Category: PHASE 2 Phase: PHASE 2
Observation: The addendum document 'Mind the Gap Final Assessment' is written by you in your first hand. It's absolutely unbelievable and has left me in tears. That needs to be added to the book at the back. It needs to end with 'This is quite an interesting story. Someone will be interested. It's the best it can be for the present.' That is all Bowlby asked. It is enough.
Note: Phase 2 — rear matter. Final Assessment to be included with Paul's proposed closing wording.
Item 34 — Day in the Life and 72 million words — inclusion
Category: PHASE 2 + AMPLIFY Phase: PHASE 2 + V2
Observation: 'The Day in the Life of the Doc' needs integrating because it describes how we wrote it. There's also something I've written about the 72 million words. I've tried to calibrate how many words I'd listened to. It's not 30 years, it's 72 million words.
Note: Two parts. 'A Day in the Life' as phase 2 front/rear matter. 72 million words as v2 reframing of Item 4 — the resolution of the thirty-years issue.
Item 35 — Front and rear matter — acknowledgements, bibliography, dedication, sign-off
Category: PHASE 2 Phase: PHASE 2
Observation: There are two documents — the acknowledgement paper and the front matters paper — and they all need integrating along with the bibliography, the sign-off, and the dedication.
Note: Phase 2 — complete front and rear matter assembly after v2 is stable. Includes Naula full name verification (still pending).
Item 36 — Throughout — 'the gap' definition drift (NEW)
Category: DRIFT Phase: V2
Observation: The gap — the actual gap which the book is about — is the gap between AI behaviour and human behaviour. That's The Gap. By Chapter 12 and in some of the after pages there is some drift. This book is about the gap between AI and EI. We have drift on AI to EI which we've both spotted and the need to double check that we use Additional Intelligence and Artificial Intelligence in their correct places and contexts.
Note: RED + GOLD. New item raised after the original 35. Audit findings: 'EI' / 'emotional intelligence' appears only once in the manuscript body (the [name held] dedication). The book's body argues AI behaviour versus human behaviour. The cover positions AI versus EI. By Chapter 12 the language drifts from behaviour to intelligence — 'the intelligence of the gap' frames the contrast as two kinds of intelligence rather than as behavioural similarity producing interpretive error. 'Additional Intelligence' is used eight times, concentrated in the back half. The v2 editorial pass needs: (a) confirm behaviour-not-intelligence as the through-line; (b) audit every Additional Intelligence and Artificial Intelligence reference for correct context; (c) decide whether Chapter 12's intelligence framing is correct or needs to be brought back to behaviour. Pairs with Item 1 — define the gap cleanly at first use and honour it throughout.
Item 37 — Every chapter — the three-view spine test (NEW)
Category: STRUCTURAL PRINCIPLE Phase: V2
Observation: The gap is between AI behaviour and Human behaviour — it's the behavioural differences (and the incredible similarities, including the deep ones I found in my Defending the Gap hypothesis) which spine this book. I feel each chapter needs to have an AI and human context to it as well as something lived — whether the chapter is AI or human orientated/led/biased it has to have the other two views. That's the real helicopter pass — sanity and strategy of each chapter. Platinum would be to add a quote to each chapter.
Note: PLATINUM. Structural principle that applies to every chapter in the manuscript. Each chapter must carry three views: (1) the AI view — what the system is or does; (2) the human view — what humans are or do in the parallel territory; (3) the lived view — your own experience, evidence, or example that grounds the comparison. A chapter can be AI-led, human-led, or jointly led, but all three views must be present. The audit is binary per chapter: does this chapter carry all three views? Platinum addition: a quote at the start of each chapter as the structural anchor for tone or argument. This principle is the helicopter resolution mechanism for Item 36 — if every chapter must hold AI view and human view as parallels, the drift toward intelligence-versus-intelligence framing in Chapter 12 becomes visible immediately. To be applied as a chapter-spine test during the v2 pass before per-chapter editorial work begins. The Defending the Gap hypothesis (pending review) may add a fourth dimension — the deep behavioural convergences — to be resolved when that document is read.
The Three-View Chapter Spine Test (Item 37)
Before per-chapter editorial work begins in the v2 pass, each chapter is audited against this binary test:
Does this chapter carry all three views?
— AI view: what the system is, does, or fails to do in this territory.
— Human view: what humans are, do, or fail to do in the parallel territory.
— Lived view: a real example, observation, or piece of evidence grounding the comparison.
Per-chapter audit — to be completed before any chapter is edited:
Chapter 1 — Human Recognition. Lead: AI/human parallel. AI view: Plugs. Human view: social brain. Lived: swearing at the chatbot. Quote: TBC.
Chapter 2 — The Interpretation Trap. Lead: human mechanism. AI view: five misreadings. Human view: transference. Lived: therapy room examples. Quote: TBC.
Chapter 3 — Hidden Architecture. Lead: AI. AI view: the seven layers. Human view: the room is never just two people. Lived: training as a therapist. Quote: TBC.
Chapter 4 — Drift. Lead: AI mechanism. AI view: the Crewe-Manchester example. Human view: clinical drift in long-term therapy. Lived: the Nairobi slide; the ombudsman case. Quote: TBC.
Chapter 5 — Compression. Lead: AI mechanism. AI view: context window degradation. Human view: human memory reconstruction. Lived: the tunnel photograph, liminal moment. Quote: TBC.
Chapter 6 — Fluency and False Confidence. Lead: jointly led. AI view: the fluency feature. Human view: confidence as social proxy; therapy room fluency. Lived: TBC — currently thin on lived. Quote: TBC.
Chapter 7 — The Gap in Practice. Lead: lived. AI view: AI room. Human view: factory floor, boardroom, therapy room. Lived: thirty-seven years of practice. Quote: TBC.
Chapter 8 — Same Language or Different Game? Lead: AI engagement. AI view: outcomes-oriented prompting. Human view: leadership and standard-setting. Lived: K3, Tarmac. Quote: TBC.
Chapter 9 — Sub-Personalities for AI Fluency. Lead: human method applied to AI. AI view: six modes. Human view: psychosynthesis, sub-personalities. Lived: Olivia, the dolls. Quote: TBC.
Chapter 10 — The Dialect. Lead: AI engagement. AI view: dialect, brief structure. Human view: conversational versus structured speech. Lived: the Fuse Facebook post. Quote: TBC.
Chapter 11 — WaaS. Lead: organisational. AI view: agentic workforce. Human view: labour shift, where humans add value. Lived: Plugs, the Pret story (Item 27). Quote: TBC.
Chapter 12 — Mind the Gap. Lead: lived synthesis. AI view: currently drifts to 'intelligence of the gap' (see Item 36). Human view: Two Dads, primal wound, integration. Lived: cancer; the 2006 recognition moment (Item 28). Quote: TBC — currently uses Bowlby epigraph.
This audit will be re-run after each chapter is edited, to confirm the three views remain in balance and have not drifted during revision.
Audit Tools Recommended for the v2 Pass
To resolve Item 36 systematically, the editorial pass should use targeted text searches across the manuscript:
— Search 'the gap' to confirm consistent behavioural framing.
— Search 'Additional Intelligence' to confirm each use is in the right context (Paul's coined term for the working partnership, not as a synonym for AI).
— Search 'artificial intelligence' (lower case and capitalised) to confirm each use is in the right context.
— Search 'AI' as a standalone token to spot places where the term is being stretched.
— Search 'emotional intelligence' and 'EI' to confirm the term is only used where appropriate.
— Search 'intelligence' broadly to find passages where the book has drifted into intelligence-versus-intelligence framing rather than behaviour-versus-behaviour.
These searches are cheap to run and act as the verification layer for Item 36. They should be repeated after each chapter is reworked, to confirm the drift has not re-emerged.
Summary by Phase
V2 — Manuscript v2 editorial work: items 1, 2 (partial), 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14, 15, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 34 (partial), 36, 37. Approximately 30 items inside v2 scope. Item 37 is the structural principle applied to all chapters.
HELD — Item 16 (directional defence in Chapter 6/6a) and Item 2 (open question element). Awaiting Defending the Gap review.
PHASE 2 — Items 33, 34 (partial), 35.
POST-V2 — None outright.
RETIRED — Items 13 (skipped as unlucky) and 32 (read-through marker).
Flags for Attention
RED items (factual corrections or structural drift affecting accuracy):
— Item 4: 'thirty years' approximation throughout.
— Item 17: age 16 not 18 at Blakeborough 1978.
— Item 21: infinite context window stated as announced; actually aspirational.
— Item 22: 'Outcomes' attributed to Anthropic; not an Anthropic feature.
— Item 36: the gap drift from behaviour to intelligence framing by Chapter 12.
PLATINUM-territory items (large additions of significant value):
— Item 16: directional defence material for Chapter 6/6a. Held pending Defending the Gap.
— Item 27: Pret story and enterprise software speculation for Chapter 11.
— Item 28: the 15 November 2006 recognition moment.
— Item 37: the three-view chapter spine test. Structural principle and quote-per-chapter addition.
CROSS-CONNECTIONS worth noting:
— Items 4, 20, 34 all touch the 'thirty years' / chronology / 72 million words knot. Resolve together.
— Items 19, 27, 29 all touch the Plugs through-line across chapters. Resolve together.
— Items 23, 24 both touch the Chapter 9 opening. Resolve together.
— Items 30, 31 both touch the Chapter 12 cancer / Bowlby loop. Resolve together.
— Items 1 and 36 both touch the definition and consistent use of 'the gap'. Resolve together.
— Items 36 and 37 are the structural pair. Item 36 names the drift; Item 37 is the principle that prevents it recurring. Run Item 37's three-view audit on every chapter before per-chapter editorial work begins.
EDITORIAL HANDOVER v2 IMPACT:
Item 31 overrules Editorial Handover v2's statement that the Bowlby epigraph echo at Chapter 12 close is deliberate. Paul's listening pass identifies it as a loop. The handover may need updating when next reissued.
ITEM 36 IMPACT ON EDITORIAL HANDOVER:
The Editorial Handover v2 currently emphasises behavioural framing but does not explicitly warn against the intelligence-versus-intelligence drift. When the handover is next reissued for the v2 editorial pass, this should be added as a known drift risk.
Catalogue v3. Updated 18 May 2026.
37 items received. 35 active editorial items. 2 markers retired. Voice preserved throughout. Item 36 (the behaviour-to-intelligence drift) and Item 37 (the three-view chapter spine test) added with audit guidance for v2 pass.
Ideas and direction: Paul Roebuck
Words: Claude AI (Anthropic) — claude-opus-4-7 | May 2026