Bilton has ADHD. He's incredibly impatient. You would think this is a terrible tool for a writer working on 300-page books. He treats it as a feature. I lean into that.1
The method: write six random paragraphs that he already sees in his head. Get bored. Write six more paragraphs about something else. Then two more. Then a whole chapter. Copy and paste. Move things around. Put pieces in different places. At the end, force myself to pull it all together. That's the hard part.1
He builds books the way someone with ADHD does any sustained task — not linearly, but in bursts, with the structure imposed at the end rather than designed at the beginning.
Imagine a folder of 200 small files (or one giant document with 200 marked sections). Each file is a passage Bilton wrote in a single sitting because the passage was vivid to him in that moment. Some are dialogues. Some are scene descriptions. Some are paragraph-long analyses. Some are 2,000-word chapters. They're not in order. They're not labeled by chapter. They're labeled by content (or unlabeled, just chronological by composition date).
The composing move is to arrange these into a book. Bilton's character-anchor-traits (Rubik's cube guy, IRS three-time-reader) help him navigate his own draft material — he can find passages by character. The Tetris-database research provides the structural skeleton (what happened when, who was where) that determines which passages belong in which chapters.
The book emerges from the back end of this composition process. He didn't outline first. He drafted by impulse, organized by retrospective imposition.
Three reasons:
1. The energy is in the passage, not the structure. A passage written in the moment the writer sees it vividly is alive on the page. A passage written because the outline said it had to be written next is dead. Bilton's method maximizes the time spent on alive passages and outsources the structural assembly to a less energetic later phase.
2. Visual memory carries the structure. Bilton has a strong visual memory and weak name memory. I can still picture every detail visually, but I'm terrible with people's names.1 He can remember what's in his draft folder by image rather than by chapter-position. This makes the copy-paste reorganization mode work for him in a way it might not for a writer with different cognitive strengths.
3. The reporting is already done. The Tetris database has stored the structural facts. Bilton doesn't need an outline because the database is the outline — he knows when Ross was where, what happened on which date, what conversations preceded which decisions. The drafting can be impulsive because the structural backstop is in the database.
This method requires:
For writers without these affordances, Karlsson's incubation-cluster-fast-writethrough or Pink's process-engineering methods will produce better results. Bilton's method is a specialty case — not generalizable advice, but a specific operationalization for a specific cognitive style.
The hard part is the end. I have to like really force myself to pull it all together.1 This is where the linear writer's discipline reappears. The impulsive-drafting phase generated material; the pull-together phase imposes book-shape on the material.
This phase has its own moves:
The pull-together phase is structurally analogous to what a linear writer does while drafting — but Bilton outsources it to a discrete phase rather than carrying it through every paragraph.
Bilton and Macfarlane's mosaic-from-set-pieces converge completely on the architectural principle. Both treat the book as an assembly of pre-written set-pieces rather than a linearly drafted text. The convergence is striking because the two writers come from opposite ends of the prose tradition — Bilton from journalism and screenwriting, Macfarlane from poetry and nature writing. Both reached the same composing architecture independently. This convergence is evidence that mosaic-composition is the natural mode for any writer working on book-length material whose substantive content was generated non-linearly (Bilton's reporting; Macfarlane's field journeys).
The split is the source of the material. Bilton's set-pieces come from impulse-drafting against a database. Macfarlane's set-pieces come from notebook-fragments expanded at the keyboard via the pull-thread of memory. Same architectural composition, different upstream generation.
Bilton diverges from Karlsson's 5-step writing method on whether structure is pre-composed or emergent. Karlsson's method builds magnetic-field-shaping and cluster-sequencing as explicit pre-write structural moves; the writing-through is fast precisely because the structure was deliberately built. Bilton's method skips pre-write structural moves entirely; structure is imposed retrospectively. The split reveals two genuinely opposed theories of when structure earns its keep — Karlsson says before drafting; Bilton says after.
Jigsaw-puzzle non-linear writing maps onto two adjacent domains where the same impulse-driven-composition principle appears: behavioral-mechanics (the modular-message campaign that assembles pre-built creative units into a final campaign late in production) and business (the LLM-assisted writer who generates fragments and uses the AI as composition substrate).
Behavioral-Mechanics: Modular Campaign Asset Architecture — political and commercial campaigns are increasingly designed as a library of pre-generated creative assets (videos, headlines, social posts, talking points) that are composed into specific deployment plans late in the process based on what's working. The campaign doesn't pre-design the timeline; the campaign assembles the timeline from the asset library in response to feedback. This is structurally identical to Bilton's jigsaw method: generate units impulsively when they're alive; compose into final structure based on what fits. The structural identity: both crafts have moved away from upfront linear design toward asset-library generation plus late-stage retrospective composition because reality changes faster than upfront plans can accommodate. The insight neither domain generates alone: Bilton's writing method and modern campaign-asset architecture are the same composition philosophy operating in different mediums. The writer and the campaign-strategist are both betting that impulse-generated units will compose into better wholes than planned-linear assembly produces — and the bet has paid off in both domains as the production environment has accelerated. A novelist who studied campaign-asset architecture would build draft-libraries more efficiently; a campaign-strategist who studied Bilton would push asset-impulse capture earlier.
AI-Collaboration: LLM as Standardization Pressure — the LLM-assisted writer who generates many candidate passages and assembles them into structure (rather than writing linearly with LLM autocomplete) is operationalizing Bilton's jigsaw method with AI-generated units. The structural identity is the same: generate fragments fast; impose structure retrospectively. The insight neither domain generates alone: Bilton's jigsaw method anticipates the LLM-era composition workflow — the writer who has been drafting in jigsaw mode for twenty years is structurally better positioned to integrate LLM-assisted draft-generation than the writer who drafts linearly. The cognitive habit of impulse-write-then-compose is the same habit, regardless of whether the impulse comes from the writer's brain or from an LLM-prompted passage. A writer who has internalized Bilton's method has a head start on AI-assisted prose because the composition architecture is already in place.
A third connection worth marking: jigsaw-writing has structural similarity to Karlsson's cluster-sequencing (discussed in Author T&C) but inverts the temporal logic — Karlsson sequences before writing; Bilton sequences after.
The Sharpest Implication
If Bilton's jigsaw method is what works for an ADHD-driven writer with strong visual memory and a comprehensive reporting database, then writing-method generalization advice (the kind of universal "outline first, then draft" or "write a daily word count" rules) is structurally lying — different writers' cognitive constitutions and material-types demand different composing architectures, and the wrong method applied to the wrong writer produces blocked-up output. The uncomfortable implication: the dominant writing-advice tradition (Pink-style discipline, daily word counts, outline before drafting) is silently optimized for one cognitive style and one material-type, and writers who don't fit either are being told their failure-to-produce is a discipline problem when it's actually a method-mismatch. Bilton's lean-into-your-cognitive-style principle is the deeper move: find the method that fits your particular brain plus your particular material, even if that method violates received writing advice.
Generative Questions