Specialist Tracks: future offerings. Join the list for updates.
Specialist Track · Future Offering

Fiction Writer Track

Author voice, character voice, and scene-level decomposition for sustained narrative work.

Self-paced · 3 modules · $699 · Pedagogy: Workshop / Studio
Waitlist

The situation you recognize

You've been working on this novel for months. Dozens of sessions. The world is built, the characters are real, but the voice has drifted. You can't point to the chapter where it shifted. Some scenes feel like you. Others feel competent but generic — the AI's average version of how someone in your genre writes. Your character spoke one way in chapter three and a subtly different way in chapter nine, and neither version is exactly wrong, but neither is fully yours.

What this means for fiction

Fiction breaks every AI assumption. Sessions expire. Context windows fill. The model doesn't remember your character's voice from last Tuesday unless you bring it. Over months, drift compounds: the author's voice flattens, character consistency frays, and the manuscript starts sounding like a committee wrote it in the same general direction. The Accommodation Design practices solve these problems individually. This track wires them together for the specific demands of sustained narrative work — a manuscript with real continuity, real characters, and a voice that has to stay yours across the full arc.

The distinction between author voice and character voice matters here. Your author voice governs the prose. Each character has their own voice that must stay consistent across scenes. Both need protocols. This track builds both.

Module 1

Input Inversion for Fiction

Workshop: brain dumps as raw material for character and world

In fiction, input inversion works at two levels. The first level is the author: your unstructured thinking about the story — voice memos about a scene, arguments with yourself about a character's motivation, half-formed observations about the world — contains richer material than any structured prompt. The second level is the character: how does Henka think when you're not performing her? What does she say when you're just talking, not writing?

You build an input practice that accumulates over the manuscript. Voice notes become character development. Unstructured rambles become worldbuilding archives. By chapter ten, the AI draws from months of your thinking about this specific world and these specific people, not just whatever you can fit into a session prompt.

Deliverable: A working corpus strategy for your manuscript. Input channels, character archives, accumulation method.

Module 2

Voice Governance for Prose

Studio: author voice and character voice, extracted and encoded

Your author voice lives in your conversation, not your published prose. Published writing has been edited toward a final state. The raw version — how you talk about your characters when no one is reading — is closer to what makes your writing yours. This module teaches you to extract both your author voice and your characters' voices from conversation and build protocols that constrain generation during production.

The protocol runs before the first sentence. Sensory priority, emotional logic, sentence rhythm, what you refuse — encoded before generation. The AI's defaults occupy none of the space. Character voice protocols do the same thing: specific speech patterns, decision logic, what each character would never say. Consistency is structural, not editorial.

Deliverable: An author voice protocol and at least two character voice protocols, tested against your manuscript.

Module 3

The Full Pipeline

Studio: wire the entire writing process into one system

This module connects everything. Scene-level decomposition runs specialists independently: setting, interiority, dialogue, plot beat — each given full attention before you provide the synthesis. A narrative lens array evaluates the work from multiple critical perspectives simultaneously. Continuity management keeps characters, world logic, and voice consistent across dozens of sessions. The coordinator wires the full process into one system that runs your methodology so your attention stays on the story.

Deliverable: A working fiction pipeline applied to your own manuscript. Voice protocols, decomposition practice, lens array, continuity system.

Where this comes from

This track was built from direct experience. Peter developed the methodology inside a novel across 40+ sessions: kishotenketsu as narrative structure lens, voice governance constraining prose generation, savepoints holding character continuity across months of work. The practices are not theoretical. They were developed inside the work they describe.

This track is a future offering. Pricing and structure to be announced. See all offerings.