Video summary
Full Beginners Guide to Writing Books With AI
Main summary
Key takeaways
Main Ideas, Concepts, and Lessons
1) Core Promise of the Course: write a book with AI without losing your voice
The creator positions the course as a step-by-step blueprint to write a full book using AI while preserving the author’s:
- Voice
- Quality storytelling
- Reduced burnout compared to traditional “manual everything” publishing
They emphasize that AI should be used to work smarter, not to avoid work.
The course is presented as covering this end-to-end flow:
- mindset
- tools
- templates
- brainstorming
- characters + worldbuilding
- plotting
- drafting
- editing
- formatting
- publishing
2) Author Credibility / Background Claims
The speaker claims:
- Experience writing 14 traditional books
- Experience managing content for authors at Kindlepreneur
They contrast themselves with “tech bros,” whom they claim haven’t written books themselves.
3) Mindset: burnout prevention and business thinking (Module 1)
Lesson A: Focus on the right success targets
- Burning out came from focusing too much on money (external and uncontrollable).
- Recommendation: assume you might not make money for a while and focus on:
- craft mastery
- improving output quality
Lesson B: Focus on what energizes you
Writing includes many steps, including unpleasant ones the speaker dislikes (e.g., “butt-in-chair typing”).
Suggested responses:
- eliminate unpleasant steps
- reduce them
- or outsource them:
- to humans (editors, designers)
- or to AI (for the drudgery parts)
Business mindset
- Treat writing/publishing as a storytelling-driven business if you want income later.
- Invest time and money into compounding assets:
- tools
- education
- systems
Abundance vs. scarcity
Scarcity examples:
- “The market is saturated”
- “If they win, I lose”
- “I only get one shot”
Abundance examples:
- Readers can be reached over time via multiple books
- Other authors’ success proves demand
- A failed launch yields data
Practical mindset practices
- daily meditation
- daily movement
- daily reading of books from successful people
4) Tools Overview and Setup Approach (Claude + n8n + Story Hacker OS)
The course uses a tool stack approach.
Claude (beginner-friendly chatbot)
Claude is recommended for writers because it’s viewed as better suited for writing than general chatbots.
Includes:
- chat interface
- research mode and web search
- model family selection (e.g., Haiku / Sonnet / Opus, plus others)
- an “effort” setting (higher models may think more but use more credits)
Key product concepts:
- Projects: upload files/text and provide reusable instructions so chats share context
- Artifacts/documents: AI can generate editable documents within the chat
- Skills/connectors: advanced customization (optional integrations)
n8n (automation system)
n8n is positioned as a time-saving and consistency-improving automation system that chains steps.
Differentiation from Claude:
- Claude needs manual prompting at each step
- n8n runs multi-step workflows automatically
Claimed benefits:
- integrate with Google suite and multiple AI providers
- run expensive tasks only when needed (use cheaper models for some steps)
- implement pipelines such as:
- generate a draft
- have a model check for “AI tells”
- have another model fix them
Hosting options discussed:
- cloud SaaS (simple, but pricier)
- local install (free, but not always-on and more technical limitations)
- self-hosted server (recommended balance: always-on, privacy, less cost than SaaS)
Technical setup flow highlights:
- local option:
- Docker Desktop
- run n8n container
- self-hosted option:
- use a hosting provider (example: Hostinger VPS plans)
Google credentials setup (described as hardest):
- create a Google Cloud project
- enable Gmail/Drive/Docs/Sheets APIs
- configure OAuth consent screen
- create OAuth client ID + client secret
- store secrets securely
OpenRouter integration:
- get an OpenRouter API key
- add it to n8n credentials
- use OpenRouter as a model router to access many model options
Story Hacker OS (speaker-developed writing system)
Presented as a “combined best of both worlds”:
- automation power of n8n
- chat/document flexibility of Claude
Core features:
- projects act like separate “books” with templates/context
- chatbot with access to selected context docs (e.g., characters/worldbuilding/outlines)
- personas (e.g., writing assistant, brainstorming assistant)
- “brain dump” → “start pipeline” style automations
- templates section (genre-specific packs)
Access is described as membership-based (not sold separately).
Detailed Methodology / Instruction Lists (By Process)
Course workflow (high-level)
- Module 1: Mindset
- Module 2: Tool stack setup
- Module 3: Templates (genre-specific frameworks first)
- Module 4: Brainstorming + story dossier
- Module 5: Characters + worldbuilding
- Module 6: Plot/outline
- Module 7: Drafting prose
- Module 8: Editing
- Module 9: Packaging (cover + formatting)
- Module 10: Publishing (Amazon/KDP)
Module-by-Module Methodology (What to do)
Module 1: Mindset (tactical guidance embedded in lessons)
Core guidance:
- prioritize craft mastery over short-term money outcomes
- redefine success to avoid burnout:
- assume a learning curve
- avoid “I must succeed immediately” thinking
- identify unpleasant steps and:
- eliminate
- outsource to professionals
- or outsource to AI
Business framing:
- treat the work as a business to create future income
- invest in systems and tools
Abundance mindset:
- treat other authors’ success as evidence of demand
- view experiments as learning
- treat each book as part of a long-term catalog
Recommended practices:
- daily meditation
- daily movement
- daily reading for inspiration/learning
Module 2: Tool stack (customization concept)
Decide which tool stack fits:
- beginner vs advanced
- need for automation vs need for control
- comfort level with technical setup
Use a customized “tool stack” instead of random app overload.
Module 3: Templates first (prevent generic AI output)
Instruction philosophy:
- don’t start with plotting/characters directly
- build genre-specific templates first so the AI has constraints
Rationale:
- without templates, AI defaults to generic suggestions (“AI slop”).
Template-building methods:
- Story hacking concept
- analyze existing works (scene-by-scene)
- extract tropes/themes/plot templates
- build genre guides from trends across multiple bestsellers
- DIY fallback (Claude-style research)
- ask for lists of major/minor tropes for a subgenre
- ask for a chapter-by-chapter template using those tropes
- iterate to refine themes, character archetypes, and worldbuilding elements
In Story Hacker OS:
- select a genre in templates
- create a project
- create a brain dump doc
- use the chatbot with the template context to generate ideas more specifically
Module 4: Brainstorming + story dossier (turn scattered ideas into structure)
Core instruction:
- establish genre/subgenre early
- niche down within established sellable genres
- do research by reading/perusing similar books (positioning, covers, marketing)
Brainstorming process (Claude):
- create a Claude project for a “story brainstormer” that:
- asks questions
- probes user ideas
- offers multiple suggestions
- ensures you end with:
- genre + style + tone
- novel scale
- protagonist
- list of characters
- world-building elements
- inciting incident and climax
- then produces a full dossier including:
- logline
- synopsis
- bullet lists (characters, worldbuilding elements, inciting incident, climax)
Funnel into a structured dossier.
Automation approach (n8n/Story Hacker OS concept):
- submit a brain dump (more structured, less interactive)
- pull in the right genre templates
- generate multiple pitches (example: 5 pitches) to avoid obvious output
- select the best pitch and create a dossier
- run automated “checks,” including:
- emotion check (emotional gut punch)
- character name check (avoid generic AI-like names)
- logic check (plot holes/inconsistencies)
Additional brainstorming prompts (examples):
- high concept pitches
- logline prompt (Blake Snyder-style, 25–30 words)
- expand an existing idea (100 possibilities list)
- what-if game
- intriguing first line prompt
- news-story-based story adaptation
- job titles → plot/character/setting/theme ideas
- “would you rather” scenario prompts
- plausible mashups
- social issue integration (subtle, not preachy)
- concise pitch format:
- “My story is about _ who must _ in order to ____.”
Module 5: Characters + worldbuilding (flesh elements using consistent prompts)
Character generation (Claude):
- use a character project prompt to:
- take dossier/brain dump
- create character sheets with:
- physical description
- role
- personality profiles
- core motivation
- background before story
- quirk/hobby
- dialogue style
- output only the requested details (avoid preamble)
Worldbuilding generation:
- use a prompt that requires elements categorized into:
- settings/locations
- objects/artifacts
- magic systems/technology
- groups/races
- gods/deities
- geography/nature
- population/politics
- culture/history/lore
- religion/beliefs
- languages
Rules:
- only use categories that apply
- each element must be fully fleshed with 3–4 sentences
- avoid generic “set dressing” only
Automation approach (n8n):
- use templates + dossier
- condense context as needed to reduce tokens
- loop through:
- each character individually (refine + logic check)
- each world element individually (refine + logic check)
- add additional checks:
- relationship map among characters (for later dialogue dynamics)
Story Hacker OS approach:
- use assistant personas (world building assistant / character assistant)
- or run dossier → characters/worldbuilding automation
- add baseline emotional/stress values:
- stress/calm baseline from -10 to +10
- later adjusted per scene intensity
Module 6: Plotting/Outlining (assemble a cohesive plot)
Outline philosophy:
- the speaker is a “hardcore outliner,” but discovery writers can adapt
- AI outlining needs guidance via templates to avoid becoming generic
Claude outlining instructions:
- prompt uses:
- dossier + character sheet + worldbuilding sheet + plot template
- requirements:
- follow the outline template carefully
- chapter summaries must be specific (no vague allusions)
- write as if handing to a ghostwriter for the first draft
- include only asked-for outline details
In n8n/Story Hacker OS:
- automations create outline documents with:
- trope summary
- full outline pass
- emotion check
- rewrite
- “plot sliders” rubrics (e.g., tension, dread, emotional intimacy, pacing energy, humor)
- logic check for continuity/plot holes
- second rewrite
Human step:
- review and rewrite vague chapters
- example: rewrite chapter 1 for clarity and grounding
- continue from there
Module 7: Drafting Prose with AI (and managing risk)
Attitude + safety guidance:
- you don’t have to use AI for prose, but it’s recommended for steps you dislike
- AI “slop” is “only slop if it’s bad”—effort is still required
- Amazon stance:
- AI-written text is allowed if quality is good
- disclosure must be accurate in Amazon’s AI-related fields
Drafting prompt requirements (Claude):
- when writing a specific chapter:
- write the entire chapter using:
- outline
- character sheet
- worldbuilding sheet
- story dossier
- use deep point of view
- “show, don’t tell”
- don’t rush; write a specific word count
- ask the user for desired word count
- follow stylistic rules (short punchy + long descriptive mixes, dialogue formatting, reduce metaphors, etc.)
- write the entire chapter using:
- continuity context:
- include only the previous chapter (or limited prior text), not the whole book
Automation drafting system (n8n) (advanced):
- loop over chapters
- for each chapter:
- estimate word count and adjust upward
- generate a “scene brief” with multiple subtasks:
- plot scene brief (beats/blocking)
- character scene brief (slider adjustments + continuity like injuries)
- worldbuilding scene brief
- chronology check vs outline
- rewrite scene brief
- write the first draft from the scene brief
- do a second chronology check using the most recent word window (example: last 20,000 words)
- do style check vs style guide
- rewrite and append chapter to manuscript doc
Operational guidance:
- don’t run the entire book at once
- do 1–3 chapters, then review before continuing
Story Hacker OS drafting:
- “chapter generator” automation:
- reads outline
- generates selected chapters
- supports POV/tenses (e.g., first/third-person limited/omniscient; past/present tense)
Module 8: Editing (quality control for AI output)
Editing types mentioned:
- developmental editing
- line editing
- proofreading (often copyediting is merged)
AI helps most with proofreading/line-level improvements, but may introduce “AI-isms.”
Top “AI-isms” list:
- Contrastive hook (“It’s not X, it’s Y”)
- Rhetorical question with instant answer
- Triple emphasis lists (“faster, smarter, better” cadence)
- Overused AI words (e.g., “delve,” “leverage,” “holistic,” “testament”)
- Inflated stakes for mundane topics
- Grand abstract metaphors (“tapestry of,” “landscape of,” “journey through”)
- Generic importance adjectives (“significant,” “impactful,” “key”)
- Bland high-level generalizations (floaty summary tone; lacks sensory specifics)
- Repetitive sentence templates/rephrasing
- Overly neat “bow-tied” conclusions
Editing method in Story Hacker OS:
- save AI output to project and read through the chapter
- reading strategies:
- read aloud
- optionally use a screen reader
- spot fix examples:
- cut over-dramatic metaphors
- tighten sentence structure
- adjust genre-appropriate language (e.g., remove a curse word early in cozy sci-fi)
- fix hallucinated/incorrect specifics (example given: “blood pressure” for an AI figure)
- use the “AI edit” tool for targeted edits (e.g., “leave dialogue, make the rest concise”)
Workflow suggestion after editing chapter 1:
- review outline for chapter 2 for continuity improvements
- edit chapter 2–N in small batches (write/edit/generate loop) to reduce continuity errors
Module 9: Packaging (covers + formatting)
Book covers:
- warning: a bad cover = bad sales (regardless of book quality)
- recommendation:
- match cover to genre conventions
- hiring an artist is generally preferred early on
- tools mentioned:
- Get Covers (budget option)
- Miblart (premium option)
Formatting:
- recommended tools:
- Atticus (Windows/Linux/cloud)
- Vellum (Mac; more expensive)
Workflow in Atticus:
- upload manuscript as a DOC file
- adjust chapter breaks/merges
- add front matter:
- dedication
- copyright templates
- images/maps
- table of contents (auto-generated)
- insert scene breaks
- add back matter:
- author note/about the author
- reader magnet / email list CTA
- sample of next book
- create and reuse formatting “themes”:
- templates per series
- print details (gutter margin, widows/orphans, typography, headers/footers, trim sizes)
Export:
- PDF and EPUB
Module 10: Publishing (Amazon-first, with expansion options)
Publishing strategy:
- Amazon dominates the ebook market (speaker claims ~85%, especially with Kindle Unlimited)
- recommended path:
- start with Amazon
- later decide whether to go wide via distributors/other retailers
KDP Select:
- enroll ebooks for Kindle Unlimited access
- tradeoff:
- ebook must be exclusive to Amazon (print can still go wide)
Other distribution options:
- non-Amazon platforms listed:
- Apple
- Barnes & Noble
- Kobo
- Google Play Books
- aggregators recommended:
- PublishDrive
- Draft2Digital (speaker favors it for generally lower fees)
Advanced: selling direct from an author store Claimed pros:
- keep ~95% revenue (vs ~70% or less via retailers)
- get buyer emails (helps build an email list)
- more control over delivery and bundling
Claimed cons:
- likely need ads
- harder with only a few books
- example recommendation: aim for ~12+ books
Amazon upload checklist:
- Prepare book description
- analyze a high-performing indie bestseller in your genre using Claude
- generate your own description using a similar template
- Prepare keywords
- use Amazon keyword suggestions