Writing as the Primary Interface Tutorial

A practical tour of writing treated as the interface it has quietly become: the medium through which you think, coordinate with your team, reach the public, and direct AI.

Who This Is For

• Anyone whose job is more typing than talking, which is most knowledge workers in 2026
• Founders and operators building teams that work async
• People using AI tools daily and wanting the prompts to actually work
• Writers whose craft you'd respect (newsletters, essays, blogs) who want frameworks beyond intuition
• Students and early-career professionals who never had "writing for work" explained

This tutorial isn't about literary craft. For that, the sibling tutorial at content/writing/ is better. This one is about writing as an operator: writing that does work.

Contents

Fundamentals

1. Introduction: Why writing became the primary interface of work in the 2020s
2. The Four Audiences: Yourself, your team, the public, and AI

Core Concepts

3. Clarity: The core skill: unambiguous, self-contained, precise writing
4. Writing as Thinking: How writing externalises and sharpens thought
5. Writing for Async Teams: Memos, docs, PRs, tickets, decisions
6. Writing for AI: Prompts, specs, and what AI rewards

Advanced

7. Structuring a Document: How to organise anything longer than a paragraph
8. Editing: Drafting vs editing, the cuts that matter, revision as craft
9. Naming and Framing: Names, labels, titles, taglines as compressed interfaces

Ecosystem

10. Tools and Workflow: Editors, notes, outlines, templates
11. Writing Over the Long Run: Managing years of written output: searchable history, templates, archives

Mastery

12. Best Practices: Habits, anti-patterns, and the long arc

How to Use This Tutorial

1. Read sequentially. The four audiences frame everything after; skip them and later chapters land oddly
2. Apply to your current writing. Take a real doc, memo, or prompt you wrote this week and run each chapter's lens over it
3. Notice your defaults. Most people have one default mode (usually "write for my team" or "write for the public"). Noticing your defaults is half the work

Quick Reference

The Four Audiences

Yourself        notes, todos, journals, second-brain entries
Your team       memos, docs, PRs, tickets, design specs
The public      essays, posts, docs, marketing copy
AI              prompts, system messages, configs, instructions

Each audience rewards different moves. Writing well for all four is a small number of transferable skills plus some audience-specific polish.

The Clarity Checklist

Concrete        names, numbers, specifics; not vibes
Self-contained  a reader landing cold can follow
Ordered         the sequence matches how a reader will process
Minimal         no filler; every sentence earns its place
Visible         the structure shows (headings, lists, emphasis)

Five properties, all testable. Most bad work-writing fails on at least two.

The Common Failure Modes

Meeting-mind writing        stream of consciousness as if you're on a call
Hedge-piled writing         "it could potentially be that perhaps"
Jargon blankets             specifics hidden behind generic terms
Burying the lead            the conclusion appears on page 3
No reader model             written for yourself but labelled as a team doc

Most revision is correcting these.

Learning Path Suggestions

The overwhelmed knowledge worker (roughly 4 hours)

1. Chapters 01 and 02 for the frame
2. Chapter 03 on clarity
3. Chapter 05 for async-team writing
4. Chapter 12 for habits

The founder or leader writing for a team (roughly 5 hours)

1. Chapters 01 through 05 for the framework
2. Chapter 07 on structure
3. Chapter 09 on naming and framing (high-impact work)
4. Chapter 12 for habits

The AI-heavy user (roughly 3 hours)

1. Chapters 01 and 02 briefly
2. Chapter 03 on clarity (prompts reward it enormously)
3. Chapter 06 on writing for AI
4. Chapter 09 on naming (half of prompts are about framing)

Why This Matters

• Writing is the interface with modern tools. Your laptop takes text input; so does your editor, your terminal, your issue tracker, and your LLM. The quality of text you produce determines the quality of what you get back
• Async work runs on docs. Remote and hybrid teams coordinate through writing. The team member who writes well coordinates three times as many things as one who writes poorly
• AI rewards precision. Vague prompts produce vague output. Precise writing produces better AI work. Most people's AI frustration is unreviewed prompting
• Written output compounds. The notes, docs, and memos you write pile up into a searchable history of your thinking. Most people's don't because they were written poorly

Additional Resources

• On Writing Well by William Zinsser
• The Sense of Style by Steven Pinker
• Bird by Bird by Anne Lamott
• Writing Tools by Roy Peter Clark
• The Amazon "six-page memo" articles and essays
• Paul Graham's essays (for density and voice)
• Stripe Press: The Revolt of the Public, High Growth Handbook, and others (models of well-written business writing)
• Julian Shapiro on writing: https://www.julian.com/guide/write/intro
• The sibling tutorial at content/writing/ for literary craft

A Note on Tools

This tutorial barely mentions specific tools. Tools change; the underlying skills don't. Whether you write in Notion, Google Docs, plain Markdown files, a paper notebook, or an AI prompt box, the clarity, structure, and editing skills are the same. Chapter 10 covers tools briefly, as infrastructure.