Skip to content

Author Voice

The hardest problem in AI-assisted writing isn't getting the AI to write. It's getting the AI to write like you. This section covers the 6-file voice system that made 81,000 words sound like one human wrote them -- and the techniques that kept voice consistent across 12 chapters written over weeks.

Contents

  • Building a Voice System -- The 6-file system: gold standard reference, author voice guide, blog-to-book adaptation, audience empathy profiles, authenticity markers, and quick reference checklist. How they compose into the master system prompt and why no single file does the job alone. Two-day investment that pays off by chapter 2.

  • The Gold Standard Method -- How to extract a benchmark from 400 words of your best writing. The line-by-line analysis process that produced 19 distinct techniques from a single blog post, plus the density test scorecard that converts "write well" into measurable targets.

  • Voice Drift Prevention -- Kill lists, de-AI prompts, periodic voice audits, and 10 encoded learnings from real writing mistakes. A layered defense against the gradual erosion of voice that happens across months of AI-assisted writing. Includes the scoring rubric (target: 85+) and the specific phrases that make readers think "an AI wrote this."

Key Takeaways

  • "Write in my voice" is a no-op. You need structured reference files that cover different dimensions of how you write, who you write for, and what you refuse to sound like.
  • The gold standard analysis is a one-time investment. The benchmark lasts the entire book.
  • Voice drift isn't a sudden failure -- it's entropy. Without active prevention, your voice erodes so gradually you don't notice until a full read-through reveals the damage.

Previous: Overview | Next: Prompt Engineering | Back to AI Writing Process