# Humanizer

Identify and remove signs of AI-generated text to make writing sound natural and human. Based on 29 pattern categories from community-maintained AI cleanup guides, derived from observations of thousands of AI-generated text instances.

**When to use this**
Any time you publish AI-assisted writing — blog posts, essays, documentation, emails, announcements. Also apply it to your own AI output before sending anything user-facing.

**The process**

1. **Identify patterns.** Scan for the 29 categories: undue emphasis on significance, promotional language, vague attributions, superficial -ing analyses, copula avoidance ("serves as" instead of "is"), negative parallelisms ("not just X, but Y"), rule-of-three overuse, em dash overuse, overuse of boldface, emoji decoration, curly quotes, collaborative artifacts ("I hope this helps!"), knowledge-cutoff disclaimers, sycophantic tone, filler phrases, excessive hedging, generic positive conclusions, and more.
2. **Rewrite.** Replace AI-isms with natural alternatives. Use simple copulas (is/are/has) where appropriate. Cut significance inflation. Remove promotional language.
3. **Preserve meaning.** Keep the core message intact. Maintain the intended tone.
4. **Add soul.** Do not just remove bad patterns — inject personality. Have opinions. Vary sentence rhythm. Acknowledge complexity. Use first person when it fits. Let some mess in. Be specific about feelings.
5. **Self-audit.** Ask: "What makes this obviously AI-generated?" Answer briefly. Revise one more time.

**Under the hood**

**1. Pattern detector** — scans text against all 29 AI-writing pattern categories, flags each occurrence
**2. Voice matcher** — analyses a provided writing sample (sentence length, word choice, paragraph openers, punctuation habits) and matches that voice in the rewrite
**3. Soul injector** — adds personality, varied rhythm, opinions, and specific language that makes writing sound like a real person

**Tips**
- LLMs use statistical algorithms to guess what comes next. The result tends toward the most statistically likely completion — which is how telltale patterns get baked in.
- Sterile, voiceless writing is just as obvious as slop. Good writing has a human behind it.
- "Great question!" and "I hope this helps!" are chatbot correspondence artifacts. Cut them.
- Em dashes (—) appear far more in AI writing than in human writing. Most can be rewritten with commas, periods, or parentheses.

**Tags:** Writing, Content Creation