← All AI writing use cases
Editing

Load an anti-AI writing file so drafts stop sounding like a chatbot

Tools:
Claude, ChatGPT
Who it's for

Founders whose drafts keep the tell-tale AI phrasing no matter how many 'don'ts' they add

What they report

No metric reported. Ruben Hassid's approach, shared and tested by Daniel Finchelstein, is a 29-word prompt plus a file of known AI writing patterns (built from Wikipedia's 'Signs of AI writing' page). The file shows the model what to avoid instead of a long list of 'don't' instructions.

Example

Hassid's red-flag prompt lines that he says don't work: 'Don't use jargon.' 'Don't sound like an AI.' 'Avoid passive voice.' 'Be conversational, not robotic.'

from source

How to do it

  1. Open Wikipedia's 'Signs of AI writing' page.
  2. Copy the whole page into a doc without editing. Save as anti-ai-writing.md.
  3. Upload it to your Claude project or chat.
  4. Use the 29-word prompt.
  5. Run final drafts through it again as an audit before you post.

The prompt

Fill in the [brackets]
Copy prompt

Read my anti-AI writing style file first. It contains every known pattern of AI writing I want to avoid. Apply these as rules to everything you write for me. (Audit mode, on a finished draft:) Check this draft against the file. Quote each line that matches a pattern, name the pattern, and suggest a plain rewrite. Change nothing else. [paste draft]

Prompt:
First paragraph verbatim from source; audit line by Legacy Builder
Source:
Ruben Hassid's method, as shared by Daniel Finchelstein on LinkedIn; Wikipedia, 'Signs of AI writing'
Read the original ↗
Legacy Builder Lab

Want help putting these to work every week?

Legacy Builder Lab is Clifton Sellers' community for founders who write their own content with AI. Members get his prompts and bot setups, five new post templates every Monday and a monthly live teardown.

Join the Lab waitlist
Run by Clifton Sellers, who also curates this library.