AIToday
Large Language ModelsSimon Willison's WeblogPublished: Jul 18, 2026, 04:00 JST2 min read

Tool detects LLM writing clichés in text

Key takeaway

  • A new online tool helps readers spot clichéd phrases commonly found in text written by language models.

  • The tool works directly in your browser and highlights ten common patterns, including repetitive chains like "no X, no Y" and expressions such as "sit with that." Users can toggle detection patterns and navigate through matches to quickly identify whether text has LLM fingerprints.

3 Key Points

  1. What happened

    A new browser-based tool has been created to identify and highlight common phrases that frequently appear in language model outputs, including patterns like "no X, no Y" chains, "sit with that," and "you already know."

  2. Why it matters

    LLM-generated writing often contains recognizable clichéd phrases that can make content feel formulaic and repetitive. The tool lets readers quickly spot these patterns in any text they paste, helping identify when content may have been written or heavily influenced by AI.

  3. What to watch

    The tool runs entirely in your browser with localStorage support, meaning no data is sent to external servers. Users can toggle pattern detection on and off and navigate through matches with quick navigation features.

Ask the AI about this article →

Context & Analysis

The creator's frustration with repetitive LLM-generated writing — particularly the prevalence of phrases like "no fluff, no filler, no jargon" — motivated the development of this detection tool. Language models tend to fall into recognizable stylistic patterns, and these clichés have become increasingly visible to readers as AI-generated content proliferates. By making these patterns visible and easy to spot, the tool addresses a practical problem: distinguishing between human writing and writing that bears the hallmarks of AI generation or heavy AI editing. The tool's design choice to run entirely client-side (in the browser) reflects a privacy-conscious approach, ensuring that text analyzed remains on the user's device.

FAQ

How do I use the tool?
Paste text into the analyzer to identify clichéd patterns. The tool highlights sentences containing matches and shows a count of each pattern found, with quick navigation between results.
Does my data get sent anywhere?
No — the tool runs entirely in your browser with localStorage support for persistence, so no data is sent to external servers.
Simon Willison's WeblogRead Original Article

Get the latest Large Language Models news every morning

For example, today's edition would include:

  • Visko raises $10M, launches live AI video model OrbisSiliconANGLE AI · 14m ago
  • Runway unveils Solaris, an AI that generates app interfaces in real timeTHE DECODER · 14m ago
  • Google AI Search flags Facebook users as dangerTHE DECODER · 14m ago

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleGraduate student builds NASA robot that assembles satellites without cameras