ChatGPT has become a prevalent tool for drafting emails, documents, and even blog posts. However, its distinct style sometimes makes the content easily identifiable as machine-generated.

Here are some tips on how to recognize ChatGPT’s writing and how to make your use of AI in writing less uncanny.

Interestingly, the usage of “delve” in research papers has increased significantly since 2022, suggesting ChatGPT’s influence extends beyond casual content creation.

Identifying ChatGPT Text:

  1. Unnecessary Conclusions: ChatGPT tends to wrap up paragraphs with obvious or redundant conclusions.
  2. Long Bullet Point Lists: Lists in ChatGPT outputs can often extend to 8-10 items, more than the typical human-generated list.
  3. Overuse of Adjectives: Phrases like “greatly reduced” instead of “reduced” are common, adding unnecessary flair.
  4. Frequent Use of Certain Words/Phrases: Words like “streamline”, “tapestry”, and especially “delve” are overly prevalent in texts written by ChatGPT.

How to Avoid Sounding Like ChatGPT:

  • Use the Memory Feature in ChatGPT Plus: This allows you to customize the AI to avoid phrases that are telltale signs of AI writing.
  • Personalize Your AI: Train ChatGPT or other AI tools like Claude on samples of your writing style. This tailors the AI’s outputs to more closely mimic your unique voice.

Takeaways

  • Be aware of unnecessary conclusions and overly long bullet lists in AI-generated text.
  • Watch for excessive use of adjectives and specific repeat phrases, which can signal AI involvement.
  • Utilize advanced features of AI tools to customize and minimize AI-like phrasing in your writing.
  • Consider training your AI tool on your personal writing style to enhance authenticity.