Following up on our post last week on Tips on Detecting & Improving AI-generated Text, I discussed a new study that suggests one of the “tells” of AI-generated text is the use of excess, sophisticated-sounding words. Inc discusses that study in more detail, identifying these 7 Words That Suggest a Text Was Written With AI.
-
Delves
-
Showcasing
-
Underscores
-
Comprehensive
-
Crucial
-
Intricate
-
Pivotal
As suggested last week, you, the human author, can and should provide specific prompts and actively refine AI-generated content to maintain your unique voice. I suggest that one of your prompts should be to tell the AI to limit the use of these seven “showoff” words.
This tracks with a prediction from Ars Technica’s Kyle Orland from the Inc. article:
“As knowledge of LLMs’ [large language models like ChatGPT] telltale marker words starts to spread, human editors may get better at taking those words out of generated text before it’s shared with the world.”
Hat tip to Steve Miller for the Inc. article.