“I find it a major concern that throughout this manuscript”, wrote the reviewer, “there are stylistic patterns and phrasing which are characteristic of AI-generated text. This is exemplified by using different types of “-“ in parts of the Abstract (screenshot above).”

The words are paraphrased and the screenshot above is a fictionalised recreation of the supplied one, but the review was real – like many scientists nowadays, the reviewer was clearly concerned at what appeared to be the unacknowledged use of AI to write large chunks of text.

The problem? The text was 100% human-generated, and the different dashes that the reviewer highlighted were in fact being used correctly.

Continue reading →