I was asked to review an article before it was submitted for publication, and it had many problems. First and foremost, it was more an ad than a “thought leadership” piece.
While it is easy to reject the article, it’s harder to explain to the author what makes “ad-like.”
I asked ChatGPT to
- Create criteria for evaluating thought leadership articles.
- Use those criteria to evaluate the article.
- Improve and re-evaluate the article.
At that point, ChatGPT said that the most significant remaining problem was the lack of evidence in the article, so I told it to make up evidence (which should be marked in italics) to make it the best possible article.
The resulting article (in addition to being filled with lies) wasn’t nearly good enough for me to be willing to publish it, but it was exactly what I needed.
I sent the thread to the original author as guidance for writing a better article. This took me only a few minutes, and I expect he is closer to being able to write a publishable article of his own than he would had I spent an entire afternoon with him.
I also expect that he is further ahead than he would have been had he spent an entire afternoon with ChatGPT.
I don’t know which is more surprising to me.
Either way, I expect him to be a better author or at least write a better article by the end of the week. And a better ChatGPT user.
More importantly, I expect him to understand the nature of a thought leadership campaign in a way that no amount of editing or ghostwriting could ever have enabled.