AI is Revolutionizing Proofreading.

This is the second of five editions focused on everyday uses of AI to assist with tasks commonly performed by Project Managers. Today, we’ll dive into one of the simplest and most effective applications: using AI as a proofreader for reports and presentations. We’ll walk through how to use it and highlight potential pitfalls to avoid.

This week, I used last week’s article as an example to test AI’s proofing. To my surprise, ChatGPT identified errors that I missed—even though I had used it to proof the article shortly before publishing. After making a few adjustments and introducing new errors, I didn't think I needed to recheck it with Dr. C. That was a mistake.

For this example, I copied the article from LinkedIn by clicking “Edit Article,” pressing Control + A to select all, and pasting the content into ChatGPT. Initially, Chat tried to rewrite my article entirely. Avoid that. While ChatGPT reliably identifies grammatical and typographical errors, its rewrites often introduce new issues—errors in tone, word choice, or meaning. These errors can be harder to spot and can mislead the reader. Instead, I ask ChatGPT: “Just show me what errors you would correct.” From there, I filter the suggestions. Typically, I incorporate 20-40% of its recommendations.

Here are the first nine issues ChatGPT flagged, rated with my feedback icons on the left:

• Green Check: Great catch, Dr. C.
• Yellow Question Mark: A little picky, but worth considering.
• Red X: Incorrect suggestion.

#6 was a big deal--I somehow duplicated a sentence in two contiguous paragraphs. Thank you Dr. C! On the other end of the spectrum, #9 says I should capitalize "Expect" but it already was capitalized. Not a big deal. You might wonder—if only two of the nine issues were serious errors, is the process worth it? Absolutely! Finding that repeated sentence invaluable; dismissing the others took seconds. And ChatGPT reliably catches almost all serious errors. The process took about a minute, compared to the 10 minutes I would spend manually proofreading this short article.

This Capability Scales

In the body of this article, we covered a simple example--a short article with one or two issues. But it scales to large work. It can proofread long documents and Power Point presentations in little more time than it takes to drag-and-drop the file into the web page. I used it to help me proof an 8000-word white paper--about 30 pages long including figures. That process took 20 or 30 minutes (most of it being me cutting-and-pasting the corrections I wanted). I know Chat GPT is famous for making errors, but I didn't see one grammatical or typographical error after the process was done. Chat even found text errors in bitmaps of screen images and made sure they aligned to the copy. In another example, I asked it to find errors in a 200-slide training deck--it found a few that had been there for more than a year. I never saw one it missed. These large documents take hours to proof...or rather, they used to. Now they take minutes.

ChatGPT has doubled my speed when writing and proofing is a large part of that. I used to think I was a pretty good writer—not a professional, but I’ve authored five books and have half a million words in print. When a colleague suggested using ChatGPT for writing, I was skeptical. My initial experiences weren’t great either—I let the tool make too many decisions, and it often rewrote my articles, adding as many new errors as it removed. Frustrated, I abandoned it for months.

Today, things are different. Between ChatGPT’s extraordinary proofreading and content generation capability (more on that in an upcoming edition), my productivity has skyrocketed. My words-per-hour output has doubled, and writing has become more enjoyable. ChatGPT takes care of the tedious parts, freeing me to focus on the creative aspects. I now consider it an indispensable writing companion.

One More Feature: Tone Adjustments

As Jasmine pointed out in the comments last week, you can also ask ChatGPT to adjust the tone of your copy. For fun, I asked ChatGPT to rework the first paragraph of last week's article in two contrasting tones: "Very formal" and "Make it Fun".

• Original (the tone I generally write in): Each Friday, we dive into “Friday is AI Day,” a weekly discussion exploring practical uses of AI in tasks commonly performed by Project Managers. These sessions aim to make AI approachable and immediately useful for your daily work. By focusing on actionable insights, we’ll help you get the most out of AI tools without the unnecessary hype.
• Very Formal: Each Friday, the initiative entitled "Friday is AI Day" convenes to conduct a structured and systematic discussion regarding the practical applications of artificial intelligence (AI) within the scope of tasks routinely undertaken by Project Managers. The primary objective of these sessions is to render AI both accessible and directly applicable to professional workflows. By prioritizing actionable insights, the discussions are designed to optimize the utility of AI tools while eschewing excessive promotion or unwarranted enthusiasm.
• Make It Fun: Every Friday, we kick back and dive into "Friday is AI Day," a laid-back, super-useful chat about how Project Managers can actually put AI to work. These sessions are all about making AI easy to get into and immediately helpful for your day-to-day. No boring tech-speak or over-the-top promises—just real, actionable tips to make your work life better.

The results are just amazing! Adjusting tone used to be one of my most tedious tasks as a writer; I dreaded it. It required countless iterations, changing one phrase at a time and re-reading paragraphs repeatedly. Chat takes just three seconds and produces a draft usually better than I could manage alone.

Let’s Keep the Conversation Going

If you found this article valuable, please give it a thumbs up ?? or leave a comment. Your engagement helps LinkedIn’s algorithm share this content with more people who can benefit. Also, if you’d like more tips on using AI in everyday tasks, check out Jordan Wilson’s blog, "Everyday AI."

. Thanks for being part of this journey!