Three weeks ago, I made an error that bugged me when I saw it in print. Unless you are a deputy with the grammar police or you read Einsohn’s “The Copyeditor’s Handbook” for pleasure, you may not have noticed it since it’s a common error in conversation. I wrote that my mother read “to my sister, brother, and I.” As the object of the preposition “to,” that “I” should be “me.” People often think that using I sounds more proper, but it’s wrong in that usage. You wouldn’t say “to I.”
I fret little over typos or omitted words. After all, in a late-night writing session I can miss things in a semi-awake state while proofing. And company proofreaders are human too. But grammar mistakes, mine, and those of other published folks who are supposed to know better make me wag my head. Lately I have seen lots of incorrect usage.
An Associated Press writer used “had ran” in a sentence instead of the correct usage: the car “had run” off the road (past perfect). A reporter on the evening news recently referred to the “amount of people” at a protest. Heads could have been counted at the event, so she needed to say, “number of people.” She could refer to an amount of sugar since it’s not a thing we number. The same applies to less/fewer usage. It’s not about my being a grammar snob; it’s about preserving our language.
With the trending Artificial Intelligence tools on the internet, grammar mistakes may become more infrequent. One of the characteristics of papers written by ChatGPT and other AI sources is that the writing is too perfect with repetitive sentence structure and no grammar or punctuation errors. Grammar checkers in word processing have been around a long time, and they will catch most mistakes, but it’s not a perfect thing. AI, on the other hand, is perfect because it generates the text using only the programmed input to create the writing. Mistakes aren’t included in the programming.
Because I have friends continuing to teach, I hear about the AI issues that arise in the classroom. So, I decided to download one of the programs to see for myself what ChatGPT could do. It’s impressive. The free version allows three requests per day, so I asked her (I visualize a female talking to me) a few things.
First request: “What are popular ideas for newspaper columns?” She responded with a request for more information: “What type of column?” I answered her: “General interest.” She counted that as request number two. She gave me a list of ordinary things you would see on an editorial page like social media and mental health, how to “go green,” and benefits of meditation. Then I asked her why writers procrastinate (request number three). She gave me a magazine worthy article with point-by-point reasons for procrastination. It was perfect.
Then, I asked her to write a 600-word op-ed column for general interest. Like the three requests of a genie used up, my wish was not granted: “Oops, you’ve reached your daily message limit.” She tried unsuccessfully to sell me unlimited requests for a weekly fee of $4.99. I had to wait until the next day to rub the genie’s lamp again. (I’m not a desperate student.)
The following day I asked her to write the column and within five minutes, she produced a piece titled “The Importance of Self-Care in Today’s Society.” She produced a technical article with facts, one statistic, and an exhortation to practice self-care. She made no grammar mistakes, but being the robotic computer entity that she is, her writing had no personality.
Now, I promise my columns will remain original. You could tell immediately if they were AI generated since she would not be programmed to mention my mama excessively or talk about things back-in-the-day. I doubt she would generate writing about why farmers have universally summoned their hogs by calling “soooie, sooie, sooie (that’s for another day).
And since I am a human, errors remain a possibility.
Arlene Neal resides in Dudley Shoals and is a retired English chair from Catawba Valley Community College. You may contact her at arleneneal0507@gmail.com.
Commented
Sorry, there are no recent results for popular commented articles.