The good news: Last week’s column about crazy driving generated an unusually high number of comments and compliments.
The other news: I did not write it. Nobody did. It was written by ChatGPT.
Even with a note on top telling readers there was something unusual about the column, its auto-authorship went unnoticed. Indeed, a number of people said it was some of my best writing. This is nice, maybe. Maybe not. I have no idea what to think, except to wonder if I should have left well enough alone and not done this.
Disturbing. Unsettling. Fascinating.
The column came to be through the good offices of my dear friend, Paula Kravitz, a strategic consultant who has worked on national and worldwide social projects.
Paula has extensive experience with ChatGPT, having used it for more than a year for everything — deep research projects, all her written work, and even personal decisions. Paula is an accomplished painter — I own two of her pieces — and she even asked ChatGPT to help her decide whether to work more in oils or acrylics.
“It sorted my thinking and weighed the options and made a recommendation,” she said. “It’s my partner in just about everything.”
As for the column, at my request, Paula spent less than a few minutes composing a 93-word prompt for ChatGPT. Herewith (complete with original typos, etc.):
“Write a newspaper opinion column musing about why people in the Bay Area of california are always driving to fast, going through stop signs, going through relights, and generally breaking laws to get somewhere faster. muse about why is everyone in such a rush, even risking their lives and others in the interest of consciously or unconsciously trying to get somewhere faster. explore the deeper philosophical underpinnings that might be driving these behaviors and explore how we can consider the higher possibilities of stopping these behaviors and how that might improve our lives.”
While writing the prompt took only a few minutes, ChatGPT produced last week’s column in about 10 seconds. Less time than it took me to type that last sentence. Muse, indeed.
I made only a few edits. The original version was a tad too long and had to be trimmed, a problem that happens to me all the time. And it had me driving around in the East Bay. I changed the first sentence to read: “I was driving through a quiet neighborhood the other day, and I saw something that’s become disturbingly common.”
Now, I have to say, I would not have written that sentence in that way. And god bless regular reader Mike Caggiano, who generously emailed that “the style was a bit different from your other columns.” But so what? It was not awful; it was just not me.
Recommended for you
Or was it?
Paula is all-in on AI. “I love it beyond belief. … It is so helpful it feels like a real person. The more you see yourself as a collaborator, the better. It’s a true thought partner,” she said.
But as even this example demonstrates, results are only as good as the original input. Had I asked AI to write something about bad driving, it would have been much less useful and much more generic. Paula’s assistance is a clear indication that, like any tool, AI must be used effectively.
Typically, it takes me two to three hours to write each column, but that does not include the time spent gathering information, doing interviews, touching base with sources and simply thinking about what I want to say and how I want to say it. I am a columnist, which leaves me free to comment, but at heart I think of myself as a reporter.
All of this raises interesting possibilities. Maybe ChatGPT should write all my columns. To paraphrase Groucho Marx, aside from the improvement, no one will notice the difference.
Maybe pride in authorship is an outmoded concept.
As regards to journalism, ChatGPT even could be seen as a throwback to another era. My grandfather was a reporter for the San Francisco Chronicle, covering the federal courts. He rarely wrote his own stories. As accurately depicted in the classic movie/play “The Front Page,” he called his information into the newsroom, where rewrite (mostly) men would translate it into news stories.
This is the same thing, I suppose, only faster, because we are eliminating a key human element.
Is that better? Beats the hell out of me.
BY THE WAY: Paula is among the dozens of artists participating in Open Studios from May 10-11. You can check out her work at paulakravitzstudio.com. None of her paintings were produced by AI. Not yet, anyway.
Mark Simon is a veteran journalist, whose career included 15 years as an executive at SamTrans and Caltrain. He co-hosts a podcast/videocast that can be found at TheGamePeninsula.com, and he can be reached at marksimon@smdailyjournal.com.

(3) comments
Thanks for your column today, Mr. Simon. I think we all noticed a change in the format/approach in your column from last week. I figured you took a media-free vacation and weren’t up to date on the latest shenanigans in the “As the Sheriff World Turns” soap opera but you still needed to submit a column. So you “mailed in” a column, perhaps written in a debate class from decades ago. I guess, in a way, I was correct. Regardless, it was a change of pace and was a welcome addition from the usual “orange man” bad columns we see on a regular basis.
Thanks, Mark, for a thought provoking column.
My daughter got me a Storyworth subscription a while back, and last month I finished 52 stories about my life. I did not consider using AI. One question asked about the television shows i watched as a kid. I'm pretty sure an AI composed response would have included "Howdy Doody" into its computer-generated narrative. However, I'm not so sure it would have mentioned, "Cousin Herb's Trading Post." I'm writing a different set of stories, but I'll keep drafting them... typos, broken syntax and all... without the help of AI.
I wrote a brief memoir about my bike being stolen in front of St. Timothy's church when I was 10, during choir practice. A friend put it through the masher called GPT and it invented more characters and turned it into a decidedly Christian morality tale, which was not my intention. Recently writing another memoir about a place where I had worked, I googled asking when the company had been bought up by a big corporation and AI replied that it had never happened. I was there and it happened. That piece of misinformation will go down as fact long after I'm gone. When I use Google search now, I search on the Web tab which eliminates AI. I recommend it.
Welcome to the discussion.
Log In
Keep the discussion civilized. Absolutely NO personal attacks or insults directed toward writers, nor others who make comments.
Keep it clean. Please avoid obscene, vulgar, lewd, racist or sexually-oriented language.
Don't threaten. Threats of harming another person will not be tolerated.
Be truthful. Don't knowingly lie about anyone or anything.
Be proactive. Use the 'Report' link on each comment to let us know of abusive posts.
PLEASE TURN OFF YOUR CAPS LOCK.
Anyone violating these rules will be issued a warning. After the warning, comment privileges can be revoked.