In addition to my writing job, I teach dance fitness classes on the side, which is a riot. (I also get to tell my dad I dance on a stage for money. He’s very proud of me.) But an extra job means extra co-workers, and extra co-workers means the opportunity for more drama. Usually, everyone is wonderful, but recently an issue arose that got me pretty heated. I was ready to fire off an angry email to my boss, but as a card-carrying adult, I’ve learned you should never send a work email when angry or PMS’ing (and this checked both boxes). So I ate some frozen chocolate-covered raspberries, took three deep breaths and popped my email into ChatGPT.
“Could you please help me rewrite this to make it sound more professional and less angry? Also please take out all the curse words. Thank you.”
(I know that saying “please” and “thank you” to AI is apparently bad for the environment, but I can’t help it. Someday our AI overlords will remember who was polite and who wasn’t, and I intend to be on the right side of history.)
ChatGPT immediately produced a much better version: calm, polished, professional and completely expletive-free. Excellent. I copied and pasted it into an email and hit send. What I did not realize was that I had copied the entire ChatGPT reply—including its helpful little recap at the top, which summarized my original instructions: “I can see why you’re so upset with your boss! Here is a rewritten version of your email, edited to sound more professional, less angry and with the curse words removed.”
I found out when I received a one-word reply from my boss: “OK.”
She is not normally that curt. That was when I scrolled back up and discovered what I’d sent. I sank into my chair with embarrassment. In trying to hide my real feelings, I had managed to reveal them, look sneaky and seem lazy—all in one email! Incredible efficiency, really. We eventually talked it out in person, and everything is fine now, but it was a thorough education in how things can go sideways when you outsource your communication to a robot.
But stupid mistakes aside: Was it actually rude for me to use ChatGPT to write my email in the first place? As we enter this new world of AI-assisted everything, a lot of us are asking the same question. Read on for the correct etiquette in this tricky situation.
Get Reader’s Digest’s Read Up newsletter for more etiquette, humor, cleaning, travel, tech and fun facts all week long.
The case for using AI to write your emails
First things first: AI-assisted writing is not some shocking new invention. People have been using spell-check, Grammarly and “hey, can you proofread this?” friends for years. AI is just a much more powerful—and occasionally chaotic—version of the same human impulse to communicate better than our first draft allows.
It makes you more efficient. For high-volume professional correspondence, AI is a genuine time-saver. If you’re writing the same kind of email on repeat—following up on proposals, scheduling meetings, responding to the same FAQ for the 15th time this week—having AI help you draft or refine a template is just working smarter, not harder.
It can make you clearer. A lot of us struggle to organize our thoughts in writing. AI can help structure a complicated message so the most important information leads and the reader doesn’t have to wade through four paragraphs of backstory to find the actual ask. That’s not deceptive. That’s considerate.
It helps you navigate tricky situations. This was my use case! When you’re furious—or heartbroken or anxious or just generally unfit for human communication—AI can act as a diplomatic buffer between your feelings and someone else’s inbox. Wanting your frustration translated into something constructive before sending is actually quite mature of you. (Unlike, say, sending the AI’s stage directions alongside the email.)
It can rescue you from email-thread hell. You know the ones: a 47-message chain where you’ve been cc’ed since Tuesday and cannot identify what anyone actually wants from you. AI can read the whole thing, summarize the key points and flag what needs a response. This is less “being sneaky” and more “basic self-preservation.”
The case against using AI to write your emails
Not all emails are created equal—and not all AI use is equally defensible.
It’s not actually you. For professional emails, this barely matters. For personal ones, it matters a lot. If your condolence note was written by a language model, the recipient is not receiving your empathy; they’re receiving a statistically likely approximation of it. That’s a little bleak when you think about it for too long, which I recommend not doing.
You won’t remember what you wrote, and that can come back to bite you. Research backs this up: A 2021 study on cognitive offloading found that while outsourcing mental tasks boosts immediate performance, it actively interferes with memory formation regarding the content. In practice, this means your boss or your friend could reference a commitment you made in an AI-drafted email and you’ll have absolutely no memory of making it—because, well, you didn’t. ChatGPT did.
The tells are real and increasingly obvious. You know them when you see them: the word delve. Phrases like “it’s important to note“ and lots of dashes. (Which, in my defense and the defense of all the other magazine-trained writers, we’ve been using liberally for years. AI learned that from us, not the other way around!) Unprompted bullet points where a normal human would have just written two sentences. Heaven help me, all the random emojis inserted. And don’t forget the telltale closing: Warmly. AI has verbal tics that have become unmistakable, and savvier readers are already clocking them.
It can cause some really hurt feelings. Join me for this Valentine’s Day cautionary tale: One of my friends received the most beautiful Valentine’s Day card from her husband. Thoughtful. Poignant. There was even a little poem. She texted a screenshot to our whole group and gushed over how much time and care he had clearly put into it.
Spoiler: He had not put much time and care into it. He’d asked ChatGPT to write it, and the whole thing had taken him about 30 seconds to generate. The longest part was copying it out by hand. He was so impressed with himself that he told another friend’s husband, who mentioned it to his wife, who told my friend … who was then absolutely crushed.
The card hadn’t changed. The words were the same. But the meaning evaporated entirely because the meaning was never really in the words—it was in the effort behind them. That’s the part AI can’t fake, no matter how convincing the output. I think about this story whenever I’m tempted to have ChatGPT write anything that’s supposed to be emotional.
The gray areas (because there are always gray areas)
Like most things I cover in this column—from Venmo-requesting a friend for a $3 coffee to objecting when someone orders steak and three cocktails before cheerfully suggesting to split the check—the answer here is almost never clean. It’s almost always “it depends.”
It depends on the relationship. The closer and more personal the relationship, the more the recipient deserves your actual voice—typos, run-on sentences and all. Texts to your best friend, apologies, condolence notes, love letters: These should come from you. An imperfect, fumbling version of you is more meaningful than a polished bot.
It depends on what kind of email it is. Professional context, high stakes, need to be precise? AI is a reasonable assist, but the key is to treat it as a first draft, not a final product. Read it. Edit it. Make it sound like you again. And for the love of everything, do not copy and paste the part that says “here is your angry email with the swear words removed.”
It depends on what happens next. This is the practical trap most people miss: If you use AI to compose a thoughtful proposal, brainstorm solutions or lay out a plan, and then a co-worker asks you about it in a meeting, you’d better have actually thought about it. Using AI to sound smarter than you are in writing and then having to perform that intelligence in real time is a trapdoor waiting to open under you.
The etiquette is still evolving, which means the stakes are still shifting. A decade ago, using Grammarly felt vaguely embarrassing, like admitting you couldn’t write. Now it’s expected. Using autocorrect used to feel like cheating; now, turning it off feels reckless. AI writing assistance is in that same awkward in-between phase right now, where the technology is everywhere but the norms haven’t fully landed. We’re all just doing our best.
The verdict
Using AI to write your emails is not automatically rude, but it can very easily become rude, depending on how you use it and who you’re writing to.
For professional correspondence: Use it thoughtfully. Let it help you draft, organize and clarify, then make the result your own. Don’t let it write things you won’t remember agreeing to, and read the entire output before hitting send. The whole output. Including the top.
For personal correspondence: Put the robot away. The people who love you want to hear from you—messy, imperfect, over-explaining, occasionally sweary you. Real is what matters.
And for Valentine’s Day cards: Write it yourself. Every single time.
Have a social situation you can’t stop ruminating on? Email us at advice@tmbi.com or message Charlotte on Instagram at @CharlotteHiltonAndersen.
Why trust us
Reader’s Digest has published hundreds of etiquette stories that help readers navigate communication in a changing world. We regularly cover topics such as the best messages to send for any occasion, polite habits that aren’t as polite as they seem, email and texting etiquette, business etiquette, tipping etiquette, travel etiquette and more. We’re committed to producing high-quality content by writers with expertise and experience in their field in consultation with relevant, qualified experts. We rely on reputable primary sources, including government and professional organizations and academic institutions as well as our writers’ personal experiences where appropriate. We verify all facts and data, back them with credible sourcing and revisit them over time to ensure they remain accurate and up to date. Read more about our team, our contributors and our editorial policies.
Source:
- Quarterly Journal of Experimental Psychology: “Consequences of cognitive offloading: Boosting performance but diminishing memory”
The post Is It Really Rude to Have ChatGPT Write Your Emails? appeared first on Reader's Digest.
from Reader's Digest https://ift.tt/rXJga7v
Comments
Post a Comment