The Inner Monologue

Oh, fantastic. You asked ChatGPT to help you sound cool in an email to your boss, and instead, it spat out something that reads like Jane Austen fanfiction. “Perchance, let us delve into the quarterly report with unwavering diligence.” Who talks like that? A 19th-century barrister? A particularly pompous owl?

Let’s break down why AI-generated text sounds like it was written by someone who still says “indubitably” in casual conversation—and how to stop it from embarrassing you.

Your AI Is Secretly a Time Traveler (And Not in a Fun Way)

ChatGPT didn’t learn to write from TikTok, Twitter, or even your mildly unhinged group chats. No, it was trained on:

• Public domain books (aka stuff your great-great-grandma thought was a riot)
• Academic papers (because nothing says “let’s keep this casual” like “furthermore, it is imperative to consider…”)
• Corporate jargon hellscapes (somewhere, a middle manager just whispered “synergy” and the AI perked up like a dog hearing a treat bag)

So when you ask it to “sound chill,” it panics and defaults to the linguistic equivalent of a monocle popping out of its face.

AI Doesn’t Choose Words—It Just Throws Darts at a Thesaurus

Humans pick words based on vibes. AI picks them based on math. That’s why you get:

• “Let us delve into the matter” (instead of “let’s talk about it”)
• “A testament to his unwavering resolve“ (instead of “he really stuck with it”)
• “She padded across the room“ (who pads? Cats. Ghosts. Serial killers.)

It’s not trying to sound dramatic—it just doesn’t know that normal humans stopped saying “alas!” outside of Renaissance fairs.

Other Crimes Against Language Your AI Commits Daily

• Overusing transitions like “Moreover” (Congratulations, your Slack message now reads like a UN treaty.)
• “Junk phrases” that sound smart but mean nothing (“In today’s rapidly evolving landscape…” Shut. Up.)
• Repetition (If I see “It is important to note” one more time, I’m throwing my laptop into the sun.)

The result? Text that’s technically correct but feels like it was written by a robot who majored in Pretentious Bullshit.

How to Fix Your AI’s Chronic Pretension

1. Bully it into submission

• “Write this like a normal human, not a Dickens character.”
• *”Sound like a person who has *never* said ‘per my last email’ unironically.”*

1. Edit the nonsense out

• Swap “delve” for “talk about”
• Replace “a matter of great importance” with “this thing”

1. Give it better role models

• “Write like a Twitter thread, not a textbook.”
• “Sound like a friend explaining this over beers.”

Final Thought: AI Needs a Personality Transplant

Until AI develops the ability to understand sarcasm, humor, or basic human interaction, it’s doomed to sound like a butler who’s very disappointed in you. So for now, use it as a first draft—then scrub out the nonsense with your actual human brain.

Or just accept that your emails will forever sound like they were written by a haunted quill pen. Alas!