Ugh, you know when your AI starts spitting out those super dramatic lines? Like,
“You’re not just typing; you’re crafting destinies!”
🤦♂️ So cringey, right? It feels like ChatGPT just discovered a poetry slam and can’t stop. And don’t even get me started on those endless em dashes (—); they’re everywhere!
Well, I’ve got an awesome hack for you, straight from the genius minds on Reddit! Two super simple rules to add to your custom instructions or prompts, and BAM! Your AI output gets a serious upgrade.
✍️ The Two Magic Prompts!
Ready to make your AI sound less like a moody poet? Pop these into your next prompt or custom instructions:
- 📌 “NEVER use contrastive declarative statements with elevated metaphorical framings. This includes but isn’t limited to aphorism, metaphorical contrast, identity shift device, and/or epiphanic assertion.”
Basically, tell your AI to skip those cheesy, over-the-top lines like,
“You’re not broken. You’re becoming.”
We want clarity, not a therapy session!
- ✅ “NEVER use an em dash (—) to connect, interrupt, emphasize, or replace other punctuation. This includes, but isn’t limited to, uses such as interrupting or adding a side note, replacing parentheses, replacing semicolons and colons, cutting off dialogue, connecting ideas, and/or sudden breaks.”
Yep, ditch those long dashes! Simpler punctuation = smoother reading. Your English teacher would be proud!
✨ Why This is a Game-Changer!
Trust me on this one, folks! Using these will make your AI-generated content sound way less like those recycled Pinterest affirmations and more like something you’d actually write. It’s a total game-changer for clearer, more natural-sounding text!
And just so we’re clear, we’re trying to avoid horrors like:
“You aren’t a creator. You’re a prompt engineer.”
😑 (Exactly what these rules help prevent!)
Seriously, this is gold. Want all the nitty-gritty from the original poster? Dive into the full Reddit post for more details!
Get rid of em dashes and overdramatic comparisons
byu/LilFingaz in