Chatbot words
Words chatbots overuse, where a plain word would do.
delvetapestrytestamentpivotalunderscoreintricatemeticulousvibrantlandscapegarneralign withfosterenhanceshowcaseleveragemultifacetedcommendablegame changer
Free tool
AI Slop Checker for posts and replies
Paste a post, a reply or an email. The checker marks the words, lines and sentence shapes readers now link to AI slop, and shows what to write instead.
What it found
The check runs in your browser. Twick doesn’t send or keep your text. It reads the writing only, and can’t tell who or what wrote it.
What it finds
Four kinds of AI slop it looks for.
Every line on the list is one readers already spot, from our articles on what AI slop is and the AI words and phrases everyone recognizes.
Sentence shapes
Sentence shapes that sound deep and say little.
It’s not just X, it’s YHere’s the thingServes asGreat questionI hope this helpsThree in a rowIs everything
Stock reply lines
Lines that could sit under any post.
Great insightsThanks for sharingCouldn’t agree moreCouldn’t have said it betterLove this perspectiveSpot onValuable perspectiveYour take on the future of AI
Em dashs
The mark readers now link to chatbots, where a period or a comma goes.
—
How to fix it
Five changes that take the slop out.
- Use the plain word. “Shows” for the fancy verb, “use” for the business one. One plain word reads like a person.
- Cut the opener and the sign off. Start with your point, and stop when you’ve made it.
- Say one thing only this post has. A number from it, a word the author used, or what’s in the picture. A reply that could sit under any post says nothing.
- Swap the em dash. A period or a comma does the same job, and readers no longer stop on it. More in why AI uses so many em dashes.
- Read it out loud. If you wouldn’t say it to the person, don’t post it.
A text with nothing on the list can still say nothing. The list catches the habits; only you can tell whether the reply adds something.
Replies with 100% no AI slop.
Twick drafts replies in your voice right under the post, about the post itself. You pick one and post it yourself.
FAQ
AI slop, checked.
What is an AI slop checker?
A tool that marks the words, lines and sentence shapes readers link to AI slop in a text, so you can rewrite them. This one runs in your browser and shows what to write instead next to each one.
Can it tell if a text was written by AI?
No. It reads the writing, not who wrote it. People use these words too, and a chatbot can write without them. Use it to clean up your own writing, not to judge someone else’s.
What does it look for?
Words chatbots overuse where a plain word would do, sentence shapes that sound deep and say little, stock reply lines that could sit under any post, and em dashes. Every one of them comes from our articles about AI slop.
My text has nothing on the list. Is it free of AI slop?
Not always. A reply without a single word from the list can say nothing about the post. Read it once more and make sure it mentions something only that post has.
Does it work for Threads, Bluesky, LinkedIn and email?
Yes. It reads any English text: posts, replies, emails, cover letters and articles.
Why is the em dash on the list?
Readers now link it to chatbots, which used it far more often than people do. A period or a comma does the same job. It has its own article: why AI uses so many em dashes.
Is my text sent anywhere?
No. The check runs in your browser, and Twick doesn’t send or keep your text. It’s free, with no sign up.