Does Quillbot actually help avoid AI detection?

[JackBlow]

Alright, let's talk about the elephant in the room.

We all use AI sometimes. Don't lie. Whether it's for brainstorming, outlining, or... okay fine, for writing whole paragraphs. We've all been there.

But lately my uni has been cracking down hard with Turnitin's AI detector. Two kids in my study group got flagged last week and had to go through the whole academic integrity meeting. Scary stuff.

So I've been experimenting with Quillbot to paraphrase AI-generated text. The theory is: if you run ChatGPT through Quillbot enough times, it becomes "human" enough to fool the detectors.

Here's my test:

1. Wrote a paragraph about climate change policy using ChatGPT.
2. Ran it through Quillbot on "Fluency" mode.
3. Ran THAT through Quillbot on "Creative" mode.
4. Ran THAT through Quillbot again on "Formal" mode (lol).

I put the final result into GPTZero (the free detector) and... it came back as 98% human!

But here's the thing: the text was barely readable at that point. It was like a robot having a stroke. Words were technically correct but the flow was completely wrecked. No professor would believe I wrote that garbage.

So my questions:

1. Does Quillbot actually work for avoiding detection, or do detectors see right through it?
2. Is there a "sweet spot" where the text is still readable but not obviously AI?
3. Has anyone actually been caught using Quillbot-ed AI text?

I'm not trying to cheat on everything, but sometimes I just need a little help you know? Like, I have ideas but putting them into "academic" words is hard. ADHD gang where you at??

Drop your experiences below. Also if you know any tools BETTER than Quillbot for this, I'm all ears.

(pls don't judge me, we're all struggling out here)