
mcmentalonthemic
Bronze
- Joined
- Sep 23, 2023
- Posts
- 462
- Reputation
- 678
Beyond common sense, there are a few indicators that a thread is GPT made
ALSO, AI DETECTORS DON'T MEAN SHIT. Though they are good indicators.
What I mean is, if something is written with AI it'll 90% chance show up on AI detectors, but if something isn't written with AI it can still show up, like 20% tbh
Think of it like this:
All thumbs are fingers, not all fingers are thumbs
All AI written threads will show on AI detectors as AI written, not all threads that show on AI detectors as AI written are written with AI
1: BULLET POINTS
GPT fucking loves it's bullet points. Most computer and laptop keyboards and probably the one you're using right now don't have bullet point buttons. I highly doubt this nigga wrote all of this on his phone so I'm going to say it's AI. It's a dead giveaway a thread was made with AI.
99% of computer users will just use dashes
- like
- this
BONUS POINTS: If the bullet points goes to the right
It's dead giveaway, literally nobody does this except GPT which does this ALLLL the time
I will actually use this nigga's responses as an example cause the retard used GPT to try and prove me wrong
Proof: I was doing some GPTing because I wanted to compare Anterior Maxillary Distraction and BImax [personal research cause curious] and it did it.
It does it ALLL the time.
IF YOU'RE TALKING TO SOMEONE AND HE DOES THIS, 100% CHANCE HE COPY AND PASTED FROM AI, OR COPY AND PASTED FROM AI AND CHANGED THE TEXT TO SOUND MORE HUMAN. NOBODY DOES THIS WITH THEIR BULLET POINTS.
2: Rapidly changing typing pattern
Look at how this person talks, quite immature, like a normal .org user, nothing wrong.
Then, when I press him further, out of nowhere he starts talking like he's writing his highschool thesis
This is actually a really good case study. If reading this has made you think back to a time where something similiar happened to you with someone else, then they were most likely using AI
3: Formatting
I wrote this thread here as a joke with AI to make fun of another thread from a nigga claiming he made 1,400,000 dollars by 18
The way it's formatted is VERY AI-like. If you see a thread written like this it's 100% AI or originally written with AI as it's template
4: How to catch out threads written with "type this like you're a person writing on a forum" prompts
AI actually is able to make some pretty good puns and quips, which it probably will do with prompts like this. Look at the screenshot above, that is literally classic-AI speak when it comes to the jokes it's making.
Look at this next example
This guy copied my microneedling guide from screenshots of my drafts I posted to hype it up.
It doesn't matter since I view mogged in half the time but the biggest issue with it is how much it gets wrong, mainly because of AI.
AI is great with explaining general topics but if you get to specifics and I mean really specific, it usually gets it wrong [for example he got needle deepness completely wrong]
BUT THAT DOES NOT MATTER, LET'S LOOK AT THE WRITING STYLE.
Reading this 100% confirmed it for me that it was AI. I know I keep saying "classic ai" but this is literally classic AI speak. He 100% wrote his prompts with "write it like you're an intelligent forum user writing casually" or some shit


If your pattern recognition is high, you will start noticing alot more of these when reading AI threads.
An AI thread in the wild:
Hahaha this is actually such Lolfuel, the bullet points, the "let's start off with the basics" then "the basics" turning out to be the whole guide

@a4dm Be ashamed.
looksmax.org
btw literally nobody called it out for using AI in the replies jfllll 

ALSO, AI DETECTORS DON'T MEAN SHIT. Though they are good indicators.
What I mean is, if something is written with AI it'll 90% chance show up on AI detectors, but if something isn't written with AI it can still show up, like 20% tbh
Think of it like this:
All thumbs are fingers, not all fingers are thumbs
All AI written threads will show on AI detectors as AI written, not all threads that show on AI detectors as AI written are written with AI
1: BULLET POINTS

GPT fucking loves it's bullet points. Most computer and laptop keyboards and probably the one you're using right now don't have bullet point buttons. I highly doubt this nigga wrote all of this on his phone so I'm going to say it's AI. It's a dead giveaway a thread was made with AI.
99% of computer users will just use dashes
- like
- this
BONUS POINTS: If the bullet points goes to the right

It's dead giveaway, literally nobody does this except GPT which does this ALLLL the time
I will actually use this nigga's responses as an example cause the retard used GPT to try and prove me wrong
Proof: I was doing some GPTing because I wanted to compare Anterior Maxillary Distraction and BImax [personal research cause curious] and it did it.
It does it ALLL the time.

IF YOU'RE TALKING TO SOMEONE AND HE DOES THIS, 100% CHANCE HE COPY AND PASTED FROM AI, OR COPY AND PASTED FROM AI AND CHANGED THE TEXT TO SOUND MORE HUMAN. NOBODY DOES THIS WITH THEIR BULLET POINTS.
2: Rapidly changing typing pattern
Look at how this person talks, quite immature, like a normal .org user, nothing wrong.


Then, when I press him further, out of nowhere he starts talking like he's writing his highschool thesis

This is actually a really good case study. If reading this has made you think back to a time where something similiar happened to you with someone else, then they were most likely using AI
3: Formatting
I wrote this thread here as a joke with AI to make fun of another thread from a nigga claiming he made 1,400,000 dollars by 18

The way it's formatted is VERY AI-like. If you see a thread written like this it's 100% AI or originally written with AI as it's template
4: How to catch out threads written with "type this like you're a person writing on a forum" prompts

AI actually is able to make some pretty good puns and quips, which it probably will do with prompts like this. Look at the screenshot above, that is literally classic-AI speak when it comes to the jokes it's making.
Look at this next example

This guy copied my microneedling guide from screenshots of my drafts I posted to hype it up.
It doesn't matter since I view mogged in half the time but the biggest issue with it is how much it gets wrong, mainly because of AI.
AI is great with explaining general topics but if you get to specifics and I mean really specific, it usually gets it wrong [for example he got needle deepness completely wrong]
BUT THAT DOES NOT MATTER, LET'S LOOK AT THE WRITING STYLE.

Reading this 100% confirmed it for me that it was AI. I know I keep saying "classic ai" but this is literally classic AI speak. He 100% wrote his prompts with "write it like you're an intelligent forum user writing casually" or some shit
If your pattern recognition is high, you will start noticing alot more of these when reading AI threads.
An AI thread in the wild:

Hahaha this is actually such Lolfuel, the bullet points, the "let's start off with the basics" then "the basics" turning out to be the whole guide
@a4dm Be ashamed.

ONLY SKINCARE GUIDE U NEED TO ACHIEVE MODEL SKIN
Lets start off with the basics: avoid trash products full of alcohol, fragrance, essential oils, and harsh actives Dont eat processed trash and cut added sugars completely prioritize products with simple, non-comedogenic, and proven ingredients LESS is MORE don’t overload your skin always...
Last edited: