There are a few signs which becomes obvious once you’ve seen lots of LLM text. For example lots of usage of the long dash symbol, the word “actually” all the time. Overuse of the word “honest”. And more.
You can copy and paste their text into gptzero.me to get a report. Here is an example
Over on piefed.social I banned about 5 new accounts in the last week because their first post was LLM generated link spam.
It’s a real problem.
In PieFed there’s a setting that makes it so that a new account’s first post is held back for review by admins before it goes public. It cuts down on stuff like this.
Woah the PieFed dev :O But huh? Wdym
There are a few signs which becomes obvious once you’ve seen lots of LLM text. For example lots of usage of the long dash symbol, the word “actually” all the time. Overuse of the word “honest”. And more.
You can copy and paste their text into gptzero.me to get a report. Here is an example
Oh. I had no idea what you were referring to because this wasn’t under their post. My bad
I totally agree – Now that you say it, it looks very generated
Over on piefed.social I banned about 5 new accounts in the last week because their first post was LLM generated link spam.
It’s a real problem.
In PieFed there’s a setting that makes it so that a new account’s first post is held back for review by admins before it goes public. It cuts down on stuff like this.