Live data from Hacker News

Students using “humanizer” programs to beat accusations of cheating with AI

nbcnews.com

11–20 of 91 posts

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#11

Everyone knows about emdashes, but there are so much more! Here is a wiki article with all common tell-tales of AI writing: https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing

I used em dashes heavily 15 years ago when writing my PhD thesis.

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#13

Everyone knows about emdashes, but there are so much more! Here is a wiki article with all common tell-tales of AI writing: https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing

I used em dashes heavily 15 years ago when writing my PhD thesis.

Same, but 5 years ago. Now they're ruined for me lol.

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#15
post #3
post #2

I'm guessing this "humanizer" actually does two things: * grep to remove em dashes and emojis * re-run through another llm with a prompt to remove excessive sycophantry and invalid url citations

You’re absolutely right! Ha. Every time an AI passionately agrees with me, after I’ve given it criticism, I’m always 10x more skeptical of the quality of the work.

Why? The AI is just regurgitating tokens (including the sycophancy). Don't anthropomorphise it.

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#16
post #14

[flagged]

The fraud is all around, also when honest writers get falsely accused of using AI when they didn't, which is the point. It's why everyone needs humanizers. The risk of not using them is too high for everyone. As such, the fraud I see above all is the one in your comment.

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#17
This ship has sailed.

It's how it was with the internet. I grew up in the 90s, and teacher didn't know how to deal with the fact we no longer had to go through multiple books in the library to get the information they needed. We barely needed to write it.

Now nobody expects students to not use the internet. Same here: teachers must accept that AI can and will write papers and answer questions / do homework. How you test student must be reinvented.

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#18
post #3

Earlier quoted context omitted.

You’re absolutely right! Ha. Every time an AI passionately agrees with me, after I’ve given it criticism, I’m always 10x more skeptical of the quality of the work.

Why? The AI is just regurgitating tokens (including the sycophancy). Don't anthropomorphise it.

Because of the way regurgitation works. "You're absolutely right" primes the next tokens to treat whatever preceded that as gospel truth, leaving no room for critical approaches.

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#19

This ship has sailed. It's how it was with the internet. I grew up in the 90s, and teacher didn't know how to deal with the fact we no longer had to go through multiple books in the library to get the information they needed. We barely needed to write it. Now nobody expects students to not use the internet. Same here: teachers must accept that AI can and will write papers and answer questions / do homework. How you t…

I know a lot of teachers are reverting back to handwritten papers. People can generate it but at least you're doing something.

Re: Students using “humanizer” programs to beat accusations of cheating with AI

#20

Everyone knows about emdashes, but there are so much more! Here is a wiki article with all common tell-tales of AI writing: https://en.wikipedia.org/wiki/Wikipedia:Signs_of_AI_writing

I saw someone created a skill to weaponize that exact list to humanize the AI's output

There are no clear signs, at least for anyone who cares to hide them

Post reply on HN