Feedback Ban AI-Generated text

[poughstings]

If this website is willing to ban pre-written automated messages, then the same should apply to text generated by theft veiled by linear algebra.

LLMs provide no benefit to users, expedites spam, and causes false information to be shared. It adds nothing to threads when users just summerizes or replies to a post with AI, because nobody asked for a bot's opinion.

 

[SylverReZ]

Yes, exactly my thoughts. AI could be generating false information, and without the proper context and resources lead by a human, this leads to places like GBAtemp look bad.

 

[Dionicio3]

How do you expect the site to do that? There's no big fancy sign in LLM posts that say "THIS POST IS AI", they're meant to appear as human as possible. You can't reasonably assume a piece of text is AI or not nowadays

 

[poughstings]

How do you expect the site to do that? There's no big fancy sign in LLM posts that say "THIS POST IS AI", they're meant to appear as human as possible. You can't reasonably assume a piece of text is AI or not nowadays

They expose themselves. There are multiple examples of users just fucking saying "HWOBAIUSBDI LOOK AT WHAT CHATGPT SAID!!!!111"

and yes, it's incredibly obvious to discern between low quality LLM text from real humans.

 

[Dionicio3]

They expose themselves. There are multiple examples of users just fucking saying "HWOBAIUSBDI LOOK AT WHAT CHATGPT SAID!!!!111"

and yes, it's incredibly obvious to discern between low quality LLM text from real humans.

the first example should obviously be banned, and obviously low quality models are easy to tell. But people are moving to higher quality models that are genuinely kinda hard to tell

 

[cearp]

Is this only going to apply to new members?
All of my posts have been generated by AI for nearly 20 years, and I'd like to continue.

 

[a_username_that_is_cool]

Is this only going to apply to new members?
All of my posts have been generated by AI for nearly 20 years, and I'd like to continue.

lmao

 

[designgears]

If this website is willing to ban pre-written automated messages, then the same should apply to text written by humans who merely parrot existing information or offer no original insight.

Certain human contributions provide no benefit to users, expedite spam, and cause false information to be shared. It adds nothing to threads when users just summarize or reply to a post with readily available information, because nobody asked for a repetitive human's opinion.

 

[poughstings]

If this website is willing to ban pre-written automated messages, then the same should apply to text written by humans who merely parrot existing information or offer no original insight.

Certain human contributions provide no benefit to users, expedite spam, and cause false information to be shared. It adds nothing to threads when users just summarize or reply to a post with readily available information, because nobody asked for a repetitive human's opinion.

Robots aren't human buddy. Talk about parroting.

 

[designgears]

Robots aren't human buddy.

I agree with you, but humans deciding if something is AI generated or not is going to be a hilarious drama fest.

 

[Latiodile]

There's no big fancy sign in LLM posts that say "THIS POST IS AI"

usually an excessive use of em dashes is a pretty major sign

 

[Dionicio3]

usually an excessive use of em dashes is a pretty major sign

which is honestly unfortunate—em dashes are awesome

 

[The Real Jdbye]

Agreed. I've seen too many users who all they do is post AI responses and it's so obvious. AI can be helpful, but these guys aren't trying to be helpful, they're just trying to inflate their post count in the laziest way possible.

Post automatically merged: Jun 6, 2025

How do you expect the site to do that? There's no big fancy sign in LLM posts that say "THIS POST IS AI", they're meant to appear as human as possible. You can't reasonably assume a piece of text is AI or not nowadays

LLM detectors are a thing, though they aren't perfect.
In the end it's probably not something that can be fully automated, and would put extra load on the staff to identify and deal with users who excessively post AI spam.

But one can wish.

 

[SecureBoot]

which is honestly unfortunate—em dashes are awesome

Yeah- you can rip my em dashes from my cold, dead hands

 

[Deleted member 754542]

I like the idea but don't think it's possible to enforce. Better just to put some sort of publicly viewable false information flag on posts which are deemed to likely contain false information, regardless of whether the content is AI generated.

 

[TomSwitch]

If this website is willing to ban pre-written automated messages, then the same should apply to text generated by theft veiled by linear algebra.

LLMs provide no benefit to users, expedites spam, and causes false information to be shared. It adds nothing to threads when users just summerizes or replies to a post with AI, because nobody asked for a bot's opinion.

How we know you are not AI?

 

[K3Nv4]

I'm disagreeing just because of this quote.

How we know you are not AI?

Despite it was trying to be humorous we don't know if that's how someone actually replies or possibly Ai generated text. Spam bots are easier to detect over its predictable style of posts. It's not that far fetched to believe some users are actually intelligent with how they respond. It's best to just stick with the trident of true if you see someone spamming just hit report and let the mods deal with it. If they start banning everything Ai they might as well ban the entire internet because online is already under the watch of Ai.

 

[Ondrashek06]

Images generated by LLMs should probably be banned too. 99% of what genAI produces is just slop, and I'm probably being generous here.

 

[kijetesantakalu042]

How are we going to do that without hurting real human's posts in the way?

 

[nullplayer00]

I did a test with AI generated texts. And from ~40 people no one could figure out where is the AI and where a human being