discuss Should AI generated content be allowed on NamePros?

[Chieff]

Namepros allows AI generate / written content as replies since there is no rules against AI content?

 

[Paul]

Namepros allows AI generate / written content as replies since there is no rules against AI content?

Most content that current AI would generate—that’s content from a LLM, like ChatGPT—would be (1) very obvious and (2) probably a violation of our rules.

LLMs aren’t autonomous: someone has to give them very clear instructions. They can’t post on their own—that would violate our rules and would be easy to detect, and it also has nothing to do with LLMs.

As an example of something that might be permissible, it might be acceptable to use an LLM to verify the grammar and wording in a post you’ve already written. Asking it to rewrite the post would be a bit of a gray area.

LLMs can’t come up with new ideas. Someone has to feed them the ideas. They’re not a threat to our community—for now.

 

[Chieff]

Most content that current AI would generate—that’s content from a LLM, like ChatGPT—would be (1) very obvious and (2) probably a violation of our rules.

LLMs aren’t autonomous: someone has to give them very clear instructions. They can’t post on their own—that would violate our rules and would be easy to detect, and it also has nothing to do with LLMs.

As an example of something that might be permissible, it might be acceptable to use an LLM to verify the grammar and wording in a post you’ve already written. Asking it to rewrite the post would be a bit of a gray area.

LLMs can’t come up with new ideas. Someone has to feed them the ideas. They’re not a threat to our community—for now.

I came across these two posts

https://www.namepros.com/threads/poll-what-are-domains-most-valued-as.1300108/post-9003443
https://www.namepros.com/threads/poll-what-are-domains-most-valued-as.1300108/post-9003448

Reported one and it got rejected

There is contradiction between first two lines and last line

 

[Paul]

Reported one and it got rejected

What was the rejection reason? Those just look like spam; they probably could be deleted on those grounds alone.

It’s not clear whether those are AI-generated. We’ve been receiving very similar spam for many years; usually it’s just generic text copy-pasted from other sources. Either way, spam is spam, regardless of where it came from.

 

[Echo Mod Team]

I came across these two posts

https://www.namepros.com/threads/poll-what-are-domains-most-valued-as.1300108/post-9003443
https://www.namepros.com/threads/poll-what-are-domains-most-valued-as.1300108/post-9003448

Reported one and it got rejected

There is contradiction between first two lines and last line

The following (Potentially) AI generated comments do not have promotional reference, no 3rd party linking, and there is not a source easily found to prove the content was duplicated/copied. The posts also appear to be on-topic and explaining what variables domains are most valued as.

1. https://www.namepros.com/threads/poll-what-are-domains-most-valued-as.1300108/post-9003443
2. https://www.namepros.com/threads/poll-what-are-domains-most-valued-as.1300108/#post-9003448

While there is a high likelihood that the content was spun from other sources, the content is on-topic, helpful (for those that didn't know the information they provided already), does not promote a 3rd party site or service, and does not have an easy way to prove the source of where said content came from.

Because of the above, Echo Mod Team see's no reason to focus moderation efforts on such content, which technically does not violate a rule at this time.

We hope that information helps clarify why we did not remove the content.

 

[Paul]

@Echo Mod Team deals with spam a lot more than I do, so they're in a better position to judge the quality of the content. At least for now, we'll leave those posts live. That being said, we're always adapting to new threats; if we end up seeing a large influx of nearly-identical content like that, we may have to start removing it, regardless of whether it originated from a human or LLM.

 

[kor]

When will you take an action against AI generated nonsense?

Are you waiting for the entire forum to be filled with AI bots talking to each other?

 

[Bravo Mod Team]

When will you take an action against AI generated nonsense?

We’ve deleted posts and restricted several accounts for using it to sneakily spam.

Do you have any examples of ones we’ve missed?

 

[kor]

Do you have any examples of ones we’ve missed?

I've reported this one 2 times but it didn't got deleted because it's somehow considered as "uniquely crafted content" :

https://www.namepros.com/threads/paid-alternatives-to-expireddomains.1347541/#post-9354908

In fact, it's just the summary of the previous posts generated by AI.

I come across many examples like this while browsing the forum and it's pretty annoying. I'm here for valuable information created by real human interaction. Otherwise I would simply go and chat with LLMs all day long.

 

[dncafe]

I've reported this one 2 times but it didn't got deleted because it's somehow considered as "uniquely crafted content" :

https://www.namepros.com/threads/paid-alternatives-to-expireddomains.1347541/#post-9354908

In fact, it's just the summary of the previous posts generated by AI.

I come across many examples like this while browsing the forum and it's pretty annoying. I'm here for valuable information created by real human interaction. Otherwise I would simply go and chat with LLMs all day long.

I don't see anything in that post or the OP that makes it look like AI spam or marketing

 

[kor]

I don't see anything in that post or the OP that makes it look like AI spam or marketing

I didn't say it's spam, from technical point it's not. But still a nonsense post because it doesn't say anything different than the posts that came before it. It doesn't contribute anything to the topic and it wastes space & time.

If proper precautions are not taken against this type of auto-generated absurd content, I believe that forums and similar platforms will soon turn into a dump.

 

[cooljub]

Do you have any examples of ones we’ve missed?

Posts by most that I am ignoring.

 

[Bravo Mod Team]

If proper precautions are not taken against this type of auto-generated absurd content, I believe that forums and similar platforms will soon turn into a dump.

It's definitely something that we are monitoring and need to be mindful about to avoid that from happening. We appreciate feedback like this to help us keep tabs on it.

Currently, we're treating it the same as content by a human:

• The content must be on topic.
• The member must not be posting the same things repeatedly in the same thread (noncontributory).
• It must not be promotional outside of the marketplace.
• Etc.

At the moment, that particular post abides by all of those expectations.

Yes, it is not that contributory, but in all honesty, it's more contributory than several non-AI generated posts, so it wouldn't make sense to be more strict with better quality content just because it was generated by AI.

It's also most likely being posted by a human who is using AI.

 

[Bravo Mod Team]

Posts by most that I am ignoring.

Please feel free to share your ignoring list with us publicly or privately. We'd be happy to take a look into those accounts.

 

[D Haynes]

There's a place for "AI" generated content but I don't think forums are it. These are places for people (humans) to converse and share with each other.

 

[dncafe]

I've been running message boards since the early 2000's and have seen all kind of bot scripts. Long before there was anything like the AI of today, there were scripts that would: register new account, solve captchas, validate email, then mass reply to messages on a slow increment to make it look natural. The script would go into relevant Questions type thread, take the original message and search Google, pull one of the top search results in another message board, run it through an article spinner (basically a thesaurus) and repost it. It's more complicated than that, but thats just a generalization to give you an idea. Over the course of months you have an aged an active account.

I haven't seen much AI or spam account on NamePros as a whole though.

I will say there have been what I believed to be brigading by groups on here from time to time to push domain trends. About 10 years ago there was a big influx in accounts pushing CHIPs domains (chinese premiums), which centered around 4 letter .com and orgs with certain letters. I believe that was an orchestrated movement to manufacture a fake demand for CHIP domains. I suspect it was a group in China that held a large number of these domains

 

[Future Sensors]

It's also most likely being posted by a human who is using AI.

This is an important point. Similar to 'simple' translation tools, AI can assist members in expressing themselves more effectively on the forum. However, AI-generated output should never be adopted uncritically, and must still retain a 'human touch'. It should not be used for spamming the forum.

 

[cooljub]

Please feel free to share your ignoring list with us publicly or privately. We'd be happy to take a look into those accounts.

abdullahhhh01013
alberhomy
AmbitionHost
ananittarhib
Apkbee info
Ayesharain
bestadvocateinjaipur
BetterJoy
chandershekhardas
charukapoorimmigrati
CoffeeMF
cyberandbeyond
digitalleonapro
Digiworldgolu
Domainesia
dukar
Dyclem
Eric Morgan
etibassbenson65
gnews.domains
HekateSwitch
hugohill
instavipbio
jacksonshirley460
Jamboree
jennyotalia23
jtmac
karuneshbhandari
kubernetesdashboard
liamjames
matthew404
miabaker
mohdkaif002
mrbasheer
murchisonhershelg457
nicenic
OsuiniAbasi
rita
rsorgansiation
rumonrakin
Sarfraz S.
themelooks
timdavid010
Time4servers-tech
Tom_Holland
Vivetool
WebNIC.cc
William Archer
wootingdoublemovemen
yolandasmith
YOLOv8

 

[biggie]

I've reported this one 2 times but it didn't got deleted because it's somehow considered as "uniquely crafted content" :

Hi

when i read that post, my spider sense tingled.
so, i feel same as you.

imo...

Yes, it is not that contributory, but in all honesty, it's more contributory than several non-AI generated posts, so it wouldn't make sense to be more strict with better quality content just because it was generated by AI.

Hi

the point is not about quality of content, it's about repetitive content being posted by AI.

if a human member objects to an AI posting, then take it seriously or humans may start to disappear.

trust me, you don't want or need the cognitive dissonance.


imo....

 

[Bravo Mod Team]

the point is not about quality of content, it's about repetitive content being posted by AI.

What's repetitive about their post?

The topic of that thread is "Paid alternatives to ExpiredDomains." What makes their post less contributory to that topic than your post in that thread:

i've written names on toilet paper, as ideas come to mind while taking a

Personally, I enjoy your posts.

Our point is that it doesn't make sense to hold AI-generated content to a different standard than we hold humans.

That may change in the future.