What should NamePros add?

Good morning!

NamePros is working on a lot of new and exciting features, and many of them are thanks to member suggestions in our Comments and Feedback section.

We'd love to hear even more! What else would you like us to add for you?

All suggestions are welcome, from small and simple changes to big and complex features!

Please post below, or submit your suggestions privately.

Let us know what you'd like us to build!

 

[kor]

How would you expect that to work? Do you have examples of LLM-generated content on NamePros?

I have reported a few users before and you have not taken any action against it.

We also discussed this issue over PM and you were completely indifferent. You can see it by looking at the message history.

Simply ban users who post AI-generated content.

 

[MasterOfMyDomains]

How about an adult bickering room, no holds barred.
Namepros needs some quality control
48 hours and before all noobs posts must be moderated. Sorry make your job harder @Paul but dingleberries like this are really hurting integrity.

contact me here to discuss about that : [email protected]

^ https://www.namepros.com/threads/drivingstunt-com-aiprintprofit-com-appraisal.1305999/
Pure garbage. And get rid of that invisible crap.
Bring back our chat room too
Heavy hitters flying under radar
Still laughing about that a year later
Quality>quantity

 

[Paul]

^ https://www.namepros.com/threads/drivingstunt-com-aiprintprofit-com-appraisal.1305999/
Pure garbage. And get rid of that invisible crap.

That's not LLM-generated. Please don't conflate LLMs with the spam we always receive--it complicates discussion of the matter quite a bit. Traditional spam is easy enough to block with a low false positive rate; LLM-generated content is not.

Since the content has since been deleted by moderators, here's what it used to say, with the email address redacted:

contact me here to discuss about that : [email protected]

 

[Paul]

We also discussed this issue over PM and you were completely indifferent.

Me, specifically? I don't see a DM between us. I'm not indifferent toward it, but I'm also not sure how people expect us to block LLM-generated content.

Simply ban users who post AI-generated content.

Please provide a link to something you consider to be LLM-generated. I have a strong suspicion people are getting confused over what's AI-generated versus what's just low-quality content from humans.

 

[Paul]

I'm not a tech bod but I read there are such things.

https://www.namepros.com/threads/th...ing-a-supportive-network-on-namepros.1304038/
https://www.namepros.com/threads/ma...ing-a-domain-for-your-gaming-website.1306206/

Those just look like the usual spam we receive. They might've been generated with the help of AI, but not the modern AI that's been in the news lately (e.g., ChatGPT). ChatGPT doesn't usually make formatting issues or use awkward wording like we see on those posts--it sounds more human, oddly enough.

I'm going to leave them up as examples so other people can follow the discussion, but I consider those to be spam. They're probably not from ChatGPT; they look like the same spam we've been deleting for years.

 

[Paul]

I was fairly sure I had commented on this exact sort of issue publicly, but it looks like it was in a DM.

We had an issue a few months back where a reputable member of the domaining community wrote something that several people insisted was AI-generated. It wasn't--it was written by a human--but that accusation is scary, and it hurts people. It really disrupted the public discourse that would've normally taken place.

Here's what I sent to that member at the time (pardon the typos):

It's difficult to say precisely what impact generative AI will have in the long term. It's often ascribed a level of ingenuity bordering on the supernatural, while simultaneousy being treated as a mysterious black box that its detractors fear. Of course, none of that is true. It is, at its best, a precisely tuned auto-complete engine.

But excitement and fear are never assuaged by reality. What you're observing here is a harbinger of social challenges that we're likely to experience over the coming decades. We'll grow increasingly suspcious of content that could potentially have been generated by AI, and we'll be able to dismiss anything we don't like as being fabricated by a computer. That's a bit of a wrench to throw into the machine of public discourse, and I don't expect humanity to fully adapt to that during my lifetime.

I'd like to say you shouldn't be offended by what happened, or to promise that it won't happen again. I would, of course, be lying.

We don't have a way to detect LLM-generated content. As the paper by @Lox mentions, we can, at best, get in the mid-90%'s for accuracy. Can you imagine if we incorrectly flagged 1 in 20 posts as being AI-generated? How would you react if one day we deleted 5% of your posts, claiming they were generated by ChatGPT?

 

[kor]

Me, specifically? I don't see a DM between us. I'm not indifferent toward it, but I'm also not sure how people expect us to block LLM-generated content.

No, not with you. I corresponded with Echo Mod Team. I've reported a few AI posts I've come across before, but each time they said "no action need to be taken".

The argument they argued in the PM:
There's technically no rule against using an AI bot as long as the comment is on-topic and not duplicated from another source without a source credit.

In response, I gave the following answer, but they could not come up with any logical explanation:
If the users have to refere a credit, they have to write "generated by chatgpt" at the end. Otherwise it's a "duplicated" content from a digital mind without a source credit.

If you can't jointly address critical issues about the forum, which is all about content, it means that there is a serious communication breakdown within the team and this is a much worse situation than the AI-generated content.

You need to stop playing ostrich before things get out of hand.

---

I have a strong suspicion people are getting confused over what's AI-generated versus what's just low-quality content from humans.

You're not the only tech person here. And you won't get anywhere with suspicion.

 

[Paul]

The argument they argued in the PM:
There's technically no rule against using an AI bot as long as the comment is on-topic and not duplicated from another source without a source credit.

That's an oversimplification.

It's simple: we have no way of determining whether content was generated by good LLM. Most of the content that gets reported as LLM-generated almost certainly wasn't. It's not a very good rule if we can't enforce it.

If content violates our rules because it's spam, misleading, or otherwise problematic, report it for that. Don't report it just because you think it's LLM-generated. There is no automated system to detect that--as the paper @Lox mentioned explains--and humans aren't great at it, either, unless they're able to converse with the AI.

You're not the only tech person here. And you won't get anywhere with suspicion.

I can't prove that content isn't LLM-generated, so it will remain suspicion--as will any claims that content is LLM-generated. If you have concrete examples, this discussion would be furthered by links. Otherwise, I can't really help, and our policy remains the same.

 

[Traderas]

I think it would be useful to add a transaction range, when someone gives it feedback.
There are members with dozens feedbacks of small transactions, like 5-10$ from auctions, that proves little about their credibility.

 

[NameInbox]

add some Real Traffic/real people against bot Traffic

 

[Samer]

The ability to separate ”.com” and “non-com” in appraisals section— Not all asked appraisal are equal.

 

[Samer]

Get rid of anonymous downvote shills.
You love being toxic.
NP can keep “dislike“ and or “disagree”.

 

[ARCCLOUD]

Live domain auctions.

 

[NameInbox]

Live domain auctions.

You mean this(in which I am participating, see att) are dead or what ?
https://www.namepros.com/marketplace/buy-domains/?prefix_id=325

 

[ARCCLOUD]

You mean this(in which I am participating, see att) are dead or what ?
https://www.namepros.com/marketplace/buy-domains/?prefix_id=325
Show attachment 243072

Real-time auctions for domain names, similar to traditional auction houses, but online. Unlike our current 48-72 hour auctions, these would kick off at a set time (e.g., 8pm EST) and continue until no further bids are made.

This change could enhance member participation and potentially increase final sale prices due to the competitive atmosphere.

 

[MasterOfMyDomains]

This place would have thousands of auctions going if that happened.
Crap names galore. If you dont get bids there is a reason and its not a lack of views.

 

[NameInbox]

Oh, I see.
Yep, it may help even NamePros to grow in some way

 

[NameInbox]

views.

I believe views maybe from bots as well, I wrote about this before, even here:
https://www.namepros.com/threads/what-should-namepros-add.920208/post-8967007

https://www.namepros.com/online/
Form Kyrgyzstan there were even much more sometimes(more than UK or IN)

p.s. I'm using similarweb for free. It works fine for me.

 

[NameInbox]

Also, now statistic info is hidden(before there was a link, like in case with replies, so after click on link you may see more information). Was it removed by NP or forum update I'm not sure.

 

[ARCCLOUD]

This place would have thousands of auctions going if that happened.
Crap names galore. If you dont get bids there is a reason and its not a lack of views.

I envision these live auctions primarily for our premium domain names that are listed here occasionally. Perhaps a moderator could review and approve these listings to ensure quality? Just an idea.