AI usage, 3rd time in a few months

[xLays]

I tagged the member in question.  When there's no rule in place to deal with this sort of shenanigans, one must use the tools available to them to combat it.

I don't care about sig payment because I make x5 as a VIP wager.

Whaaaaat? Do you think we will believe you? You post 30 posts a day, and yet you're saying you don't care about the signature campaign payment? You must be kidding me. LOL

[1714607264]

1714607264

[1714607264]

1714607264

[1714607264]

1714607264

[Ultegra134]

Now you're messaging me in private.

Hey,

I don't know why you're battling me but within yourself you know I'm 100% not using AI. Got you pretty good merits for defamation. Cool.

Thanks anyway... Have nothing to say to you.

Within myself, I know it's pretty clear that you're using AI, you're not fooling anyone and it's showing even without using a scanner. So from how you're seeing things, I'm "battling" you, just to get a few merit points?

You literally wrote 30 posts yesterday, from 6 am to 10.30 pm, doesn't sound like you don't care. I don't understand how it's even possible to write so many posts within a day. That's pure spam, you're posting within 10 minute timeframes. That means you're in front of a computer for over 16 hours, so there's no way in hell you can claim you don't care, that's practically your job.

[FinneysTrueVision]

The posts quoted all seemed vaguely familiar that I thought they were older posts. They are actually quite recent but the reason they seem familiar is because this user has a habit of repeating the same thing over and over again but phrasing it a little differently each time.

Here are some examples of them talking about chasing losses in gambling.
https://ninjastic.space/search?author=borovichok&content=chasing

The most recent of these results all have that uncanny valley effect where you can just tell they were written by AI. Some of them are detected as being partially or wholly written by AI. Some others are not but I suspect they could be using something like the “more casual” setting in Google Gemini or perhaps an AI humanizer.

[borovichok]

Within myself, I know it's pretty clear that you're using AI, you're not fooling anyone and it's showing even without using a scanner. So from how you're seeing things, I'm "battling" you, just to get a few merit points?

You literally wrote 30 posts yesterday, from 6 am to 10.30 pm, doesn't sound like you don't care. I don't understand how it's even possible to write so many posts within a day. That's pure spam, you're posting within 10 minute timeframes. That means you're in front of a computer for over 16 hours, so there's no way in hell you can claim you don't care, that's practically your job.

Exposing a PM without my consent shows your lack of trustworthiness. It's been more than 8 hours, and no one, including you, who created the so-called "human and AI" mixing report, has provided/named at least one Chatbot capable of producing discussion on topic. There is a $50 bounty.

Do you know what spam is?  Writing more than 30 or 50 posts per day does not necessarily imply spamming; what matters is the content, not the quantity. My posts are human-written, but a random profile can't be so good unless he's using some form of AI.

Let me educate you on how AI works....

AI-written contents are words collected from the Internet, and most plagiarism detection tools detect them since they are randomly combined content from many sources.  No plagiarism tools have yet detected any AI-claimed posts of mind. You folks are playing assumptions game with someone's reputation.

The most recent of these results all have that uncanny valley effect where you can just tell they were written by AI. Some of them are detected as being partially or wholly written by AI. Some others are not but I suspect they could be using something like the “more casual” setting in Google Gemini or perhaps an AI humanizer.

Still doesn't make sense for anyone to use AI on few of his posts and go Human on most of them. The only tools I've ever used are quilbot and Grammarly( Noticed they're AI powered tools so stopped using) to enhance my sentences and punctuations. I'm not familiar with any of the tools you mentioned. Honestly appreciate your inputs.

[Ultegra134]

Exposing a PM without my consent shows your lack of trustworthiness. It's been more than 8 hours, and no one, including you, who created the so-called "human and AI" mixing report, has provided/named at least one Chatbot capable of producing discussion on topic. There is a $50 bounty.

Do you know what spam is?  Writing more than 30 or 50 posts per day does not necessarily imply spamming; what matters is the content, not the quantity. My posts are human-written, but a random profile can't be so good unless he's using some form of AI.

Let me educate you on how AI works....

AI-written contents are words collected from the Internet, and most plagiarism detection tools detect them since they are randomly combined content from many sources.  No plagiarism tools have yet detected any AI-claimed posts of mind. You folks are playing assumptions game with someone's reputation.

Exposing you? There's an open report here, everything you'd like to say, say it here, there's no reason to send me a PM and claiming that I have something personal against you in order to gain merit. You're free to share your though here, unless you're afraid to be proven wrong from other members who know me and my history, that there's no personal gain by this report.

[yahoo62278]

Within myself, I know it's pretty clear that you're using AI, you're not fooling anyone and it's showing even without using a scanner. So from how you're seeing things, I'm "battling" you, just to get a few merit points?

You literally wrote 30 posts yesterday, from 6 am to 10.30 pm, doesn't sound like you don't care. I don't understand how it's even possible to write so many posts within a day. That's pure spam, you're posting within 10 minute timeframes. That means you're in front of a computer for over 16 hours, so there's no way in hell you can claim you don't care, that's practically your job.

Exposing a PM without my consent shows your lack of trustworthiness. It's been more than 8 hours, and no one, including you, who created the so-called "human and AI" mixing report, has provided/named at least one Chatbot capable of producing discussion on topic. There is a $50 bounty.

Do you know what spam is?  Writing more than 30 or 50 posts per day does not necessarily imply spamming; what matters is the content, not the quantity. My posts are human-written, but a random profile can't be so good unless he's using some form of AI.

Let me educate you on how AI works....

AI-written contents are words collected from the Internet, and most plagiarism detection tools detect them since they are randomly combined content from many sources.  No plagiarism tools have yet detected any AI-claimed posts of mind. You folks are playing assumptions game with someone's reputation.

The most recent of these results all have that uncanny valley effect where you can just tell they were written by AI. Some of them are detected as being partially or wholly written by AI. Some others are not but I suspect they could be using something like the “more casual” setting in Google Gemini or perhaps an AI humanizer.

Still doesn't make sense for anyone to use AI on few of his posts and go Human on most of them. The only tools I've ever used are quilbot and Grammarly( Noticed they're AI powered tools so stopped using) to enhance my sentences and punctuations. I'm not familiar with any of the tools you mentioned. Honestly appreciate your inputs.

Anyone writing 30-50 posts a day, no matter the content, is going to get themselves on some users radar. Someone should run a check on the number of posts you made prior to being put in the stake campaign, then the number you make upon being accepted and see if there is a huge change in your habits. If so, at the very least you should be hit for sig spamming as 30 posts in 4 hours is a bit ridiculous and might show that you're only posting for money.

As far as the AI goes, the mods agree apparently since there are so many reports marked good. Being as AI is still fairly new, there might be some errors in detecting AI content, but if it walks like a duck........

[Aanuoluwatofunmi]

It still bothers me when i see a set of five or six people overseeing the affairs of a signatory campaign and couldn't detect for the use of AI from their participants, am sure it may not have to be this same guy alone, there may be more other participants found on the same shade for using AI, now, it is clear that this guy in question had been caught severally and reported, but the campaign managers are doing nothing to take action against such, which could in other way implies that they accept the use of AI for their postings in the campaign, if it were to be other managers, they would have taken quick action on that.

[The Sceptical Chymist]

The trust feedback cannot block all the holes in the forum moderation policies or it starts to lose its own relevance.

The trust system, in my eyes at least, lost its relevance a long time ago and as far as me misusing it in a case like this, I don't think I'm doing that at all.  The threat of AI ruining this forum seems very real to me, and it's unknown if Theymos intends to take any meaningful action about it.  Handing out a negative trust to a member who's most likely using AI to shitpost for a campaign reminds me of the months before the merit system was created; if the trust system is the only tool you've got to fight these morons, then that's the one you use until someone with authority tells you otherwise.

As far as the AI goes, the mods agree apparently since there are so many reports marked good. Being as AI is still fairly new, there might be some errors in detecting AI content, but if it walks like a duck........

This (in addition to the imminent threat) is why I'm inclined to shoot first and ask questions later.  I'll sleep just fine, and the rest of you can do what you like, but AI-generated shitposts might potentially be an overwhelming problem very soon.  And honestly, this is the first member I've tagged for this.  It isn't like I'm on a rampage circa December 2017.

[yahoo62278]

It still bothers me when i see a set of five or six people overseeing the affairs of a signatory campaign and couldn't detect for the use of AI from their participants, am sure it may not have to be this same guy alone, there may be more other participants found on the same shade for using AI, now, it is clear that this guy in question had been caught severally and reported, but the campaign managers are doing nothing to take action against such, which could in other way implies that they accept the use of AI for their postings in the campaign, if it were to be other managers, they would have taken quick action on that.

AI as far as I know is not deemed illegal. Is it good to allow the use of AI in users posts? No, not at all. If we allow AI there will be no more genuine conversation on the forum and 5000 users spamming a day to earn money.

I cannot speak for the managers of any campaigns, some of them may not be aware of any AI detecting tools. Detecting AI is not 100% perfect either.

As far as the AI goes, the mods agree apparently since there are so many reports marked good. Being as AI is still fairly new, there might be some errors in detecting AI content, but if it walks like a duck........

This (in addition to the imminent threat) is why I'm inclined to shoot first and ask questions later.  I'll sleep just fine, and the rest of you can do what you like, but AI-generated shitposts might potentially be an overwhelming problem very soon.  And honestly, this is the first member I've tagged for this.  It isn't like I'm on a rampage circa December 2017.

You or someone should make a list listing all suspected AI posters and get them all blacklisted from campaigns. Let other users see that there will be consequences for their actions.

[Lucius]

~snip~
This (in addition to the imminent threat) is why I'm inclined to shoot first and ask questions later.  I'll sleep just fine, and the rest of you can do what you like, but AI-generated shitposts might potentially be an overwhelming problem very soon.  And honestly, this is the first member I've tagged for this.  It isn't like I'm on a rampage circa December 2017.

AI generated posts have been a problem for years, at least I've been aware of it since mid-2020 when @fillippone opened this topic AI writing messages on Bitcointalk.org. I cannot say how much content on the forum is currently generated with the help of AI, but I believe that it is already a significant percentage.

~snip~
You or someone should make a list listing all suspected AI posters and get them all blacklisted from campaigns. Let other users see that there will be consequences for their actions.

That would only have a partial effect, because there are always managers who don't care, as is the case here. I think I even read that the user in question here was somehow promoted (rewarded) for his good work in the campaign. Tags, flags and black lists for such managers are unfortunately worthless.

[FinneysTrueVision]

Anyone writing 30-50 posts a day, no matter the content, is going to get themselves on some users radar. Someone should run a check on the number of posts you made prior to being put in the stake campaign, then the number you make upon being accepted and see if there is a huge change in your habits. If so, at the very least you should be hit for sig spamming as 30 posts in 4 hours is a bit ridiculous and might show that you're only posting for money.

As far as the AI goes, the mods agree apparently since there are so many reports marked good. Being as AI is still fairly new, there might be some errors in detecting AI content, but if it walks like a duck........

According to Ninjastic.space, prior to the Stake campaign he had 2550 posts since joining the forum in October 2017, most of them being made after his account woke up in September 2021 and started spamming with a high level of activity. After joining he’s ramped up his spamming and made 3885 posts in 9.5 months.

[JollyGood]

This is a theory I never thought of before but it fits in with the bigger picture that they have many super spammers and low quality posters on their signature campaign and not only that, they have many members that have received multiple neutral and negative tags.

Having said that, I would be more inclined to think that in this sort of scenario the campaign manager is the one that would be controlling the selection process therefore the company might be completely innocent and the whole problem was created as a smokescreen by the campaign manager as he runs the multiple accounts and takes the money for himself.

Is that more probable in this case considering Stake is a massive enterprise and are not exactly short on cash for running campaigns?

~snip~
Another explanation may be that multiple people use the same account, or multiple accounts, which is why you can notice differences in his posts, in terms of writing style, grammar and AI usage. It's extremely evident in posts that he's expressing his opinion and in others that actually require some knowledge to answer, in terms of definitions, technical specifications etc.

I would be inclined to exactly that theory, considering that the manager wholeheartedly defends exactly such shitposters who could actually be part of the alt network that the company bought to promote itself on the forum. I wouldn't be surprised if they then hired one or more people to manage that network, and of course they use AI and paraphrasing to do their job.

It is obvious that such a way of promoting pays off, because the investment in buying alt accounts is returned in the way that they do not actually pay unique members to advertise them. Unfortunately, apart from the persistent report such post and the hope that some of the global mods will react with a ban, red tags or flags are quite ineffective for such things.

[borovichok]

I will remove signature and Avatar once my resignation message is approved. I've left all my post unedited, when a good AI generated detector is finally developed, I will come back to this thread. Until then Good bye everyone, it was a nice ride.

My last advice: Do not trust any AI detector tools yet. I should be the last victim.



Edit:
Resignation approved! I've asked the admin to deactivate the account. Cheers

[LoyceV]

Does bitcointalk consider AI generated post here in bitcointalk as plagiarism?

I think this applies:

To be so lazy that you won't even make up your own low-quality post, but instead save a few minutes by importing some garbage from the wider Internet, is a massive insult to this forum's community, and is proof that at least in that moment the only thing you cared about was money-making.

If yes, he's candidate for a permanent ban. I think the forum admin needs to step up for this issue because it's not just one person who uses AI generated posts; it happens every day here on Bitcointalk. It's time to make forum rules regarding this matter.

I've been surprised (and demotivated) by the lack of bans for chatbot plagiarism. It's very easy to produce, and time consuming to detect and prove. And because the forum barely bans them, the number of chatbot posts is only going up.

I tagged the member in question.  When there's no rule in place to deal with this sort of shenanigans, one must use the tools available to them to combat it.

I hadn't looked at is that way yet. I always assumed it falls under the plagiarism rules, and thus should be handled by forum rules. But the lack of action from the forum could indeed be interpreted as a lack of rules against it, and in that case negative feedback might lead to change the same way it was used to tag shitposters before theymos introduced the Merit system.

[lovesmayfamilis]

That is, if you follow the thoughts of The Sceptical Chymist and how he reacted to the decision to leave a negative review, is it worth giving users several chances to correct their posting?

For example, we currently leave neutral reviews for those who have been caught using AI and have had their posts deleted by moderators. Everything happens logically: a report is sent to the moderators, they agree and delete the posts, and the user receives a tag.

But often accounts, even after receiving this tag, continue to use AI in the same way. How many times does it take for the user to wake up and realize that the neutral tag will soon turn into a red tag?

[LoyceV]

How many times does it take for the user to wake up and realize that the neutral tag will soon turn into a red tag?

Doesn't matter, they can just as easily create a new account. Chatbots are patient.
All it takes now it for someone to automate the whole process: create 10 new accounts per day, make each of them them automatically post 5 "AI" generated posts, and see what happens.
After a week: 70 accounts, with 350 posts per day. A week later: 140 accounts, with 700 posts per day. If half of them get banned, half remain. If some get Merited, they might rank up. After a year, the spammer may have enough surviving accounts to start earning money from signature campaigns. If they get banned, it doesn't matter because he still creates 10 new accounts per day. I don't have high hopes for the future of forums on the internet.

[nutildah]

I will remove signature and Avatar once my resignation message is approved. I've left all my post unedited, when a good AI generated detector is finally developed, I will come back to this thread. Until then Good bye everyone, it was a nice ride.

My last advice: Do not trust any AI detector tools yet. I should be the last victim.



Edit:
Resignation approved! I've asked the admin to deactivate the account. Cheers

Amazing... the guy basically said he'll come back to the forum when an undetectable AI tool has been developed, lol. The sheer audacity to deny posting dozens of straight-up ChatGPT texts and then declare themselves a "victim." Bravo. Stunning, yet also brave.

[LoyceV]

I've left all my post unedited, when a good AI generated detector is finally developed, I will come back to this thread.

Amazing... the guy basically said he'll come back to the forum when an undetectable AI tool has been developed, lol.

Actually, he said the opposite. He's saying the current AI detectors are inaccurate. That's kinda true, but in his case I don't believe that.

[ABCbits]

How many times does it take for the user to wake up and realize that the neutral tag will soon turn into a red tag?

Doesn't matter, they can just as easily create a new account. Chatbots are patient.
All it takes now it for someone to automate the whole process: create 10 new accounts per day, make each of them them automatically post 5 "AI" generated posts, and see what happens.
After a week: 70 accounts, with 350 posts per day. A week later: 140 accounts, with 700 posts per day. If half of them get banned, half remain. If some get Merited, they might rank up. After a year, the spammer may have enough surviving accounts to start earning money from signature campaigns. If they get banned, it doesn't matter because he still creates 10 new accounts per day. I don't have high hopes for the future of forums on the internet.

But without some automation, i wonder how many people actually patient enough to do that. And if someone have skill to perform automation (where they also need to consider CloudFlare), they probably can land IT job which offer better stability.

[Lucius]

But without some automation, i wonder how many people actually patient enough to do that. And if someone have skill to perform automation (where they also need to consider CloudFlare), they probably can land IT job which offer better stability.

In some cases, it is certainly about individuals who try to cheat the system, but I believe that there are also groups with many more members who operate alt accounts on this forum in a much more sophisticated and thoughtful way in order to position themselves in the long term.

Also, the possibility that some people do it against their will should not be completely dismissed, because as we can often read, in some parts of the world, people are dragged into forced labor in various ways, especially when it comes to "jobs" related to cryptocurrencies.