Shouldn't there be a forum rule against obvious AI-generated content?

[Odohu]

I know what plagiarism is, and I know it's technically plagiarism, but I said it's not provable or clear plagiarism because you can't find concrete evidence for it since the content generated is only available to that specific user and no one else, so if you report it as plagiarism, you can't prove it, technically.

You can use that argument to say that you can't prove with 100% certainty that someone uses AI bot as well which would then make the rule against AI generated content useless.

As others said, plagiarism rule covers this nicely, but unfortunately mods are less leniant to ban people over that type of plagiarism (some would say for any type of plagiarism as well). At least for now.

Why I think the mods use the hammer on users who are engage in the use of AI is that allowing AI to thrive in this forum defeats the core purpose of the forum which I guess is a place of organic discussion of Bitcoin, Blockchain technology and every other thing that is connected to the digital economy. Allowing AI kills the originality requirements and will make here to be heavily infested with bots, the consequence of which will be that no one will take the forum seriously. So, I love the approach adopted by the mods regarding the use of AI in a forum like ours.

[nutildah]

For me AI generated content in forum fits nicely under rule number 33:

Posting plagiarized content is not allowed.

Agreed. But why don't they all get banned instantly?

This is a good point. Only some of them are getting banned. It seems we're at the point where most regular forum members have moved away from trying to use AI (except for in the Gambling section) and now the only ones who do it are newbie bots and promoters of projects. If a newbie just writes a bunch of spam posts right off the bat, they usually get banned; same thing should apply to AI spammers.

I'm thinking about how much AI spam there might be right now if I hadn't started my thread 

[Vod]

I do not think any new rule needs to be added just because of AI content creation.

It can be treated under plagiarism or even spam.

But AI is not plagiarism.  

AI is a tool that many people here use to generate text.    Could be because they are lazy, or aren't native to the language.  

Consider a user who watches a bitcoin story and has a great political cartoon idea.   He/She can't draw worth shit, so he describes the idea to AI.   He then posts the image here and starts a passionate discussion about a novel idea.  Why is that punishable?

Did you know that websites like BPIP, Nintastic, ImgTalk, etc all use plagiarised code?   It's legal in the form of "packages", so why shouldn't AI content be legal as well?

From this definition, we can say that posting an AI-generated post is plagiarism because they took a work that's not theirs and posted it as theirs.

So each of the websites above should list all authors of each open source package they use or be banned? 

[Apocollapse]

Agreed. But why don't they all get banned instantly?

Because the forum don't want to lose more traffic, not all reports in plagiarism are taken action by moderators, only the posts are being deleted, while the users didn't get banned. So, this might be the reason why all users that use AI didn't get banned.

So each of the websites above should list all authors of each open source package they use or be banned? 

There are no rules in this forum the developers will be banned for committing code plagiarism.

What I know there's a license to protect digital property, so the authors can sue them for breaching the license.

[nutildah]

But AI is not plagiarism.

I disagree. Let's look at the definition:

an act or instance of using or closely imitating the language and thoughts of another author without authorization and the representation of that author's work as one's own, as by not crediting the original author

The original author is not the poster. Its ChatGPT. If you credit ChatGPT for the post, then its not plagiarism. AI posters are misrepresenting their actual abilities, and most of the time the output is perfunctory blather that doesn't add anything to the conversation (which is why there is a lot of overlap between spam and AI text).

Consider a user who watches a bitcoin story and has a great political cartoon idea.   He/She can't draw worth shit, so he describes the idea to AI.   He then posts the image here and starts a passionate discussion about a novel idea.  Why is that punishable?

I don't care if people post AI generated images (unless they are extremely bad to the point of being offensive). I doubt most other people care either. It would be tacky for the person posting the image to say "I made this using my own brain," or imply something similar.

Let's say for instance someone designing a series of Bitcoin collectibles uses AI-generated images to "showcase their work." Its not their work at all -- it doesn't even exist. This would not be a particularly trust-inspiring act. Its all about the context in which AI-generated material is being used. If its being used to meet shitpost quotas, I'm against it.

[hd49728]

The original author is not the poster. Its ChatGPT. If you credit ChatGPT for the post, then its not plagiarism. AI posters are misrepresenting their actual abilities, and most of the time the output is perfunctory blather that doesn't add anything to the conversation (which is why there is a lot of overlap between spam and AI text).

The chatbot AI users never credit their posts to ChatGPT or other AIs because they intend to steal contents created by ChatGPT, AI and make it like their own content.

If they are in a campaign, campaign manager will not pay for posts credited to ChatGPT, AI.
If they are newbies, forum members will not merit these posts if there is credit to ChatGPT, AI.

Intention of ChatGPT, AI users is clear, and they don't forget to add credit but they want to hide source of the content.

So it's plagiarism.

I don't care if people post AI generated images (unless they are extremely bad to the point of being offensive)

They can use AI generated images in their posts for fun, like memes, but if they use these images for their participation in art contests, it's not acceptable.

[Vod]

But AI is not plagiarism.

I disagree. Let's look at the definition:

an act or instance of using or closely imitating the language and thoughts of another author without authorization and the representation of that author's work as one's own, as by not crediting the original author

Oh, the definition card.   I trump that with the details card:  An AI does not have thoughts.   AI is a tool much like Grammarly or a typewriter.

(forgive the sarcasm, you know I love you)

[LoyceV]

But AI is not plagiarism.

It's text written by someone (or something) else while the user pretends he wrote it by himself. That's plagiarism by definition.
The "AI" tool itself is highly debatable too: I've seen many accusations of them using unlicensed input data to train it, and they don't credit the real authors.

Consider a user who watches a bitcoin story and has a great political cartoon idea.   He/She can't draw worth shit, so he describes the idea to AI.   He then posts the image here and starts a passionate discussion about a novel idea.  Why is that punishable?

Now consider the same user who pays someone to draw for him, and pretends he did it by himself.

Did you know that websites like BPIP, Nintastic, ImgTalk, etc all use plagiarised code?   It's legal in the form of "packages", so why shouldn't AI content be legal as well?

The Open Source licenses allow this use, and (usually) allows you to make changes as long as you credit the original writers. By crediting them it's not plagiarism.

Because the forum don't want to lose more traffic

If by "the forum" you mean theymos, I don't think he cares about the amount of traffic. If anything, low-quality traffic isn't worth much.

The original author is not the poster. Its ChatGPT. If you credit ChatGPT for the post, then its not plagiarism.

It's probably still plagiarism, but from ChatGPT. They take bits and pieces of other people's work without crediting anyone.

If they are newbies, forum members will not merit these posts if there is credit to ChatGPT, AI.

Some even earn Merit, but I'm much more disappointed by the number of people responding to them. Even on the tech board: what's the point of responding to a made-up problem?

[lovesmayfamilis]

Yes, but it's not clear plagiarism because the content is not taken from somewhere on the web, or a forum or anywhere, it's just generated through a bot and used as if it's written by the user himself.

I'll show you an example. OP, you are misinterpreting your guesses about where the AI texts come from. And when we see this kind of behavior, where there are a lot of similar phrases torn out of a million articles on the Internet and collected together in another, next answer, it still does not belong to the person who copied and pasted it.
And yes, such posts can be seen not only by those who use AI, but also by those who often check such posts.
Just answer, can this post not be considered plagiarism when the tool shows other sources from which the sentences were torn out?

DCA is a strategy that invests a fixed amount of assets at regular intervals. The advantage of using this DCA is that it can reduce the risk of large investments, and if you invest small amounts over a period of time, you are less likely to lose significantly if the market shows a sudden downturn.
The advantage of the DCA method over single investments and large sums of money at once is that if the market is bullish, you have the potential for higher returns.

Dollar-Cost Averaging (DCA) A time-tested investment strategy, this strategy allows an investor to periodically spread an aggregate amount of money in an attempt to reduce the effect of volatility on the aggregate purchase of a target asset.

For example let's say you want to invest $200 in Bitcoin, instead of buying all $200 BTC you can buy $20 worth of BTC every week for ten consecutive weeks.

The main advantage of the DCA method is that it takes the emotion out of investing because you buy a fixed dollar amount of assets on a regular schedule and are less likely to make emotional decisions that could damage your investment portfolio.

[Lucius]

~snip~
Not get banned because mods not seeing AI contents as plagiarism? Now, I get why OP started this discussion. You nailed it @LoyceV .

It is quite easy to determine with 100% certainty that someone has plagiarized because you have the user's post and the source from which the content was copied - noting that there were cases where some users copied their content from blogs or other forums without citing the source, which then it's definitely not plagiarism if they can prove it.

On the other hand, you have a lot of AI detectors, some of which are very bad and give a lot of false detections, which is why I accused one forum user of using AI, which was not the case. Therefore, one should be careful when using such tools and always use at least three reliable tools before making a final conclusion.

I think that the mods are very careful in the case of AI reports, because there is a possibility that someone will be permanently banned due to bad detection, and that's why you need more evidence that someone is really using AI in order to get a permanent ban.

I would personally try to influence all members (personally new ones) with a clear message that the use of AI can lead them into problems and that they should think twice about using it on the forum. Of course, there are users who use AI to create images and that's perfectly OK and can be fun, as is the case in WO sometimes.

[Alone055]

Thus even if the rule is  been created(apart from the plagiarism), the content still needs to be reported and the  decisions will still have to be carried out by mod.

I know members will be reporting AI-generated content and the users being reported should get some punishment along with deleting their posts.

Because the forum don't want to lose more traffic, not all reports in plagiarism are taken action by moderators, only the posts are being deleted, while the users didn't get banned. So, this might be the reason why all users that use AI didn't get banned.

About plagiarism, I don't think a user is banned if it's proven that they have been involved in plagiarism, but when it comes to AI-generated content, users aren't getting banned, only the posts are deleted. What you said about traffic doesn't make any sense because you wouldn't let people keep breaking the rules only because you don't want to lose some traffic.

I'll show you an example. OP, you are misinterpreting your guesses about where the AI texts come from. And when we see this kind of behavior, where there are a lot of similar phrases torn out of a million articles on the Internet and collected together in another, next answer, it still does not belong to the person who copied and pasted it.
And yes, such posts can be seen not only by those who use AI, but also by those who often check such posts.
Just answer, can this post not be considered plagiarism when the tool shows other sources from which the sentences were torn out?

I understand, I know that AI-generated content is plagiarism but it's plagiarism on the part of the bot providing that content, and the reason why I believe it's wrong to use those texts as someone's own is because they are not referencing that they have used an AI model or a bot to generate that. If I use ChatGPT to generate a code or anything, and I mention that I have used ChatGPT to do this, it shouldn't be a problem, but if I don't do that, it means I'm trying to use that thing as my own and that is wrong. I understand it's plagiarism of a sort, but we currently don't have strict rules regarding it, hence this thread.

Those who say AI-generated content falls under the plagiarism rule in the forum rules list, I want to ask them the same question asked by LoyceV earlier. Why aren't users who are reported with evidence getting banned? If someone is found guilty of plagiarism and there is a rule against it, the person should get punished for that. Having their posts deleted and leaving them to keep doing what they have been doing isn't how it should be dealt with if it falls under a forum rule and they are breaking it again and again.

As an example, you can look at the post history of this guy, it's the same user that I referenced in the OP. I reported some posts from this user yesterday, all of them were AI-generated, and both I and LoyceV, reported him in the AI Spam Report Reference Thread but I upon checking his post history again today, I can see that the posts reported have been deleted but he did a few again after that, without getting any punishment for it. Why?

If this falls under plagiarism, the punishment should be the same for it, and if it doesn't, there should be a separate rule and punishment for it. That is the basic reason why I started this thread.

[lovesmayfamilis]

If this falls under plagiarism, the punishment should be the same for it, and if it doesn't, there should be a separate rule and punishment for it. That is the basic reason why I started this thread.

I think it's also worth asking why they don't ban those who have plagiarism found, and it doesn't relate to AI. And only after that can we start talking about a complete ban on those who use AI to create posts.
Probably because the number of users on the forum is not increasing, the moderators are protecting everyone, hoping that people will respect the forum rules.

[Vod]

Did you know that websites like BPIP, Nintastic, ImgTalk, etc all use plagiarised code?   It's legal in the form of "packages", so why shouldn't AI content be legal as well?

The Open Source licenses allow this use, and (usually) allows you to make changes as long as you credit the original writers. By crediting them it's not plagiarism.

I don't believe that these websites are plagiarizing anything, even if they don't credit the thousands of authors on each project.    They are not changing the code, just using it.

One thing is for sure - the lawyers will make their money on changing terms alone.   Next time my word processor suggests "you are" instead of "your", I don't want to have to credit the software, or the people that wrote it.   

[dkbit98]

Yes, but it's not clear plagiarism because the content is not taken from somewhere on the web, or a forum or anywhere, it's just generated through a bot and used as if it's written by the user himself.

This bots are not inventing that from thin air, they are using copy-pasting and combining what other human being wrote before in different websites.
Basically all AI bots are doing the same thing as a human who is doing plagiarism but than makes little changes and mistakes to make it more human-like and harder to detect for plagiarism.

Agreed. But why don't they all get banned instantly?

Because it's not always easy to detect them with high percentage of accuracy.

But AI is not plagiarism.  

Maybe only for people who are lazy and can't use their brain for thinking and fingers for typing.
If those AI bots open their source code you would found written source of every sentence and word written and copied from human beings.

[Vod]

Maybe only for people who are lazy and can't use their brain for thinking and fingers for typing.

I am humbled that the industry is spending billions to help me.  My left fingers are slow, and I can copy/paste much faster than typing. 

[EarnOnVictor]

I know that if we report AI-generated texts or content to moderators, they get deleted, and I also know that there is a thread for reporting AI-generated posts (I'm unable to find that thread at the moment, sorry), but, shouldn't there be an official rule in the unofficial list of forum rules against obvious AI-generated content and posts?

You mean you want an official rule for AI-generated content when the forum itself does not have an official rule that governs anything but unofficial ones? Regardless, I believe there is a reason for that, nevertheless, it's a good call to distinctively define the forum's view on AI-generated content.

For me, I am confused about this and might be the reason why the forum is silent about it after many reports about it. AI is being used in this advanced age where even reputable companies and educational sectors are mastering its work. Are they not? So, to take a side here is so technical. It's good that moderators have been doing a great job about it at their discretion and it's our duty to report more of such content that we suspect and give moderators the benefit of the judgmental doubt to deliver the right verdict.

[Alone055]

This bots are not inventing that from thin air, they are using copy-pasting and combining what other human being wrote before in different websites.
Basically all AI bots are doing the same thing as a human who is doing plagiarism but than makes little changes and mistakes to make it more human-like and harder to detect for plagiarism.

Okay, let me ask you a question. If I read some information over the internet, I memorize it, and then I rewrite that information somewhere else with the same meaning but different wording, etc. Will that be plagiarism? I don't think so. The AI models do the same thing. They don't copy and paste exact words or sentences from other sources, they take information from the data that is fed to them, process that information, create a different version of it, and give it to you as a user asking for it.

The point is, that if you, as a user, take that information from ChatGPT and then post it as your own, you are doing a wrong thing because it's not your work.

There is one more thing. Experiment with ChatGPT just for reference. Create a prompt that would generate some text about anything you want, and use the same prompt with two different accounts or on two different devices. You will see that the answer might be the same by meaning but it won't be the exact same by wording.

Because it's not always easy to detect them with high percentage of accuracy.

What about the ones that are reported in threads with proofs and everything? If it's about AI-detecting websites or applications that mods might think aren't accurate, one can be wrong, if multiple detectors are giving the same results, that should prove it.

You mean you want an official rule for AI-generated content when the forum itself does not have an official rule that governs anything but unofficial ones?

Hey, you are probably confused about this, but let me tell you, the word "unofficial" in the title of Unofficial list of (official) Bitcointalk.org rules, guidelines, FAQ thread is an adjective used for the word "list". So it's basically saying that the list is unofficial but it's for the official (you can see it written in brackets in the title) forum rules.  

[joker_josue]

Did you know that websites like BPIP, Nintastic, ImgTalk, etc all use plagiarised code?   It's legal in the form of "packages", so why shouldn't AI content be legal as well?

The Open Source licenses allow this use, and (usually) allows you to make changes as long as you credit the original writers. By crediting them it's not plagiarism.

In programming things will be a little different.

First, as mentioned, a lot of content can be developed based on Open Source code. Some of this code requires credits to be given, others do not.

Second, the code generated by AI is public code, that is, no one can demand exclusivity. So, in certain cases it will be the same as using public code libraries. It is then up to whoever uses this code to validate, verify and analyze whether everything will be correct and functional. In the end, several details must be evaluated to understand whether or not the script can be considered yours.

One thing is certain, even the code generated by AI does not invalidate not having programming knowledge, otherwise you are exposed to many bugs.

Writing a text, 100% by AI, is something completely different. Because you don't need any specific level of knowledge for everything to "work".

[CryptoHeadlineNews]

Maybe only for people who are lazy and can't use their brain for thinking and fingers for typing.

I am humbled that the industry is spending billions to help me.  My left fingers are slow, and I can copy/paste much faster than typing. 

Lolz😃.. You actually sounds funny.


But however, using an A.I generated content is one thing that shouldn't be frown at, as it adds no value to forum discussion, since they are not genuine and coming from a user who has got experience, but simply robot. And as such, just as many people suggested, it's punishment should be treated just like that of "plagiarism" with a temporary or permanent banned, whereas if the account seems to have been of higher rank, you just red-tag the account with (i.e negative Trust), so as not to be able to partake in signature campaign ever again.

[DYING_S0UL]

But AI is not plagiarism.

I disagree. Let's look at the definition:

an act or instance of using or closely imitating the language and thoughts of another author without authorization and the representation of that author's work as one's own, as by not crediting the original author

Oh, the definition card.   I trump that with the details card:  An AI does not have thoughts.   AI is a tool much like Grammarly or a typewriter.

(forgive the sarcasm, you know I love you)

I sometimes use similar tools like grammerly, quilbot (mostly) or a typewriter, when I'm writing a long post. (You know, just to make sure that there are no grammatical or punctual errors. ). Does that makes my post a plagiarism? Umm, I don't think so. The content was originally mine, I just fixed it. 

afaik, if you are not the original author and don't mention the source, it will be plagiarism. AI is basically giving you content that you never wrote, and people are just using it as their own. So it's plagiarism.