Topic: What's AI Written Post? (Read 928 times)

Thanks for finding this and posting it... I spent a few minutes reading through the OpenAI terms and conditions the other day looking for something similar but didn't notice that clause.

I suspect we will continue to have lazy shitposters finding ways to skirt their ChatGPT copy/pastes from being detected by the detectors, rather than bothering to use their own brains to generate thoughts. Here's one such example:


Here you can see the post is buffered by original content ("Exactly!", and "Anyway, ..."). But the middle chunk is largely AI-generated.

Copyleaks: AI content detected
Hivemoderation: 100% likely to be AI generated
Sapling.ai: 100% fake

As a matter of fact I'm going thru this person's posts and will probably find a lot of them to report here. The Gambling section is chock full of such posts. Where will the moderators draw the line? We'll find out.

 I don’t think that there is really that much room for debate on labelling a post as plagiarism, in forum terms, when it is created by simply copying and pasting here AI generated content (without adding a source so as to mirror the norm). The fact that a person performing the said routine is intentionally trying to pass as his own content generated elsewhere, is what is really being forbidden here.

The forum doesn’t really care too much (as far as I can tell) if we’re talking about plagiarism to the line of it’s definition, but rather more it focuses on the intent behind the act itself. When that intent is to pretend that the poster was indeed the author of a given content, that is what contravenes the spirit of posting on the forum, being the subjacent origin of the content secondary to a certain degree.

Note: Open AI still states that content generated using their API should:
See: https://openai.com/policies/sharing-publication-policy#content-co-authored-with-the-openai-api-policy

So even from the point of view of Open AI, user’s should credit the fact that the content is AI generated when it is. The problem obviously resides in proving it, but that does not give one a free pass on his intent.

Only the source is different. With content stolen from the internet, it was easy to compare text A to text B to check if it was plagiarized because you could find the source. With AI it's different in the sense that the bot creates the content you ask it to. And even if you ask it to do the same thing multiple times, the outputs will probably be different.

Plagiarizing from AI is a new and non-traditional way. Other non-traditional ways are plagiarizing someone's speech. For instance, you listen or record someone speak on a conference about Bitcoin, and then you log into your Bitcointalk account and post that as your own. That's also plagiarism if you didn't make it clear that those aren't your words but someone else's and/or if you don't mention where you got the information from.       

If you post AI-generated text without crediting the AI, how is that any different from plagiarism? You are attempting to take credit for something that you didn't write.


For the AI to be committing plagiarism, it would have to be demonstrated that the AI is directly copying words from another source. It does occasionally do this, but not always.

Well, the problem is that using Artificial intelligence to create post is that it widely differs from the classic definition of plagiarism and the way it has been created for most of the history of humanity. You know, until not long ago it implies the copy/pasting and small modification of information which did not belong to us but try to fool others to believe it does. The main problem is the origin of the information, if the AI was legally recognized as an author, then it would easier to accuse someone using it to generate content for their own purposes to be a plagiarist, on the other hand, it would also imply the AI itself is committing massive plagiarism.

I would personally be in favor of restricting the use of AI generated posts in the forum, unless it's origin is explicitly stated and it is used for amusement/experimental purposes. Not to spam around.

It fits perfectly the definition of plagiarism if you are hiding the fact that you have used AI to create the post. And that's what people will do, either to get merits, sound more knowledgeable, or both. If nutildah created that poll he mentioned, I am sure that a big majority would vote for non-credited AI posts being plagiarism. But a different poll would be more interesting if it had the question, should AI-written posts be against forum rules and discouraged? 

As far as I know using Artificial intelligence to generate content and then post it here is not explicitly against the rules, some users believe it is some kind of plagiarism, while other disagree. The example you provided mostly applies in the context of signature campaigns, since people is supposed to be paid by using their account while commenting or generating meaningful posts around the forum, some managers may believe that using AI is a form to cheat the campaign while doing the minimum effort to actually engage in a high quality way with the rest of the forum users.

In my opinion, it would not be no problem if someone decides to use AI to generate content here, as long as it is explicitly disclosed they are doing so and why, it would also help if they person in question is not engaging in any signature campaign with a policy against AI generated posts.

Anyways, the way I see it, it will be mater of time before we start to see more complex ways to generate commentary and content which may be more difficult to spot, hopefully the developers of AI will implement some watermark for people to realize whether the content was done by a human being or not.

Interaction between humans will produce solutions in discussing a problem, but not with AI that only correlates with existing data.
Human intelligence has a better and stronger capacity for thinking than AI and human intelligence has great problem-solving skills depending on the core of the situation.

When the Forum is only generated as a BOT chat then there is no Ego in the conversation, there will be no emotion in every discussion, this is because BOT AI does not have Cognitive abilities like humans, has no common sense, AI cannot understand the concept of "Cause and Effect".

Absolutely.


Yes, if its mentioned that ChatGPT was used to write the text, then its OK, because you are crediting the source (similar to putting a reference). It isn't great, and posts that are copy/pastes of articles or tweets with a "Source:" appended to the end of it usually suck to read... But its not technically plagiarism so the posters get away with it.

I don't know about everyone else, but when I start reading a text that is clearly AI-generated, my eyes just kind of gloss over the words... I have no interest in reading it because I know a human didn't write it. We should really do our best not to let the forum turn into a place where this is the norm.

Why would that be needed? Asking an AI bot to write you a post on a topic or searching for that topic through Google or Bing is the same if your ultimate goal is to copy-paste the results and pretend you wrote it. The punishment should, in that case, be the same. There should either be punishment for both or neither of the two.

On the other hand, if you created an ANN for a casino (since that's the example you made), and you mentioned in the sources that it's AI-generated content, you wouldn't be accused of plagiarism. That still doesn't mean you should do it because this is a forum for human interaction and exchange of ideas and thoughts. It's not and shouldn't turn into a place where bots talk to each other with humans only giving commands.     

My reply might not be suited for this thread, if required please delete it. I need to share this news with everyone who thinks AI can make a difference. One of the creators of this project and the CEO, Sam Altman got fired from his post. The reason given is strange as the board feels," he was lying to the company". Isn't AI based on the concept of replicating human intelligence, which is a big lie? Not sure if it's an internal war to take control of the project initiated by Bill Gates.

I am only sharing here my point of view of what is happening in this AI sphere. I feel that we need to update with some internal guidelines that should help the newbie come up and create topics that are AI-generated. I don't think some of us trying our level best to control this AI infestation would achieve something positive as I have seen a few neutral tags on the Gambling board where owners are now getting tagged.

Creating an ANN using AI does not make any difference as what a casino needs is to create trust around their project. It does not matter creating an ANN with AI as it is real now and I am sure some of you would recognize its importance in marketing a product for a casino. I am not a supporter but I would hesitate to use a tool that I paid for to create good marketing content in a short period.

If you opened such a poll, I think at least 80% would vote that AI generated content is plagiarism unless the poster says it's AI generated.   

I don't know how often it happens either probably because it's time consuming to check and hard to prove in some cases.

There is no written rule about it, but do we know for sure that if there is proof that a user used AI generated content and got caught with it, that admins won't take action such as deleting the post and or temp/perma ban that user?

My stance on AI usage is very clear. Read my previous posts in this thread if you are in doubt. If you are saying that the forum is officially against it, then show me the rule that prohibits the use of AI. Or, show me a post of the administrators confirming that AI posts aren't allowed.

Why are you bringing red tags into the discussion? The trust system doesn't have anything to do with it, and anyone can post any feedback they want. I can give you a positive, neutral, or negative tag right now but that action of mine doesn't represent the forum's position. 

Your failure to understand what I am saying is exactly that: your failure. It's got nothing to do with me. My opinion on AI-generated content can be found in this thread. I assume you can find it and comprehend it. And you are again confusing apples and oranges. The trust system isn't the voice of the forum administration. It's a subjective rating system of the community that they can but don't have to use.

The Forum is against it or are you telling me that is ok to make use of AI? If is cool using AI why do we have red tags on members who uses AI?
Plagiarism, spamming, use of AI are all against the rules, if caught doing any of those the offender will have to face the punishment.



I still don't understand why you refuse to accept the fact that is part of the rules of the Forum like the community kick against the use of AI. To me why it's the way it is, is because some users still choose to go against the Forum rule to use AI, like they see it as a perfect way of making their post look clean. We know some ain't that good with English but it doesn't mean that you must make use of AI.

Well, if we'll look at the definition of the word "plagiarize" from Merriam-Webster dictionary, I guess there hardly can be a double understanding of if taking a text written not by you without crediting its source (even if the source is AI) is plagiarism or not.

If it's not a catch phrase, then crediting the source (including AI) is something expected.

So I guess that moderators don't always take action against AI users because there can be a problem in case of appeal on ban. If there is a text with a date from an outer source, you can easily check it any time. But there's no generally accepted standard for AI texts. We can see that cheaters try to use many different tricks to try to say that it is a mistake in AI detection. Sometimes what is obvious is not what is easily proved.

I'm curious (and may open my own poll about it shortly), is there anyone who believes that posting unaccredited AI text isn't plagiarism? It seems to me the general consensus is yes, it is plagiarism, but I am potentially biased because that's what I believe as well.

Aside from one famous example (and kudos to Royse777 for his no-AI policy), I rarely see accounts removed from campaigns after getting caught posting AI. This could be because campaign managers are largely unaware of it happening -- usually at most people will leave a neutral trust rating for the offense, if anything at all.

So AI posters exist in a legal loophole right now:

- its often hard to know 100% for sure that a post is AI-generated
- its not considered plagiarism by moderator standards (as of the moment)
- there's no precedent for red tagging users for doing this.

The forum doesn't have any specific rules on the use of AI bots, but it does have rules on plagiarism. If the admins deem that you have plagiarized, then you are in trouble regardless of whether you used an AI or took the content from elsewhere.
 
Again, the forum rules don't discuss AI posting in any way because they were written long before AI technology was introduced. You are right about the second part of your sentence. Some signature campaign managers have rules that prohibit the use of AI software. If users are found to be using them, they will get rejected from the campaign and perhaps DT members might give them negative feedback.

But this isn't in any way connected to forum rules. We need to make a distinction between what the forum allows/disallows and what isn't encouraged by the community.
For instance, scamming someone isn't against forum rules, and you won't get banned for it. But such actions can earn you negative feedback by forum users.

When you quote spam instead of reporting it, you help spammers. It's better to delete the spam content from the post you quoted.

They are directly promoting AI tools in the topic where we talk about why is it inappropriate, and you help to keep their link after the post is being deleted by moderators for spam.

If you see some spam, including AI spam, the best way is to tap a "Report to moderator" link under the post. And definitely not to quote it to preserve.

Some of these questions are making me feel you want to try making use of AI in your post. The Forum doesn't permit or encourage the use of AI in posting, it only signify one thing, that's laziness. Those who make use of AI to polish their words are just lazy, they can't make use of what they've upstairs to drop any thing that would attract people's attention.
If you don't know, just know that using AI is against the rules and regulations of this Forum and if caught (especially when you're in a campaign) your account won't be of any use to you.

AI generated content is in direct conflict with rule #33 about plagiarism for two reasons. Firstly, it's copy/pasted content from AI software, meaning it's not your words, and you aren't the creator of the content. Secondly, by copying that from an AI bot and posting it here on the forum under your name without giving credit or mentioning that it came from a bot, you make it seem like it's yours. You claim it to be yours. Put those two together and you get copy/pasted content without posting information about the source. That's plagiarism for you.

The forum rules don't mention it exclusively because they were written years ago before we had AI technology. But it doesn't matter. Regardless of where the content came from, if you post it under your name, pretending you came up with it, you planarized.   

Why should we, anything that has to do with what is not coming from us being the author of them is as null as having someone's else's work to claim, why are some people taking their time to make use of their head in making a quality post while others can do nothing than to go through the otherwise, this means that their feedback or posts aren't genuine on a sincere ground, they are just taking leverage on other sources for help or a way out.