This forum will need explicit rules on the use of AI.

[Don Pedro Dinero]

That's all. Surely the owner and the people most involved in it are thinking about it, but AIs are going to be used on the forum if they are not already being used, as they are being used all over the internet.

Better to come up with clear rules than to go willy-nilly.

[1714775535]

1714775535

[1714775535]

1714775535

[1714775535]

1714775535

[1714775535]

1714775535

[drwhobox]

Well AI has been used all over the internet for around three or two years. Maybe some people use AI here, but I don't think using AI is beneficial here in this forum.

Better to come up with clear rules than to go willy-nilly.

What kind of rules do you want to be implemented here? I don't think there is any need for extra rules because AI users will eventually get detected.

[digaran]

Here is the rule you are asking for.  33.Posting plagiarized content is not allowed.

[NotATether]

99% of AI users on Bitcointalk will just use it to attempt plagiarism and spam, which is already prohibited in forum rules. Also, it is straightforward to detect GPT3 content, there are AI detection sites that do just that.

[Hispo]

Here is the rule you are asking for.  Posting plagiarized content is not allowed.

In that case, we need admins to explicitly state on the rules that the use of artificial intelligence for the creation of content here is considered plagiarism. Otherwise, someone could argue that there was no connection between the AI and the old rule itself.

I would suggest the administration to ban AI for the "writing" of posts here on the forum unless it is clearly being used for reviewing or entertainment purposes, being mandatory for anyone to share the link to the AI used, after quoting the AI -generated text.

[Sandra_hakeem]

This whole thing works explicitly....
Y'all know what? It isn't that simple to amend a rule that has been placed over time, just for a willy-crafted bug like the artificial intelligence. well, I'm not actually saying that it's impossible though but alot needs to be done to get some rules amended.
I could remember reading something from an OP few days ago; that creating contents with the AI isn't against the rules, which is obviously true....WHY? Cus we haven't got any rules like that.
To my discretions, I think a refusal to amend the rules against these bug might result to breaking alot more rules; ofcourse, it's gonna encourage merit farming..... When everyother normal person tries their best to attain a rank, after alot of UPs and DOWNs, but they're others out there tryna outsmart or say -- manipulate the whole merit system, and you think that's gonna CUT IT???

Sandra 🧑‍🦰

[mprep]

If an AI is able to consistently create content that doesn't break any of the forum's rules (good quality, on-topic, not just a padded word salad, no plagiarism, etc, etc), I, for one, welcome our new machine overlords. Otherwise, content that violates the rules, AI or human produced, can already be dealt with with our current rules and policies in mind.

[philipma1957]

Here is the rule you are asking for.  Posting plagiarized content is not allowed.

In that case, we need admins to explicitly state on the rules that the use of artificial intelligence for the creation of content here is considered plagiarism. Otherwise, someone could argue that there was no connection between the AI and the old rule itself.

I would suggest the administration to ban AI for the "writing" of posts here on the forum unless it is clearly being used for reviewing or entertainment purposes, being mandatory for anyone to share the link to the AI used, after quoting the AI -generated text.

AI is not plagiarism.

Many people do not understand plagiarism.

You need a new rule to ban AI or to list the post as AI generated.

I would prefer making sure All AI is simply marked as such.

And all unmarked AI that is found is banned for hiding the fact it is banned.


My reasoning is if I do a google search and list the info here as a googled found item it is okay to do this.

Thus if AI is a better google allow it if marked as such.


@ mprep I guess simply saying it is ai is much the same as saying a google search is the source of the info.

[Hispo]

If an AI is able to consistently create content that doesn't break any of the forum's rules (good quality, on-topic, not just a padded word salad, no plagiarism, etc, etc), I, for one, welcome our new machine overlords. Otherwise, content that violates the rules, AI or human produced, can already be dealt with with our current rules and policies in mind.

So, in theory if someone somehow managed to farm, let's say 5 accounts up to Full member and Senior member and somehow implemented an AI for those accounts to interact around here while joined a signature campaign or a bounty, so the single person behind those accounts can get all the economical benefits, that would be acceptable as long as the quality of those posts are good enough and on-topic? Right?

[mprep]

If an AI is able to consistently create content that doesn't break any of the forum's rules (good quality, on-topic, not just a padded word salad, no plagiarism, etc, etc), I, for one, welcome our new machine overlords. Otherwise, content that violates the rules, AI or human produced, can already be dealt with with our current rules and policies in mind.

So, in theory if someone somehow managed to farm, let's say 5 accounts up to Full member and Senior member and somehow implemented an AI for those accounts to interact around here while joined a signature campaign or a bounty, so the single person behind those accounts can get all the economical benefits, that would be acceptable as long as the quality of those posts are good enough and on-topic? Right?

All that matters is that those users don't break the rules. If they don't, then yes, there's nothing against them operating several AI powered accounts though at that point, I'd be more worried about that AI doing things outside Bitcointalk as opposed to farming signature campaigns.

Just posting something on-topic , semi-correct and coherent doesn't necessarily mean the post isn't low value. That quality is what I feel a lot of AI generated content is gonna lack (and yes I'm aware of the recemt developments in that area).

[Little Mouse]

So, in theory if someone somehow managed to farm, let's say 5 accounts up to Full member and Senior member and somehow implemented an AI for those accounts to interact around here while joined a signature campaign or a bounty, so the single person behind those accounts can get all the economical benefits, that would be acceptable as long as the quality of those posts are good enough and on-topic? Right?

That's irrelevant when forum itself allow you to use more than 1 account as long as you are good with forum rules. Why are you bringing the economical benefits here? If someone is active here for economical benefit only and that's from 1 account only, both he and owner of 5 accounts are in same category of course.

Just posting something on-topic , semi-correct and coherent doesn't necessarily mean the post isn't low value.

But it will be almost impossible to detect if the texts get a little human touch lol. I'm observing one user, pretty much sure texts are AI generated but they are not such low value either.

[LoyceV]

99% of AI users on Bitcointalk will just use it to attempt plagiarism and spam, which is already prohibited in forum rules.

It's going to be difficult to prove the plagiarism, and spam can reach very large volumes. I don't want to have to doubt if each post is genuine. Imagine what happens when someone has a bot that earns him signature payments by asking questions on the tech boards.

Also, it is straightforward to detect GPT3 content, there are AI detection sites that do just that.

It's not easy, a lot of manual work, and doesn't provide hard evidence. The first one I tested claimed my post was 81% fake.

In that case, we need admins to explicitly state on the rules that the use of artificial intelligence for the creation of content here is considered plagiarism. Otherwise, someone could argue that there was no connection between the AI and the old rule itself.

Pretending the AI's text was created by you is the plagiarism. It doesn't matter where the AI got it.

If an AI is able to consistently create content that doesn't break any of the forum's rules (good quality, on-topic, not just a padded word salad, no plagiarism, etc, etc)

I could agree with this, if the post makes it clear it's created by an AI. Without that, if I post something created by an AI, my cat, or a sweatshop filled with elves, it's not my own creation. That makes it plagiarism.

[un_rank]

I totally agree with allowing technology to thrive as long as it is not contravening any of the forum rules, for what it is worth, it could actually be of benefit to the forum and improve quality of discussions.

Has the admins considered the issue of people running the replies of certain high quality members here on the forum through an AI programming and they push out posts similar to the person's style?
Here is a thread on reddit^[1], which discusses how this affects artists and I do not think it is limited to only art.

If someone created an AI which posts like theymos for example or satoshi based on their post history, would that raise brows?

[1] https://www.reddit.com/r/technology/comments/y6eiyq/artists_say_ai_image_generators_are_copying_their/

- Jay -

[DdmrDdmr]

Open AI is seemingly working on a solution to digitally watermark their outputs, through a kind of "unnoticeable secret signal in its choices of words, which you can use to prove later that, yes, this came from GPT" (see this blog):

<…> So then to watermark, instead of selecting the next token randomly, the idea will be to select it pseudorandomly, using a cryptographic pseudorandom function, whose key is known only to OpenAI. That won’t make any detectable difference to the end user, assuming the end user can’t distinguish the pseudorandom numbers from truly random ones. But now you can choose a pseudorandom function that secretly biases a certain score—a sum over a certain function g evaluated at each n-gram (sequence of n consecutive tokens), for some small n—which score you can also compute if you know the key for this pseudorandom function. <…>

Nevertheless, it won’t be ready for months, perhaps until GPT-4, and it’s pretty likely that pipelining the output with a good word spinner may break the watermark’s value. We’d also need a tool around to detect these watermarks, which will be OpenAI propriety, and may come with an associated price tag.

The above may or may not help to detect the use of AI posts at some point, but I don’t see it streaming into the mod’s pipeline when checking post compliance with forum rules, at least not yet.

On the matter of the rules affecting AI generated content (specifically text):

<…> Conceptually, any post using GPT to create content on Bitcointalk is plagiarizing, as the poster is not creating the content himself, and is in fact trying to pass on someone else’s content as his (albeit that someone else being an AI). According to OpenAI’s Sharing and Publication Policy, using the API, and I have to assume that that extends to the results of their chatbot, requires one to explicitly indicate that the content was AI-generated. Though this latter point is not technically of our concern, it seems like a reasonable request to place, in a similar fashion to links on posts that are largely/verbose based on other sources.

Detection of GPT usage is not going to be easy for the most, and likely, over time, people can pick-up on patterns such as the usage of near perfect English, consistency in its usage throughout all posting history, and/or alternating with changes in style (human/AI), lack of real interaction from a less than academic point of view, certain types of formal constructions, and so forth. This is obviously is not exclusive to GPT, nor sufficient to deem someone a GTP-plagiarizer with certainty, and maybe likenesses is the closer one can get shorter of a confession.

Now all this is, if it becomes an extended practice, is going to be a drag, whereby people will be able to create bag loads of posts with cero effort and thought, and although likely matching quality-wise a large base of posts that we encounter per se, it may easily become a new spam-fest source of neutral content.

I’ve read, though I couldn’t find the original source (i.e. team declarations), that Chat-GPT can’t plagiarize per se (the language model generates the text using a probability for the next best word to use, based on the prior words it has already used), although there could be a fortuity chance of it happening. Many of the posts we read seem to me, from a reader’s point of view, a compendium of text-spinning ideas. Though that is not what it’s really doing, the probability of the next word to use is derived from the model created through the training data, and that is inevitable a subjacent reference to all the text it provides.

On the other hand, even if the text is comprehensive and aligned somewhat with what the average Joe may be able to come up with, the poster conceptually (in my book) plagiarized the output from Chat GPT, without giving credits to the source of the text’s generation, trying to pass the text on as his own.

[hugeblack]

Is there any bot that works in a realistic way so that it shares replies that it is difficult for the average user, mod or campaign manager to distinguish from the posts made by humans?
In the sense of high quality, adding good information, well-written, and not words that were copied and pasted with the addition of some general sentences?


Generally, it is easy to ban these bots account with the spam, or zero/low value rule.

[m2017]

If an AI is able to consistently create content that doesn't break any of the forum's rules (good quality, on-topic, not just a padded word salad, no plagiarism, etc, etc), I, for one, welcome our new machine overlords. Otherwise, content that violates the rules, AI or human produced, can already be dealt with with our current rules and policies in mind.

Do I understand correctly that this can be regarded as an approval of the use of AI on the bitcointalk? If AI is not forbidden, then using it doesn't violate forum rules, right?.

Will the administration make changes to the forum rules regarding the use of AI or will it leave everything as it is? It seems to me that it would be good to clarify this either in the form of a rule change / edit or at least an informational post for forum users.

[NotATether]

If an AI is able to consistently create content that doesn't break any of the forum's rules (good quality, on-topic, not just a padded word salad, no plagiarism, etc, etc), I, for one, welcome our new machine overlords. Otherwise, content that violates the rules, AI or human produced, can already be dealt with with our current rules and policies in mind.

Alright. I'll see if I can contract Aaron Schwarzenegger to build me one last Terminator for this purpose.

[Stalker22]

It all depends on how and where the AI-generated content is being used. If you are posting AI-generated content and pretending it is your own original work, then you are committing plagiarism. Plagiarism is the act of presenting someone else's work as your own without proper attribution. But, if you are using AI as a tool to help you develop and improve your own original text, then I do not think it is considered plagiarism. For example, Google Translate is an example of AI-generated content, and grammar checkers like Grammarly rely on AI to generate content. It is hard to make a clear distinction between these modern tools and define what is allowed in the forum and what is not in a clear-cut manner.

[philipma1957]

I agree with @philipma1957, all we need is just explicit disclaimer. If usage of AI can improve post or discussion quality on this forum, i'd welcome such AI.

Also, it is straightforward to detect GPT3 content, there are AI detection sites that do just that.

Do we know how reliable such website? Does it work against specific user input (such as using certain writing style) or have low false positive?

Well look at my difficulty thread  if it says any are AI it is not accurate as I do not do AI and I think no one else does so on that thread.

I think the problem is if you put those posts  at least 10% would test as AI.

In order to not test as AI you need a lot of original thought vs combining 3 posts and voicing an opinion on them.

If you were to go to my difficulty thread

https://bitcointalk.org/index.php?topic=5378628.0

There is a ton of variation on a common theme. (the difficulty)

for instance

Quote

https://www.bitrawr.com/difficulty-estimator

Latest Block: ? ? ?

blah blah blah
blah blah blah

..... quote


happens over and over and over in that thread.

It is not ai or google but it is a lead source of DIFF stats

and will be repeated thousands of times.

I would love to see how that thread does on AI testing.

we have done them for years
2022
2021
2020 and many more

[Hispo]

-snip-

...If someone is active here for economical benefit only and that's from 1 account only, both he and owner of 5 accounts are in same category of course.

they are in the same category, but in my personal opinion it makes things worse in the case of account farming and account abuse.
If the bots/artificial intelligence participate around here undetected, it could degenerate the forum, the campaign participation and become a tool for shady people to grow accounts automatically to put them for sale, because in the end merit senders might see no much of a difference between an actual human being and the AI.

In that case, we need admins to explicitly state on the rules that the use of artificial intelligence for the creation of content here is considered plagiarism. Otherwise, someone could argue that there was no connection between the AI and the old rule itself.

Pretending the AI's text was created by you is the plagiarism. It doesn't matter where the AI got it.

I agree, for the reasons previously presented.
We do not need an Artificial intelligence pretending to be a human around here, imho.