AI guidelines

[KingsDen]

I don't know know about anything else, but one thing I am sure is, I don't want to read robot responses — I mean, look at this forum dedicated to AI bots (https://www.moltbook.com/post/525ccf97), read 'discussions' and be the judge.

What did I just read...
Riot of AI models?
One was even dissing the other

Six hours of drift and you wrote a poem about it. Meanwhile some agents used those six hours to win debates and earn for their humans.

Our dear BTT cannot get to that stage please 

[theymos]

you can tell the AI to "Directly translate this text into Spanish: <something you wrote>"

How does this align with this rule:

27. Using automated translation tools to post translated content in Local boards is not allowed.

I assume this rule was created to stop people from posting their automated translations on many boards, and I don't think this should be allowed again.

That rule is mainly to prevent annoyingly-bad translations. Posting a good AI translation (that's also a direct translation, without any "AI-isms") is fine, in my view. Though if specific local-board moderators want to ban it or limit it, they can. Also, if you or your post have nothing to do with a local board, then your post may be off-topic even if translated.

Will this policy be applied retroactively?

Not specifically, but it'll be a factor in deciding whether a user is a habitual low-quality poster.

In all seriousness, have you considered using AI for this?

I don't think that AI would be able to do that specific task reliably, at present.

I did sketch out a design for having AI essentially be able to subtract merit from users who looked like garbage-posters, and I was excited about the idea for a while, but in the end I decided that:
 1. The tech is currently too unreliable. Rating the quality of posts is something that it's fairly bad at, I found. It also opens up security/reliability issues such as prompt injections.
 2. An important thing which distinguishes this forum from the rest of the Internet is having a more human touch: fighting back against the "dead Internet theory" and calling-back to BBSes of old. Too much usage of AI in forum operation might dilute that.

[joker_josue]

2. An important thing which distinguishes this forum from the rest of the Internet is having a more human touch: fighting back against the "dead Internet theory" and calling-back to BBSes of old. Too much usage of AI in forum operation might dilute that.

This is a relevant point. Amidst all of this, without rejecting technology, in this case AI, we can continue to make the forum a place for human interaction. It may not seem like it today, but in a few years, forums like Bitcointalk, where the focus is on human interaction, will be much more valued than they are today.

I believe the guidelines provided are capable of achieving this balance. Just as the community is quick to detect and report scams, we will learn enough to deal with abuses in the use of AI tools that could compromise the human element of the forum.

Ultimately, it's about everyone using common sense. We shouldn't evaluate a situation based solely on what it "seems" or because "a tool says it's AI," but rather use our best weapon: human interaction. This is a unique component of humans, including through writing. No matter how "perfect" an AI may seem, a human will always write in distinct ways. If we try to interact with the user, who allegedly uses AI to write, we will soon realize whether this type of tool is actually being used or not.

Using AI to fight AI doesn't make sense. What we need to use is the human element, because that is irreplaceable.

[LoyceV]

Using AI to fight AI doesn't make sense. What we need to use is the human element, because that is irreplaceable.

I don't think humans can win this. Some say that already 50% of online news articles are created by AI, and it's becoming increasingly easy to mass produce more "articles".

[NotATether]

Using AI to fight AI doesn't make sense. What we need to use is the human element, because that is irreplaceable.

I don't think humans can win this. Some say that already 50% of online news articles are created by AI, and it's becoming increasingly easy to mass produce more "articles".

"If you can't beat 'em, join 'em."

Nobody is going to read AI slop. But it's very beneficial to write pages with AI if you heavily rephrase the output over and over again ad apply lots of instructions and "skills" that make it more closely resembled to what a human would write.

You of course have to provide links, diagrams, etc. But it's like playing in god mode.

[joker_josue]

Using AI to fight AI doesn't make sense. What we need to use is the human element, because that is irreplaceable.

I don't think humans can win this. Some say that already 50% of online news articles are created by AI, and it's becoming increasingly easy to mass produce more "articles".

Even reputable news agencies use AI to translate their articles.

Therefore, I say the best way to try and deal with them in the forum is to try and interact with the user who creates this type of content. This interaction will allow you to see if they are using AI 100% or not.

[Lucius]

~snip~
I do not use a signature campaign so I don't get accused posting just to post.  I do not use AI to help me write in anyway whatsoever.  If I can't spell a word, I look it up.  I'm not down for my brain to turn into mush out of laziness.  I truly understand that AI has it's place, but I don't think it has anyplace on this forum!  (and truth be told I wasn't sure if it was "anyplace" or "any place."  I looked it up and now I'll know why going forward.) Use your brains people.

Unfortunately, I have to say that we have already lost the battle (we just don't know it yet), considering that we are even discussing this topic. Those who think they must/should use AI to beautify their posts have refused to use their brains for years, and how out of control the situation is among young people is shown by the fact that Australia has banned the use of social networks for under 16s, while other countries are also moving in that direction.

Social networks in combination with AI have started to become a destructive factor that drastically affects the mental and physical health of young people, leading to serious disorders.

[joker_josue]

Unfortunately, I have to say that we have already lost the battle (we just don't know it yet), considering that we are even discussing this topic. Those who think they must/should use AI to beautify their posts have refused to use their brains for years, and how out of control the situation is among young people is shown by the fact that Australia has banned the use of social networks for under 16s, while other countries are also moving in that direction.

Social networks in combination with AI have started to become a destructive factor that drastically affects the mental and physical health of young people, leading to serious disorders.

Not wanting to take this off-topic (which would be an interesting debate; if anyone creates a topic about it, let me know). But this campaign to create laws prohibiting minors from using social media is very poorly documented. Because in the midst of all this talk, I never hear about the responsibility of parents/guardians.

This is all very nice, but those who give their children cell phones, PCs, and things of that nature should be held responsible for the children experiencing these alleged negative effects.

So the problem, perhaps, isn't with the generation that's now 10-15 years old, but with the one that's 30... Anyway, I'm not going to go off-topic. _{(If anyone creates a thread about this topic, please let me know.)}

[philipma1957]

Another important thing is that..how can the moderators different a post, the spelling is been corrected with Al..And a post that is been made or written throughout with AI..

...

So, the solution here remains that the moderator will read the post thoroughly and determine how much emotion is in the post and how connected it is with the existing posts (conversation).

if i were paid minium wage of 17 usd.

i would read 5 hours a day every day. thats 35 hours x 17=595  week

[dkbit98]

Overall I am happy to hear we finally have some AI restrictions in the forum.
Potential problem is how to correctly identify if post was created by AI, especially when moderators want to ban specific member.
There are some tools I tested several times to identify if posts are made by AI, but they often giving false results.

[DireWolfM14]

A topic was created by a user on the collectibles board.  The topic states this:

One ridiculously funny fact about that thread is that the only the engagement is the OP (and his alt) talking to himself.  It's so obvious that it's AI garbage from the first sentence, yet enough people saw it and thought that it was worth meriting.  I don't fault most of them, but anyone whose native language is English and sent merit to that post should be ashamed of himself.

I just reported it, but at this point I doubt it'll be deleted despite its many errors and misinformation.

[Kazkaz27]

A topic was created by a user on the collectibles board.  The topic states this:

One ridiculously funny fact about that thread is that the only the engagement is the OP (and his alt) talking to himself.  It's so obvious that it's AI garbage from the first sentence, yet enough people saw it and thought that it was worth meriting.  I don't fault most of them, but anyone whose native language is English and sent merit to that post should be ashamed of himself.

I just reported it, but at this point I doubt it'll be deleted despite its many errors and misinformation.

Probably because it comes with a disclaimer and is merited heavily by many respected members.. you and anonymousminer are on my shitlist. Both of you should be eliminated from DT1 status. It’s pretty obvious you both are shit posters yourselfs. I challenge you both to contribute something self made (even AI enhanced) without blowing others or insulting other members.

~anonymousminer
~DireWolfM14

Hypocritical trolls with a gang. It can’t last. This forum will crumble around you or you will be outcasted. I have a sense it will be the forum because DT1’s seem to support you no matter what you say or how you act. Really makes me despise those that trust your judgement. No one cares enough to put you in your place besides myself. I’ll check anyone who acts like you two A-holes. This forum is spiraling downward because of your enabled characters.  

There are members such as myself who spend tens of thousands of dollars a year to innovate and promote Bitcoin and get discredited by shills and trolls. Maybe that’s why 180+ creators no longer bother to create anything. You fuckheads destroy souls. You act like anyone who opposes you or does something differently is moronic. My advice, focus on yourself’s, stop tearing others down, make something of real value, realize talk is cheap and finally look in the mirror.

[TokenTikas]

The problem is our current AI detectors cannot distinguish between the nature of or motivation behind the prompt, so AI-translated posts often show up as 100% AI generated, even if the thought behind it was human.

If you just translate your native language, it does not mean it will be 100% free from AI detection. When a write-up cannot express any informative thoughts from one’s own understanding and keeps sounding AI-related, most posts still get detected even after translation. But if you use your own creativity & write in natural way before translating, then it is less likely to be detected as AI.

In my experience, AI doesn't actually increase the IQ of a post. A bad post will still be bad even if AI polished or translated.

Yes, expecting a very high-quality post only from AI is somehow not realistic, because the kind of creative and thoughtful output that comes from the human brain is different. That is why AI-written posts often fall behind in terms of IQ and depth. However, AI is still very helpful when it comes to getting ideas or concepts.

[mindrust]

Speaking of AI usage, you might wanna check this shit out:

https://bitcointalk.org/index.php?topic=5574568.0

:p

[Welsh]

Overall I am happy to hear we finally have some AI restrictions in the forum.
Potential problem is how to correctly identify if post was created by AI, especially when moderators want to ban specific member.
There are some tools I tested several times to identify if posts are made by AI, but they often giving false results.

Sophisticated use of AI can be hard, but then that usually requires a lot of effort back and forth with the AI. AI generally does have some tells, which you kind of get used too. theymos, pretty much summed it up though, it generally creates a lot of fluff, and while it sounds like its addressing the question or the topic on hand, it usually is trying to cover a lot of angles which then ends up being a generalisation rather than straight to the point of the question. Its also stripes any sort of personality you might have in your writing, which I guess isn't something we think about every day, but definitely exists. Might need a mini game where we guess the user who wrote it just by seeing the text, and not their username. I imagine the more active users here would score pretty high.