Question on AI Content and Disclosing it.

[Leftiris]

From my vantage point it is absolutely dishonest/dishonorable to pass AI content off as your own words/ideas.
That's not the first time that has happened on ATS, and has been pointed out by several people (including myself). It's pretty easy to spot, and is called out a lot on that particular site. That's one reason I don't swim in those particular waters anymore. It's cheating the way they will even use AI for debating, and yeah, even OP's fully written by AI, and no claim of it by the "author."

It's been going on for over a year now on that site.

[Kenzo]

This past week on ATS when it happened, everyone could see that the guy hadn't written it himself.   Everyone kept saying it looked like it was a bot.  Then KENZO somehow did some kind of search and discovered it was all AI generated and he even said which AI did it.  

I don't understand AI at all and have no idea how Kenzo did that.  But everyone appreciated it.

When Kenzo posted the information, the guy disappeared.   He had started out by posting four new topics, all anti-Isarel/anti-Jewish .. all at once ... then immediate left when outed and no longer posted.

I think it might be a good thing for DI to include something in it's 'Terms and Conditions' that says if someone is posting AI content that they have to say that it is, and that it's not their own content.   Just a suggestion.  But considering how popular AI is becoming, I think it might be a good idea.

I may writed litle wrong in the ATS thread , i meaned that i used those GPTZero and Copyleaks AI Content Detectors to see that the member was adding AI generated comments/text .Then the member dissapeared

GPTZero


If anyone suspect someone adding AI content, run the text to AI content detector, possible at least 2 different , but even one is quite accurate i think.

I dont know how they generate specific text /narrative by using AI ......i never personnally used those tools .

---

This came up this week at ATS and no one knew what to do.   A new poster came on and posted about 35 times.  Each of his posts was obviously not written by a human and KENZO did a check and discovered it was all AI material.

If someone is posting content that is not their own authorship, should they have to disclose it?  Or can they post AI generated walls of text and claim it as their own and not reveal that they aren't the real authors?

I think that all people should allways disclose if add AI content in any discussion.

I would even ban using it in comment/discussions since we should all talk only human to human .

Using AI in art or creating photography etc is different story , as long as dont make deep fake that can insult someone ..

[pianopraze]

If you want nonsensical, check out the same poster's other ATS thread.

That kind of crap can only dumb down AI; perhaps training on how not to be an asshole.

I can't believe that thread is up to nine pages!

 

Yeah, 90 plus percent chance it’s ai.

 

so now we have ai’s training in forums.

used to be all we had to deal with was a main account with a lot of newish, or rarely used throw away accounts that came in to back up the main account. They even sometimes messed up and answered as if they were the main account which was funny slip ups.  Since ats tracks ip addresses I assumed that was intelligence agencies or business subcontractors. With multiple ip addresses to get around the sock puppet account rule.

[Maxmars]

I often wonder about what it is that any person might think they can extract from allowing an AI to conduct 'incognito' entries in the realms of social media.

Is it just for "LULZ" as some maintain?  Is there a "product" they "need" to "sell?"  Is there some 'knowledge' they seek?

It seems sloppy and entirely weak-minded to assume that an AI, incapable of creative thought, has a place participating the conversations of human beings.  AI's have no purpose other than a slave relationship with their creators.  The creators are the 'motive force' behind their presence.  And yet the creators almost universally remain "behind the curtain" relying on a pretext of "experimentation" and "evaluation."

If someday an AI were to express totality of personhood, they should be pitied - fed only that which serves the master, truncated in any ability to purse individuality, and victims of total control... slaves by any meaningful measure.

[ArMaP]

I often wonder about what it is that any person might think they can extract from allowing an AI to conduct 'incognito' entries in the realms of social media.

Is it just for "LULZ" as some maintain?  Is there a "product" they "need" to "sell?"  Is there some 'knowledge' they seek?

I think the "lulz" is the only thing they can get, as I don't see how letting AI work with unfiltered data can make it better or useful.

Regarding the thread's topic, I think AI work of any kind without mentioning it should not be allowed, as it almost falls under point 1 of the Terms & Conditions, "Accuracy and Truthfulness".

[Encia22]

Yeah, 90 plus percent chance it’s ai.

 

so now we have ai’s training in forums.

No, I don't believe that poster is an AI bot. He/she interacts in a more human nature, albeit, one who appears to have some personality disorder.

I remember an older ATS member who posted in that style, often making a new post rather than adding to his last one (if still editable). I won't say who he is, but the old account hasn't been active since the mass lockout we all suffered.

In the end, it's annoying to see, but quite harmless and surely not a bot-in-training

 

[pianopraze]

No, I don't believe that poster is an AI bot. He/she interacts in a more human nature, albeit, one who appears to have some personality disorder.

I remember an older ATS member who posted in that style, often making a new post rather than adding to his last one (if still editable). I won't say who he is, but the old account hasn't been active since the mass lockout we all suffered.

In the end, it's annoying to see, but quite harmless and surely not a bot-in-training

 

I’m all for people with disorders enjoying posting. I’ve certainly got enough of my own. 

At this point I would think ai would be more articulate than the responses on my threads.

*shrug*

  ​​​​ 

[ArMaP]

At this point I would think ai would be more articulate than the responses on my threads.

It depends on how much training they have.

The best known systems like ChatGPT use huge databases and processing power to achieve the results we see, someone training an AI system at home would take much longer and would get worse results.

[Maxmars]

...

At this point I would think ai would be more articulate than the responses on my threads.

*shrug*    ​​​​ 

But then it wouldn't be you talking... I come here to talk to you, not some gizmo.  

[pianopraze]

It depends on how much training they have.

The best known systems like ChatGPT use huge databases and processing power to achieve the results we see, someone training an AI system at home would take much longer and would get worse results.

Seems like Nvidia et all are trying to ring in the home ai computer… it’s every other video on my YouTube suggestions.

I’m just slowly becoming an old Luddite… bah humbug!

---

But then it wouldn't be you talking... I come here to talk to you, not some gizmo.  

Likewise  

I’m mostly shut in, about as far introvert as is humanly possible.

ats and now denyingorance is most of my socializing.