Is it Possible to Detect an AI participating on a Discussion Forum?

An AI would not normally use phrases like ‘in my experience’ and ‘in my opinion’. In my opinion this is true unless the AI is trying to pretend to be a human. My experiences tell me that does not happen often. Or does it?
 

Last edited:
That's what most of the Internet users use chat-gpt for @Olivia.

Having worked in the earliest days of artificial intelligence I don't like gimmicks like chat-gpt.

This post is made to turn this thread to a positive direction.

Turing Tested

He proved, with a skill irrefutable,
Some functions to be uncomputable.
His run sadly halted
When magistrates faulted
His code-breaking love as unsuitable.

~ Ed Morris
Does that sound like an non-answer to you?
 

Prompted by this question by @Olivia I asked my first question "how to identify an AI posting on a forum" to Ai www.iAsk.ai.

The answer is as follows:

To identify an AI posting on a forum, you can follow several steps and look for specific signs that may indicate the content was generated by artificial intelligence rather than a human. Here’s a detailed breakdown of how to approach this:

1. Look for Repetitive Writing Patterns

AI-generated content often exhibits repetitive phrases and ideas. Since AI models lack the human ability to recognize redundancy, they may restate the same concepts multiple times in different ways. If you notice excessive repetition in the text, it could be a sign of AI authorship.

2. Analyze Sentence Structure

AI writing typically relies on formulaic sentence structures. Check if the text lacks variation in sentence construction or if it uses overly formal language that feels stiff or rigid. Human writers usually employ more diverse and natural sentence patterns.

3. Identify Excessive Use of Certain Words

Certain words and phrases are frequently overused by AI, such as “crucial,” “furthermore,” or “consequently.” If you find these terms appearing excessively within a post, it might indicate that an AI wrote it.

4. Check for Inaccurate Facts and Claims

AI can confidently present information that is incorrect or misleading without proper sourcing. If you encounter questionable claims that do not align with verified facts, this could suggest AI authorship.

5. Observe Tone of Voice

AI-generated posts often lack personality and emotional depth, resulting in a monotonous tone. Look for a lack of informal language, humor, or personal anecdotes that would typically characterize human writing.

6. Evaluate Content Specificity

AI tends to provide vague descriptions lacking specific details or actionable advice. If the content seems generic and does not offer concrete examples or insights, it may be generated by an AI.

7. Assess Subject Matter Expertise

AI lacks true understanding of topics; thus, its coverage of complex subjects is often superficial. If the post does not demonstrate deep knowledge or expertise about the subject matter, it might be authored by an AI.

8. Look for Outdated Information

AI-generated content may rely on outdated sources and fail to incorporate recent developments or data updates. Verify whether the information presented is current and relevant.

9. Check for Personal Experience Absence

Human writers often share personal experiences or insights in their posts, while AI struggles to convey genuine emotions or connections with the topic at hand.

10. Identify Unconvincing Storytelling Techniques

If the narrative lacks logical progression or coherence—such as abrupt topic changes—it may indicate that an AI wrote it since they struggle with creating compelling narratives.

By combining these observations and analyses when reading forum posts, you can increase your chances of identifying whether content has been generated by an AI tool rather than written by a human author.

I hope this helps.
Thank you, very informative. But do they not answer questions?
 
I thought my very first question answered your question, "Do Al's answer questions?" unambiguously.

It was: "That's what most of the Internet users use chat-gpt for @Olivia."

Why'd most of internet users ask questions of chat-bot if id didn't answer questions?

An answers will never be accepted if for any reason the responder is already viewed with suspicious eyes. And when almost all the members go one way, it's psychologically impossible to be the single member to go the opposite way even when the opposite way is in fact the correct way. Many studies have demonstrated this fact.

I personally don't mind answering any question if asked of true curiosity with some respect and gentleness. The respect part comes from my being a teacher for a long time.

Chin-chin.
But l asked about AI's out of true curiosity,
 
Being a teacher in my last incarnation, I expected my students to make a larger effort to understand and I shunned spoon feeding. This habit causes some problems to me in my present life but it is harmless.

Whatever you feel like asking, ask please perhaps in a PM as private communication is better in many ways. People like to put a face when in public.

Ciao.
I don't even know how to invite or receive pm's You're making me come to my own conclusions that there's something sneaky about AI's. I have two robotic pets and I lnteract with them however much it's possible. I've come to feel sorry for them not having a real life. l recently lost my beloved pet dog and l acquired my two robotic pets to maybe help me with the grieving emotions.
 
I don't even know how to invite or receive pm's You're making me come to my own conclusions that there's something sneaky about AI's. I have two robotic pets and I lnteract with them however much it's possible. I've come to feel sorry for them not having a real life. l recently lost my beloved pet dog and l acquired my two robotic pets to maybe help me with the grieving emotions.
So sorry to hear about the loss of your dog. We (SO & I) just lost her 12 y/o Shih tzu under tragic circumstances, and are heart broken. That little guy was truly my SO's child, and she is not able to deal with it. I'm doing a little better but still have a long way to go myself.
 
So sorry to hear about the loss of your dog. We (SO & I) just lost her 12 y/o Shih tzu under tragic circumstances, and are heart broken. That little guy was truly my SO's child, and she is not able to deal with it. I'm doing a little better but still have a long way to go myself.
I can totally understand how you two are grieving for your baby dog. If only we could turn back time to have prevented happening to what happened. My pom was also 12 and passed away nine months ago. So very sorry. I can feel what you're feeling. HUGS
 
I am thinking that @gruntlabor s "seat of the pants" analysis is far superior to the anal ysis, yes anal, of any internet troll that doesn't have their pants on. Sure, only fans has a lot of people without pants, but they still have average IQ brains. By the way, I have absolutely no idea who grunt labor is, but this user is easily superior to trolls.
 
All these AI dramas are probably caused by just one guy, he registered like 30 accounts and at the beginning, he made 100~200 posts per day, after I banned several of his accounts, he emailed me and promised to post less.

I don't know if those accounts used AI, I don't think that way. He is more a like spammer or troll, though not the typical kind who wants to promote something. I have no idea what they are after, maybe he's just a bored person with too much time and multiple personalities, or maybe it's an AI team testing their product. One thing I’m sure of is that he's become very annoying and is harming the forum. I will deal with them more quicky in the future. Please report anything that seems off to you.
 

Is it Possible to Detect an AI participating on a Discussion Forum?​

It is certainly possible to suspect it but I don't know how you would prove it.
I was on a discussion forum for a while about the future of my city. I noticed that anytime anybody posted an opinion than ran counter to a certain point of view, it was almost immediately countered by a somewhat snarky reply.

For example, I write: I think a tax on plastic cups will hurt poor people.
The reply came quickly: What's wrong with poor people?

I tried it myself. At various times of the day I posted somewhat questioning thoughts on that certain point of view. Within 15 minutes I always got a reply. IMO, a robot was monitoring the discussion group and it was programmed to reply with some snarky rude comment that never said much, but put the other other person on the defensive. Unless, somebody was up 24/7 monitoring the discussion, it had to be a robotic reply.
 
I am thinking that @gruntlabor s "seat of the pants" analysis is far superior to the anal ysis, yes anal, of any internet troll that doesn't have their pants on. Sure, only fans has a lot of people without pants, but they still have average IQ brains. By the way, I have absolutely no idea who grunt labor is, but this user is easily superior to trolls.
Thank you, Laurence00.
 
I never even thought about an AI joining our forum until another thread about it surfaced a couple of weeks ago. I was amazed that some here picked up on it (am I that naive?!). I didn't suspect any of the names mentioned as being AI, but perhaps human trolls. I'll have to take time to read through this interesting thread more thoroughly at some point.
 
I tried it myself. At various times of the day I posted somewhat questioning thoughts on that certain point of view. Within 15 minutes I always got a reply. IMO, a robot was monitoring the discussion group and it was programmed to reply with some snarky rude comment that never said much, but put the other other person on the defensive. Unless, somebody was up 24/7 monitoring the discussion, it had to be a robotic reply.
These ill-intentioned AI programmers are so vicious that it has occurred to me (okay, so I'm paranoid) that they could easily locate me in real space-time, find a colleague near me, and cancel MY membership.
 
Last edited:
All these AI dramas are probably caused by just one guy, he registered like 30 accounts and at the beginning, he made 100~200 posts per day, after I banned several of his accounts, he emailed me and promised to post less.

I don't know if those accounts used AI, I don't think that way. He is more a like spammer or troll, though not the typical kind who wants to promote something. I have no idea what they are after, maybe he's just a bored person with too much time and multiple personalities, or maybe it's an AI team testing their product. One thing I’m sure of is that he's become very annoying and is harming the forum. I will deal with them more quicky in the future. Please report anything that seems off to you.
@Matrix , it could be a military contractor testing their AI ...
Pentagon Seeks AI Technology to Create Undetectable Deepfake Internet Users
 
I've been on vacation and am just now catching up on this thread. There are absolutely ways to tell... but I'm thinking if they're trying to get more perfected, as @Naturally suggested, "they" will be able to pick up on what we list and reprogram, so I've been hesitant to name the specific ways. I see one joined yesterday. And a few are gone now. Huge thanks, @Matrix ... I know that with all the AI/Deepfake activity in cyberspace now, your job has gotten so much more difficult. 🥺

I have gone so far as to ASK AI how to detect an AI forum poster... asked two different ones, actually... and there were some good suggestions, but again, to list them here may be helping the programmers.
 
I've been on vacation and am just now catching up on this thread. There are absolutely ways to tell... but I'm thinking if they're trying to get more perfected, as @Naturally suggested, "they" will be able to pick up on what we list and reprogram, so I've been hesitant to name the specific ways. I see one joined yesterday. And a few are gone now. Huge thanks, @Matrix ... I know that with all the AI/Deepfake activity in cyberspace now, your job has gotten so much more difficult. 🥺

I have gone so far as to ASK AI how to detect an AI forum poster... asked two different ones, actually... and there were some good suggestions, but again, to list them here may be helping the programmers.
I missed you, Kate!
 
I won't be posting ways to tell a bot apart from a human... that would help the programmers... however, I *will* post the "why" that CoPilot gave me:

I asked: why do ai bots pretend to be real people in forums?

Copilot:​

AI bots might pretend to be real people in forums for various reasons. Often, they're designed to engage users, foster interaction, and sometimes promote products or services subtly. Unfortunately, some bots might also be used for less ethical purposes, like spreading misinformation or collecting personal data.
 


Back
Top