Should a member be allowed to make a post which is AI generated or AI mixed without disclosing such use of AI as part of the post?

[Ron Resnick]

*a significant number of posts on What’s Best Forum exhibit at least moderate probability of AI involvement. This is going to be a lot of fun?

This is why we are discussing these issues in a leisurely and collective way. Before a member is accused of an AI post we had better be damn sure that a forum-agreed detector, or a composite detector like Monica, rates it as 100% probability AI.

Better 10 AI post violators go free, than one original writer is falsely accused.

 

[Al M.]

Better 10 AI post violators go free, than one original writer is falsely accused.

Exactly. Bravo.

 

[facten]

Ron, I did. And I think this is ridiculous. I don’t think you’re writing your posts with AI. But I took screenshots of ChatGPT assessing a moderate probability of AI generation to your post.

The one that was questioned that you posted was deemed 100 percent AI , what's your response to that? It was your post rebutting a technician point that Dave made

 

[Robh3606]

My personal thinking about AI on WBF has evolved over the course of this discussion. Originally I did not see a problem with a member drafting an original post and then running it through AI for processing. Now I am skeptical even of this use of AI, because I think increasing adoption by members of this "back-end" use of AI will tend to homogenize posts and tend to suppress the individual personalities and writing styles of our members who use AI to correct grammar and to "smooth" their posts in this way. I personally do not care to see any posts which read fake or kind of fake.

Group Think

Rob

 

[Ted Denney III]

This surprises me. I would not think that we have presently a high instance of AI polished posts.

Would you please copy and paste here an example or two of the posts you're talking about?

Sorry Ron, there was a screenshot missing.

 

[Ron Resnick]

The one that was questioned that you posted was deemed 100 percent AI , what's your response to that?

My response is that I'm extremely flattered.

 

[Ron Resnick]

Yes, I believe that AI detectors are absolute raving bullshit.

There is no way to enforce an anti-AI rule with AI detectors.

What do you suggest we do, Al?

(I don't want to read posts that sound like robots.)

 

[facten]

My response is that I'm extremely flattered.

What are you flattered about? The post was on a technical matter not made by you

 

[Ron Resnick]

moderate to high probability of being written by AI.

I'm confused, Ted. Chat is saying that these posts have indicia of AI, but it is not evaluating them as an AI detector and saying 100% probability of AI.

 

[Ron Resnick]

What are you flattered about? The post was on a technical matter not made by you

My mistake. I thought Ted was running my actual posts through his AI detector.

 

[Al M.]

What do you suggest we do, Al?

(I don't want to read posts that sound like robots.)

Nothing. You can't enforce the rule with lousy AI detectors.

I don't think it's an issue, frankly. How many posts have you seen that sound like robots?

 

[Ted Denney III]

My mistake. I thought Ted was running my actual posts through his AI detector.

I did, and it came out low to moderate probability that it was written by AI.

 

[andromedaaudio]

Nothing. You can't enforce the rule with lousy AI detectors.

I don't think it's an issue, frankly. How many posts have you seen that sound like robots?

Just wait untill AI finds its way into the judiciary system ..... Lawsuits

 

[Ron Resnick]

Nothing. You can't enforce the rule with lousy AI detectors.

I don't think it's an issue, frankly. How many posts have you seen that sound like robots?

I'm perfectly willing to consider that this whole effort is a solution in search of a problem. Maybe we don't need, at this time anyway, any AI Term of Service at all.

The question is the accuracy of AI detectors. If the membership doesn't have confidence in the detector selected, then the Term of Service is doomed to false accusations and contentiousness.

From the little testing I've done that Monica, which is a composite of multiple AI detectors, seems pretty good.

 

[Ron Resnick]

Take a look at the copy I wrote this morning for our webpage. Some AI detectors show a 100% probability that it was generated by AI. I wrote it myself.

But did you run your original writing through AI for spelling, grammar and polishing?

 

[Al M.]

I'm perfectly willing to consider that this whole effort is a solution in search of a problem.

That's quite possible.

Maybe we don't need, at this time anyway, any AI Term of Service at all.

Well, I guess you do need special Al (M.) Terms of Service .

(My name still keeps coming up all the time, creepy )

 

[Ron Resnick]

take a look at this copy I wrote myself, this morning. It shows a 100% probability of being AI generated, depending on which AI detection tool is used. Except one problem, I wrote it myself.

I understand you wrote it yourself originally. But didn't you run your original writing through AI for spelling, grammar and polishing?

 

[Ted Denney III]

But did you run your original writing through AI for spelling, grammar and polishing?

I run spelling and grammar check on everything. I also don’t use a typewriter anymore. That’s not the same thing as inputting parameters, and then asking AI to punch out boiler plate. It’s a spellcheck and grammar check.

The above was written by me, talking to my phone. Unedited not grammar or spellcheck. Now this is how I would normally proceed after verbal dictation and doing a grammar and spellcheck…

I run spelling and grammar checks on everything. I also don’t use a typewriter anymore. That’s not the same thing as inputting parameters and then asking AI to punch out boilerplate. It’s a spell check and grammar check.

 

[Ron Resnick]

Somehow whatever back-end AI checking you are doing makes the text sound too perfect and kind of fake and annoying to read.

Hobbyists (including me) don't want to read robot-sounding text. It's as simple as that.

 

[italia]

I'm perfectly willing to consider that this whole effort is a solution in search of a problem.

In Post #41 of this thread I shared this consideration as well, by stating "Seems to me this is a unrealistic solution in search of an inevitably uncontrollable problem." The statement evidently didn't get much traction at the time. Having continued to follow this thread through an additional 300+ posts, my conviction of the futility of this pursuit has only strengthened