People posting AI output verbatim?

[Hakeem]

Call me a Luddite but I’m not interested in seeing someone copy & paste something that some AI program generated. I understand people are posting it in good faith but I want to hear what the human members here have to say. Making matters worse - people don’t always point out that it’s AI.

Would there be any interest in “guideline” that encourages human contributions and discourages AI contributions?

 

[Tchicken]

Second - but the devil would be in the enforcement. I "think" that I'm pretty good at spotting AI or morons, but they're often interchangeable. Maybe call BS on them when you spot them?

 

[AldeanFan]

I was helping a friend replace the clutch in his old Plymouth.
We needed to know how much fluid it took.
The AI generated response was 3 pints,
It actually takes 3-1/2 quarts.

AI needs to be thoroughly reviewed and double checked before relying on it.

 

[Snaparxon]

Ai

 

[Hakeem]

Second - but the devil would be in the enforcement. I "think" that I'm pretty good at spotting AI or morons, but they're often interchangeable. Maybe call BS on them when you spot them?

Fair point, and I’m not expecting any hardline enforcement. If it were up to me, I’d post an official statement along the lines of “Please keep your contributions original and avoid posting AI generated output” and leave it to the community to self-regulate.

I believe people that post it genuinely think they are helping but I feel it’s antithetical to the spirit of a discussion board. I’d love to hear what the moderators & admins think about it.

 

[kbeefy]

I find it annoying.
These days anyone can get their own AI input. I use it for my own reference.

If I'm posting a question or looking for experience/advice, I'm looking for real experience. Not regurgitated AI babble.

 

[RalphInCA]

Much “human created” content here on GJ is ****, too.

And we’ve been able to (somewhat) control that as a community for a long time.

How is AI created content different from this?

 

[CoogarXR]

A tale as old as time... Dudes wanting to provide an answer even though they really don't know. I try my best to only answer topics that I actually know. But, I confess, I have googled some stuff and posted the answer. But always in my own words. I gotta at least pretend like I'm smart, lol.

 

[Tundra1]

This is thread so tempting to upload to chatgpt and post it's response. Verbatim hehehe
Oh the irony would be thick on that one....

But i agree ai posts are about as irritating as all the obviously ai generated web sites now that pretend to be informative but just regugitate what you already know.

 

[LXCam]

I’m with you Hakeem and perfect timing too. A couple weeks back there was a thread that asked AI the most asinine question I could ever imagine. And since then I’ve seen a half dozen more answers posted up that were real WTF.

This morning we probably saw the same post and I damn near started a similar thread.

AI will impress me the day it can pick up a shovel or a wrench as far as this forum goes.

 

[whateg01]

This is thread so tempting to upload to chatgpt and post it's response.

I thought the same thing. But I didn't stop myself!

That’s a really good question — and it’s actually being debated a lot right now. Whether AI responses should be allowed in forum discussions depends on what kind of forum it is and what purpose it serves. Here are some of the main arguments on both sides:


---

Reasons to allow AI responses

1. Helpfulness and efficiency:
AI can provide quick, accurate, and well-organized answers, especially for technical or factual questions (like coding, troubleshooting, or history).


2. Accessibility:
It can make discussions more inclusive by helping people who struggle with writing, translation, or explaining complex ideas clearly.


3. Idea generation:
AI can contribute to brainstorming or help users refine their thoughts before humans weigh in.


4. 24/7 availability:
Forums benefit from having responses even when no human experts are online.




---

Reasons to limit or ban AI responses

1. Authenticity and trust:
Forums are meant for human discussion. If people start posting AI-generated replies without labeling them, it can feel deceptive or hollow.


2. Misinformation risk:
AI can sound confident even when it’s wrong. Without human oversight, this can spread bad info fast.


3. Erosion of community:
If AIs flood threads, users might post less — it stops being a social exchange and becomes a knowledge dump.


4. Accountability:
Who is responsible for what the AI says? If a response is harmful or misleading, it’s hard to assign blame or moderation.




---

Middle-ground approach

Many forums are starting to allow AI-generated content only if it’s clearly labeled (e.g., “Generated with ChatGPT”) and used as a supplement, not a replacement for human discussion.
That keeps transparency while still letting AI add value.


---

Would you like me to outline what a good forum policy on AI posts might
look like (for example, how to moderate or label them)?

 

[Hakeem]

I thought the same thing. But I didn't stop myself!

That’s a really good question — and it’s actually being debated a lot right now. Whether AI responses should be allowed in forum discussions depends on what kind of forum it is and what purpose it serves. Here are some of the main arguments on both sides:


---

Reasons to allow AI responses

1. Helpfulness and efficiency:
AI can provide quick, accurate, and well-organized answers, especially for technical or factual questions (like coding, troubleshooting, or history).


2. Accessibility:
It can make discussions more inclusive by helping people who struggle with writing, translation, or explaining complex ideas clearly.


3. Idea generation:
AI can contribute to brainstorming or help users refine their thoughts before humans weigh in.


4. 24/7 availability:
Forums benefit from having responses even when no human experts are online.




---

Reasons to limit or ban AI responses

1. Authenticity and trust:
Forums are meant for human discussion. If people start posting AI-generated replies without labeling them, it can feel deceptive or hollow.


2. Misinformation risk:
AI can sound confident even when it’s wrong. Without human oversight, this can spread bad info fast.


3. Erosion of community:
If AIs flood threads, users might post less — it stops being a social exchange and becomes a knowledge dump.


4. Accountability:
Who is responsible for what the AI says? If a response is harmful or misleading, it’s hard to assign blame or moderation.




---

Middle-ground approach

Many forums are starting to allow AI-generated content only if it’s clearly labeled (e.g., “Generated with ChatGPT”) and used as a supplement, not a replacement for human discussion.
That keeps transparency while still letting AI add value.


---

Would you like me to outline what a good forum policy on AI posts might
look like (for example, how to moderate or label them)?

 

[Stelzer]

I think it's a fair point the OP brings up. I think a more realistic compromise is to encourage members to state at the beginning of their post when using AI so readers who aren't interested can skip, while others can read and decipher for themselves as to whether or not the content is helpful.

 

[four.cycle]

I suppose expecting adults to not attempt answering questions for which they really don't know the answers would be too far a stretch, huh?

There are all sorts of questions raised on this forum. None of us know all of the answers to all of the questions. I have no issues with the fact that I don't know everything in the universe, and in those cases where I don't know the answer, I prefer most of the time to just remain silent. Other times, I'll open my mouth (or start typing) and stuff my foot into it - like the recoil spring on the drill press question.

Maybe a better idea would simply be to point out that not knowing all the answers isn't a character flaw. Not being able to admit you don't know all the answers is most definitely a character flaw.

I know, I know.... expecting adults to act like adults is a tall order.


There are a lot of men in high places who have a great deal of money invested in this "AI" thing, and they're hoping and praying they can get it to work so they can cash in.
Problem is: it will never be more than "garbage in, garbage out" - they'd make better headway shoveling sand against the tide with a spoon.

 

[zimman]

They call it "artificial" for a reason. LMAO
Zim

 

[PCustoms]

I suppose expecting adults to not attempt answering questions for which they really don't know the answers would be too far a stretch, huh?

I've stopped responding/interacting to many threads because it's tiresome to put time into a knowledgeable response only to see it get burried with garbage from people that think they know. The BS responses treated as "fact" seem to be increasing lately, and the thinly veiled (at best) AI style copy and paste responses sure aren't helping.

I also feel like there's been an uptick in one post wonders posting some ridiculously obvious questions. I can't help but think it's just training/tuning AI modules, which has been rampant elsewhere for a month or so.

 

[NUTTSGT]

Fair point, and I’m not expecting any hardline enforcement. If it were up to me, I’d post an official statement along the lines of “Please keep your contributions original and avoid posting AI generated output” and leave it to the community to self-regulate.

I believe people that post it genuinely think they are helping but I feel it’s antithetical to the spirit of a discussion board. I’d love to hear what the moderators & admins think about it.

It sounds like you want every post reviewed and moderated. Not going to happen.

Let people spew enough BS and the community here will figure it out sooner or later. . . . be something similar to the boy who cried wolf.

 

[scooby074]

AI generated info is firmly in the trust but verify category. So many errors in my experience

 

[PCustoms]

Forgot, I got beat to it...

 

[Hakeem]

It sounds like you want every post reviewed and moderated. Not going to happen.

Let people spew enough BS and the community here will figure it out sooner or later. . . . be something similar to the boy who cried wolf.

No sir. In fact, in the post you quoted I specifically said that I wasn’t expecting any hardline enforcement. Reviewing and moderating every post is clearly unrealistic.

I would just like to know the official GJ stance on AI content. Is it encouraged? Discouraged? No particular feelings one way or another?

 

[Hakeem]

I've stopped responding/interacting to many threads because it's tiresome to put time into a knowledgeable response only to see it get burried with garbage from people that think they know. The BS responses treated as "fact" seem to be increasing lately, and the thinly veiled (at best) AI style copy and paste responses sure aren't helping.

I also feel like there's been an uptick in one post wonders posting some ridiculously obvious questions. I can't help but think it's just training/tuning AI modules, which has been rampant elsewhere for a month or so.

I can sympathize with your frustration. I’m a fan of your posts so I hope you’ll stick around!

 

[four.cycle]

I also feel like there's been an uptick in one post wonders posting some ridiculously obvious questions. I can't help but think it's just training/tuning AI modules, which has been rampant elsewhere for a month or so.

^ The only way you can find out if your script is working properly is to put it online and run it. So no doubt that is what you're are seeing.

I don't believe, however, that every first-time poster who asks a dumb-*** question should be automatically assumed to be some sort of "AI" bot. People ask stupid questions all the time.

Eventually all of this will get sifted through and will be seen for what it is: some college boy's ********* he thought would make him a billionaire. No matter how you slice or dice it, it's always going to be "garbage in, garbage out". Nothing is going to change that fact.
Sure, they may be perfectly capable of programming robots to do robot stuff like assemble automobiles, but you're never going to build a machine that is going to think - you can only build a machine that will regurgitate what is fed into it.

There's a great example in this very prescient animated short that won the Academy Award for best short in 1974: Closed Mondays

 

[Burt Shaver]

it's always going to be "garbage in, garbage out

but you're never going to build a machine that is going to think - you can only build a machine that will regurgitate what is fed into it

Are you sure? Is this not how the human mind works? Are we not fed (taught) millions of actions beginning at birth and continuing for the rest of our lives. From the most basic action like walking that we take for granted and how we react to a situation, have we not been taught different responses to each incident, do we then not go through our past experiences and pick the most appropriate response for the situation at hand? If we have never been in a situation before would we not search through experiences we have had to find the most similar experience. If we happen to chose an incorrect response would we not learn that, that was incorrect from the facial, and verbal responses we receive. I know nothing about AI, but I imagine if fed enough information for it to chose from, it could very well appear as intelligent, also I imagine it has the ability to scour the internet for information and is also programmed to decipher good from bad information based on the reactions to that information it found on the web. Don’t get me wrong, I’m not a big fan of technology but I worry AI is or could be programmed to think from the information available which is plenty

 

[PCustoms]

At least AI uses paragraphs...

 

[Burt Shaver]

I always hated English class, always preferred math, science, and phys. ed. Of course

 

[mike93lx]

Problem is: it will never be more than "garbage in, garbage out" - they'd make better headway shoveling sand against the tide with a spoon.

The same kinds of things have been said for nearly every new technology.

You may not like it and may not believe in it's potential but it will do amazing things that we can't even think of now

 

[RonnieC]

It would never have occurred to me to google a response to a posted question and post that as a response, let alone an AI answer. The value of forums like this is experienced members.
You see it all the time- someone says “I’m having trouble accessing the rear bolt on my ‘67 whosiwhatsit transmission.” Someone else chimes in “ I just did mine a few months back. If you use a short flex head with a 12” extension you’ll have just enough room to get it out.”
Whereas AI will list the pros and cons of maintaining your transmission.

 

[WildBill]

I hate it. I don't want to see confidently wrong AI BS answers in my online searches, so why would I want to see it regurgitated here. I put people who do it on my ignore list.

 

[RalphInCA]

Whereas AI will list the pros and cons of maintaining your transmission.

AI answers go well beyond this.

They will definitely go into the details of a procedure.

And they’re getting better all the time.

 

[rust in the eye]

Asked the question "How to disguise a regurgitated AI answer"


AI Overview


You can disguise an AI-generated answer by manually editing it to incorporate personal insights, vary the language, and refine the structure
. Fully rewriting the text by hand is the most effective method, as AI detectors are specifically designed to flag the predictable patterns that AI writers rely on.
Manual editing strategies

• Fact-check thoroughly. AI models can generate plausible but incorrect information, a phenomenon known as "hallucination." Always verify every fact and statistic with credible sources before presenting the information as your own.
• Add personal context. Infuse the answer with personal experiences, anecdotes, or opinions that an AI could not generate. This makes the content more relatable and authentic to a human audience.
• Vary sentence structure. AI text often follows a predictable cadence with similar sentence lengths. To make it sound more natural, combine short, abrupt sentences and break up longer, more complex ones.
• Refine the vocabulary. Replace overly formal, jargon-heavy, or repetitive words with more natural and varied language. AI models frequently use stock phrases, which can be a red flag for detection software.
• Strengthen the flow. Ensure smooth transitions between paragraphs and ideas. An AI might produce a logical structure, but a human editor is needed to make sure the arguments build on each other cohesively.
• Fix structural clichés. AI is notorious for using boilerplate introductory and concluding phrases like "In today's fast-paced world..." or "In conclusion..." Replace these uninspired cliches with more original and engaging alternatives.
• Proofread for human imperfections. While AI can be excellent at grammar, perfect text can sometimes read as sterile. A final manual proofread can introduce a more natural, slightly imperfect touch to the writing.

Risks of AI-powered rewriting tools

• Inconsistent results. Using a paraphrasing tool like QuillBot or even asking the original AI to rewrite its own output can produce mixed results. The rewritten text may still retain detectable AI patterns or require further editing.
• Potential for errors. Some tools designed to bypass AI detectors intentionally add grammatical errors or odd phrasing to trick the software, resulting in lower-quality content.
• Ethical concerns. Many academic institutions and employers view using AI to disguise generated content as a form of cheating or academic dishonesty, with potentially severe consequences.

Focusing on value instead of tricks
Rather than trying to trick a detector, the most reliable strategy is to use AI as a tool to improve the writing process, not to replace it.

• Generate an outline: Use AI to create a structural framework for your writing.
• Conduct research: Use AI to find and organize information, but always verify the sources yourself.
• Refine drafts: Use AI as a virtual editing assistant to help you polish your own drafts.

By focusing on adding your own unique voice and human insights, you can create high-quality, engaging content that will resonate with both readers and search engine algorithms.

Pretty soon it will be following its own advice.

 

[theoldwizard1]

Call me a Luddite but I’m not interested in seeing someone copy & paste something that some AI program generated.

I have done it a few times, but I ALWAYS say the source !

 

[larry_g]

It would never have occurred to me to google a response to a posted question and post that as a response, let alone an AI answer. The value of forums like this is experienced members.

I do google answers all the time for members. Example, the other day a person was fighting with a taper lock hub. He got a lot of wrong answers to his problem. I googled the mfg, found the instruction sheet for that hub, and gave him the link. I'm not about to spend a half hour developing a lesson for a member when a few clicks of the mouse will connect him with the manufactures literature on the subject. Many here have no idea that the manufactures publish a great lot of information about their products.

As far as AI answers I consider them one point of view, not the end all for the discussion.

lg
no neat sig line

 

[kj_mustang]

But do you know the source that the AI tool used to gather their info. If people would study the sources used by each AI program, they would quickly realize how much inaccurate garbage is ingested into these systems.

 

[logical]

AI isn't fully matured but it's learning from us every day. When it starts calling you a ***** for having drywall, TV and a 4 post lift in your garages, you'll know it's at full maturity.

 

[bluedog225]

I was asking X.ai some bicycle questions. It kept giving me wrong answers. Then I asked why people duck tape cats to their bicycles. It gave me a long explanation of why this was a good idea.

 

[ChevyEFI]

I was helping a friend replace the clutch in his old Plymouth.
We needed to know how much fluid it took.
The AI generated response was 3 pints,
It actually takes 3-1/2 quarts.

AI needs to be thoroughly reviewed and double checked before relying on it.

Prime example right here. The clutch requires no fluid.

 

[whateg01]

My search,

"tc-32 grade tcb" machinist tool

Selected the AI Mode

Some Google AI answers

TC-32: Insert Geometry and Size
"TC-32" is typically a code for the physical dimensions of a Triangular (T), Positive Rake (C) insert with a specific size designation (32).

• Size "32": Represents an insert with a 3/8" inscribed circle (the '3') and a thickness of 2/8" or 1/8" depending on the specific standard (the '2').
• Common Variants: You will often find TCMT-32 or TPC-32 carbide inserts used in turning and boring operations.

First source link is

GUHRING Grooving Systems - Series 27220 - Insert for back boring and profiling

2nd is

https://exhibitorsearch.messefrankfurt.com/images/original/document_downloads/10000006202501/374019/1734440222497_3932263244.pdf

3rd is

CCMT32.52MT TT8125 ID: IC065519598

www.carbideanddiamondtooling.com

Moving down

Summary of "TC-32 Grade TCB"
If you are looking at a specific insert labeled this way, it is likely a Tungsten Carbide Blade (TCB) or an insert for a Tungaloy TCB holder with a TC-32 (Triangular) geometry.

Link

Solid Carbide Knife, Chemical Fiber Blades - Huaxin

Chengdu Huaxin cemented carbide Co., Ltd. is a professional tungsten carbide knives/blades manufacture since 2003. Its former is the Chengdu Huaxi tungsten

www.huaxincarbide.com

What about this? Photo clearly shows an item that isn't a carbide insert but this is posted anyway. Excuse later was something about it being a warning not to use AI. WTH?

 

[RTM]

What about this? Photo clearly shows an item that isn't a carbide insert but this is posted anyway. Excuse later was something about it being a warning not to use AI. WTH?

Those were their source link references, provided in little link logos in the lower right corner of the paragraph on their original reply.

Please read the Undine thread, as an example of people posting AI as gospel, and me complaining about their naivete, suggesting they dig etc.

WTH?

 

[M.Brane]

Is it people posting AI or is it bots?

 

[lovetap]

I now type "-ai" before every Google search because I don't want to boil a pond in energy wasted for a demonstrably wrong AI Summary everytime I search for something.