Continuing the discussion from I am working on a code and It is using a ir sensor to sense walls and I need a code to make it sense the wall and turn my robot 4 time:
@gilshultz
I see you like to post AI answers to the forum.
Not to be rude, but I don't think that's a good idea.
At the very least, in addition to AI-generated content, the message should contain your own experience. Otherwise, these messages do not bring anything useful to the forum.
Hi @b707. Do you have any technical criticism to provide on the referenced post from @gilshultz?
Is there anything in accurate?
Is there anything irrelevant?
Is there anything inappropriate for use within the context (e.g., unnecessarily going far too deeply into technical details or advanced techniques when supporting someone having basic beginner problems)?
I think it will be much more productive to base this discussion on objective technical details rather than on "AI bad" subjective opinions.
Please keep in mind that responsible use of AI generated content is allowed on Arduino Forum.
I haven't found LLMs like ChatGPT to be a useful tool for my own work so far. However, the thing I can say in favor of LLMs is that they produce relatively high quality writing (in terms of organization/structure/formatting; not accuracy), significantly superior to the majority of the human produced writing on the forum. So if a human who is competent at prompt engineering and knowledgeable in the subject matter is willing to do a thorough validation of relevant AI-generated information and use it to provide support here, that could actually be a significant contribution to the forum. It is my impression that @gilshultz is making the effort to validate the AI-generate content they use.
Unfortunately, performing a thorough validation takes a significant effort so there will always be a temptation to skip that and simply trust the information. We know that, as with any non-authoritative source of information on the Internet, we can not cursorily trust the accuracy and suitability of AI-generated content. The forum community does need to push back when we see instances of well intentioned people cutting corners on validation of information.
Keep in mind that nobody is perfect and even those who make the strongest efforts at accuracy in their posts are going to be wrong occasionally, regardless of the source of the information they are sharing. We should always strive to get closer to perfection, but we should also try to be understanding when others make mistakes in good faith (while also being sure to correct these mistakes).
I thought it was a GREAT answer and have seen lots of responses that are similar. What do you think a politically correct answer would be.
You are right that evaluating such a post is a lot of work. And may be my suggestion in this case was wrong.
But for example, in another topic, where @gilshultz also sent a huge piece of AI-generated text - I think that his message is, at the very least, premature. And perhaps it turns the questioner in the wrong direction.
In that discussion, the author asks why using a PWM affects the Serial output. I think it is impossible to answer here accurately without a diagram and full code.
@gilshultz message immediately directs the author to eliminate standard problems of electromagnetic interference, although the problem may be completely different.
The main danger of such answers is that they look convincing and make an impression on a beginner of the answers of an experienced person, although in fact they are not.
In fact, I have nothing against anyone personally.
I don't like the idea of copying AI's answers into the forum. I think it devalues communication between real people.
I have noticed a few posters (non regulars) chiming in with what looks like an AI post, not adding very much to the topic. At least one I think contained incorrect information.
"Responsible use" seem highly subjective. I tried several times pasting the poster's question into an AI, and often it comes up with a good answer, but I am not comfortable acting as a "pair of hands" for an AI chatbot.
It's already quite off-putting being asked to debug people's AI generated code. I guess eventually a lot of web forums will devolve to AI generated questions getting AI generated answers.
A common spammer technique is to make a post from a copy of relevant content from some source on the Internet. A spam link payload is inserted into this content. In recent times, AI-generated content is increasingly used for this purpose.
So if you see a post that looks odd, definitely take a close look to check whether it contains a spam link. If so, please flag the post.
AI-generated or not, we will always take action on posts made for malicious purposes.
Surprisingly, the switch to using AI-generated content for malicious purposes has actually been a boon to moderators. The reason is that AI-generated content has a distinctive character that is easily spotted once you are familiar with it. Previously, when the spammers would use copies of human generated content it was very difficult to spot and so such spam slipped past us much more often.
And correctly so. That would not be responsible. Please only use AI-generated content after you have thoroughly validated it for accuracy, relevance, and appropriateness.
I understand, but keep in mind that nobody is forcing you to do so. If you aren't inspired to help someone with such code, you are welcome to simply move on to another forum topic. There is always plenty to do here on Arduino Forum so we can definitely pick and choose where we want to direct our efforts.
Yeah, I know a spam post when I see one, I don't think these were spammers.
Perhaps I wasn't clear. I don't feel comfortable reposting an AI answer, even if I have thoroughly validated it for accuracy, relevance, and appropriateness.
However, I am sure other people will post AI answers without checking. And as I can't tell what thought process they used, there is no way to tell if they were "responsible" or not.
Btw, please don't feel you need to defend the official policy. It is what it is.
Actually I used AI to clean up my original, not create the answer. I am very poor at english structure. AI is a lot better then the other programs I have tried. The format is what we were taught at MU in technical writing. I have tried AI searching for information and it has come up with some of my stuff I have published.
This is what Ai gave me which is better?
Actually, I used AI to refine my original text, not to generate the answer. My English structure isn’t the best, and AI does a much better job than other programs I’ve tried.
The format I use is based on what we were taught in technical writing at MU. I’ve also used AI to search for information, and interestingly, it has surfaced some of my own published work.
My English is bad enough - I don't see any important difference 
And I never in my life used an AI.
Please ignore my attack.
I had a publicist until a few years ago. She always did the correction, spelling, structure etc. I would send her a text file and she fixed it for me. She helped me for many years after I retired then had to stop.
and the polished version.
I had a publicist until a few years ago who handled all the corrections, including spelling, structure, and formatting. I would send her a text file, and she would refine it for me. She helped me for many years after I retired, but eventually, she had to stop.
Your first version is just as understandable as the second one 
The only thing that might be a little better is the last sentence. Unless you are a language purist I would not bother.
Note
I'm not native English speaking and have to pay attention to what I type 
It has its advantages 
While I am a native English speaker, I have dyslexia and so spelling is not my strong point. Back at school in the days before this condition was recognise I was called lazy, inattentive and stupid. So my first spell checker was my girlfriend, later my wife. She was studying at a university department making the first research into what is now a widely recognised condition.
Therefore I am acuity consensus of trying to get things correct. If the normal spell checker can't get what I am trying to say, then I use an internet search for a phrase that contains the word I am trying to spell. Often that causes the correct response in the suggested links.
I do struggle with homophones and small words still pose a problem.
I've enjoyed this thread so far more than any others because of the human side of you people. I also struggle with this.. thing.. yeah, words. A mix of bad memory and Swedish. Not bad Swedish but I tend to put the English words in Swedish order.
(To the quote;
Board and Bored - Norm Macdonald heckles Courtney Thorne-Smith for doing a movie with Carrot top, named Chairman of the board.
If you enjoy Norm check it out on YT. Nobody and nothing funnier, imo.)
Since you asked it just might be all the boldface. Now, I'm someone that
inline code of course, andall more than average. Of course, this is a forum where we discuss code, and those features of the forum help with clarity.
But what is quoted in the OP
A numbered list, each with the first sentence in bold Title Caps, followed by a (bold) colon? That's a bit much. And this is one of the milder examples I have seen. And all that structured bold both screams for attention and says, "not written by a human".
Some of the choices for bold, like in post #11, seem unwarranted. Maybe just use the AI to improve the structure and phrasing, but turn down the styling? Is there a setting?
And may I ask: where/what is "MU"?
My guess is University of Munich
Has it?
That appears to mean something quite different to what you originally wrote.
That illustrates the problem with AI. Even though people are supposed to check it, how is someone with less knowledge supposed to check something with supposedly more knowledge?
Several studies have shown that people rely too much on the output of AI.
I believe these critera are too narrow:
My personal take is this. In principle I don't like seeing AI generated answers but agree there are valid uses for it.
The situations where I consider it perfectly acceptable to use AI are for tidying up grammar or translation and possible structure.
If an answer is (a) focussed and (b) relevant and (c) accurate and (d) does not repeat something which has already been said in the thread and (e) does not stray outside the responder's expertise area and (f) does not repeat something that is in the standard instructions for forum users, then it is OK to use AI.
Long, boiler plate answers triggered by a keyword in the OP's problem description should anyway be handled as a link, not a copy paste operation into the thread, which simply clutters it up.
Maybe slightly off topic but I don't find it too difficult to read posts from forum members whose primary language may not be english. There are some words that prompt clarification but for the most part I/we get by.
Regarding spelling etc, I don't know if this is true for languages other than english, but I can read the following at pretty much normal reading speed:
"It deosn't mttaer in waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht the frist and lsat ltteer be at the rghit pclae. The rset can be a toatl mses and you can sitll raed it wouthit porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe."
At least in german it's the same . And it's one reason why it is often difficult to find (own) typos. You simply read over it.