Today, people use the internet and AI to find solutions to their problems, and when they solve them, they want to share their "personal" experience to help others. And then a moderator applies the forum's "guidelines" and decides you've used AI.
End of knowledge sharing.
As I told the moderator, AI is here to stay. Accept it.
1 Like
There is no problem whatsoever if you declare your AI chatbot use openly. When you post totally f***ed prompts and not bearing with conversation playing "helping" is just plain waste of time.
Obviously you come here when AI did not solve your problems, and most likely already compromised your router and home network. S in AI stands for Security as we all know.
3 Likes
I didn't want to go into too much detail, but... I replied to a post to provide a definitive solution to a problem with a Mikrotik router, and the moderator decided, without asking me, that I copied it from the AI. The procedure is entirely my own, and I've been accused of using the AI. That's why we're here. Who decides?
Thanks anyway. And please don't take offense, but we're supposed to be on a technical forum, not in a WhatsApp group.
Like you said, AI is here to stay. In technical forums (such as this) the community has a responsibility to uphold a high level of scrutiny and accuracy on its content to ensure that low quality AI ramblings are not propagated into the future training models.
If we allow AI to answer things close but not 100% right, that is a race to the bottom.
Don’t take it personally. Declare your use of AI, fact check its responses before posting and it should sail through ok.
6 Likes
Thanks @Lantis1008, you're always very constructive.
The thing is, I did countless tests for days until I found the right procedure. Along the way, I bricked the routers, but I didn't give up and managed to install OpenWRT on four routers that are difficult to install. In fact, openwrt.org doesn't have an information page for this model, nor a procedure for installing it, except for the general procedure for all Mikrotik routers.
So nobody's going to benefit, because it's not going to be published. And if they don't want to publish it, that's fine, but I still think the same thing: who's going to decide what is knowledge and what isn't?
A device page in the wiki or another page to capture your specific instructions sounds like a great place to document everything!
1 Like
Sorry for bumping a solved post.
But it’s a pain to write a long, detailed explanatory post and then see it reported and removed due to “AI Generated Content”.
What’s the best approach when explaining something extensive in a post? Should we add some kind of note for future admin reviews?
For example:
-
I use AI to learn, as a tool, to improve how I communicate and to tie all the loose ends in what I write.
-
I use AI to sift through thousands of lines of a tcpdump to find correlations.
-
I use AI to doublecheck things before stating them as facts (just like I use documentation and man's/docs pages and not the IA for the same purpose)
-
I use AI without fear because AI won’t take my job; someone who knows how to use it will.
-
Just as in formal education where you give your word and explicitly commit not to plagiarise, I don’t plagiarise in my research, my findings, or when writing in English instead of my native language.
-
Do I need to write badly on purpose, or put “I use AI” in my profile?
I’m not sensitive. But seeing a long, valuable, explanatory post removed does sting.
1 Like
The examples of what have been banned are excellent, thank you.
I am glad the mods have banned it’s use!
1 Like
On a whim, I asked a chatbot to produce an modern openwrt logo. The result was very nice, arguably superior to what is currently used. I posted it to share with the community. Guess what happened.
The internet broke? 
Seriously, that’s about the only thing AI is good at, fake images/videos.
Genuinely, I’d like to see. Will it allow for an external link to the image or better yet the complete post?
Edit:
I’m not trying to be combative. It’s difficult to convey playfulness when you can’t understand how the other side will receive the intent. I’m curious to see where the line has been drawn with respect to AI banning.
Your post was removed because it had AI generated technical content. Beyond simply having the hallmarks of the "AI writing style," it also contained many things that were technically incorrect and would cause problems if users implemented the items in your post.
I first wrote about the risks of using AI (with respect to hallucinations and other misleading information) for technical content about 3 years ago. While AI has improved in many areas, it still tends to hallucinate "solutions" for OpenWrt and networking in ways that can cause massive problems.
We do understand that this stance seems harsh at times, but it is critical to protect the integrity of the information/solutions on the forum.
The community guidelines also describes why it is so important to avoid using AI for the technical content on this forum... quite simply, it's not good enough (yet).
Thank you for understanding.
7 Likes
It's been deleted. I didn't save a copy because I never imagined anybody would find it objectionable in any way.
You should have taken a look how art projects are made. Starting from freehand draft. Not asking AI abot how to best stretch a chewing gum 6 ways.
With all due respect,I understand the points regarding hallucinations and the highly inaccurate responses and problems they cause; however, my content did not utilise AI to generate answers or technical solutions.
These are examples taken from my own configuration files, detailing exactly what I have to do to achieve what the user requested: multi-routing
The deleted post and its explanations contained an example configurations , as well as the procedures required to establish proper policy-based multi routing, as he asked.
-
The person was being guided through visual examples of what needs to be done with their interfaces:
-
They aren't simply told 'follow these steps', 'do exactly this', or 'use these commands'.
-
Instead, it is explained that 'to achieve that, you must design the interfaces in such a way, then configure X with a similar arrangement, because that is how OpenWrt handles this', etc
But, it's ok. I must be the one admited using AI for cleaning text, then wasting my time on repeating explanations that will only be deleted by a human who is mistaken in their response to me when I ask 'What's happen'
Bye OpenWrt Forum.
As mentioned here, can you please delete my account?
Thanks.
You have to post in that thread.
I think the 'changing the mindset and the guidelines' mentioned in this post is more about how random users (or randomers with multiple accounts) turn up in the comments to talk about unrelated stuff just to try and belittle others.
That’s why I reckon I should delete my account.
Let's be honest ... AI usage isn't actually being punished:
- It’s being instrumentalised to humiliate those who don't speak English and just wanted to show or talk about something, help someone out, or contribute with something constructive.
I didn't expect to find toxic people scattered throughout the community when I started here.
Bye!
That’s a reach. The Forum has always required that posts be submitted in English. As well, posters are encouraged to also append their post content in their native language to help resolve any ambiguity in meaning.
2 Likes
All of this is very strange, because I wrote my procedure in my native language, and then translated it with Google Translate. How is it possible that they're telling me it was generated with AI?
I decided to use a very well-known AI-generated content detector, and guess what... it told me that 21% was generated by AI. Unbelievable, especially after having translated it!!!
So if AI doesn't give accurate results, why do we use programs that use AI to detect if we've generated content with AI? Think about it carefully.