If you are using AI generated text, add a disclaimer?

AI is helping us…become lazy. It seems like using ChatGPT and other such tools for day-to-day communication is the new norm. From product reviews on shopping websites to emails, real humans are happily utilizing generative AI tools.

That creates a major problem. How can anyone distinguish if an AI generated text was written by a bot or a human used an AI tool to write what they wanted?

Recently, a new member joined and posted long wall of text. It was written in cowboy style. I admit that it was amusing. But if an account starts posting 2000 words of AI generated text, without proper context, it makes little sense for people to interact with them.

This forum is for human to human interaction, not human to cyborg interaction.

Should we go for AI disclaimer? Like, if you use AI to generate an answer that you wanted to write, specify that “I used xyz tool to get the text”. This adds transparency, but I’m not sure how beneficial it would be.

Your opinions are welcome.

5 Likes

That is what I have been doing.
I sometimes use AI ( mostly google) to get some info on a topic I cant help with personally.
I prefer to give a link, but sometimes with a complicated topic what is needed is a summary of all links. AI seems to do that well most of the time, but I have had it totally misread the question.

5 Likes

+1 for a disclaimer requirement.

7 Likes

ISTR that it was suggested that AI is programmed such that it cannot not answer a question so will just scrape whatever rubbish it can find that someone, somewhere has upvoted thereby just adding to the chance that it will continue in its perpetual quest to provide absolute rubbish. I think the principle is known as RIRO.

3 Likes

Make that +2

Pointing to a site with more details is good. Copying part of a text from another site is ok if its detailed and personalised such as do this then do that , perhaps with a link to an explain why or what such as a technical answer, such as images of ssd connection types where difficult to explain. Or screen shots.

The best answers i have seen reciently were from fellow member jorge, he wrote what to do, showed an image with arrow pointing to the part of the screen, then the next step….. bravo jorge.

On a personal noté I did not like the cowboy style of reply. Ok we all use english differently but he was off the wall.

5 Likes

Same here. I muted it after the 2nd reply. :unamused_face:

3 Likes

+1 for a disclaimer requirement from me too.

I didn’t read the “off the wall” post and have no wish to do so. I detest AI and the way that it is being used to influence people who should know better. Stack Exchange has a very clear policy on this

which I wholeheartedly support. I’ve seen posts intending to do nothing more than just gain rep on Ask Ubuntu clearly wrong putting users at risk of trashing there system. To their credit, Mods quickly identify these and remove them completely which in turn prevents the AI bots from scraping further rubbish to add to their algorithms.

3 Likes

@abhishek

I have used copilot with some of the issues and the learning of “ LinuxFromScratch “!! So far it has been, rather spot on!!!

5 Likes

@abu

I replied a couple of times but not care for all the “ cowboy “ style of writing!!!

2 Likes

Then you are more patient than I am. I just found it annoying.

3 Likes

Hi @nevj and all, :waving_hand:

Exactly the same with me.
There are occasions when we want to provide some help but need some help ourselves. In that case using AI on and off may come in handy..
Before forwarding the answer I try to check its contents for plausability.

Naturally I add a disclaimer or mention it like so:
" Here´s duckduckgo´s Search Assist´s opinion"…
It should be a matter of course.

Many greetings from Rosika :slightly_smiling_face:

3 Likes

Good idea, but if you are going to do that, why not simply add a link to the page or site.

2 Likes

Hi Paul, :waving_hand:

That might not always work.
E.g. with Perplexity the topic you initially created is publicly available only for a limited time. I don´t know exactly for how long, perhaps 14 days?

I tried to access some of them, which I saved as bookmarks. It said the topic has expired or something to that effect.

You could log in to perplexity using your personal account, if you have created it.
In this case your threads would be saved but of course they wouldn´t be availble to the public this way.

Cheers from Rosika :slightly_smiling_face:

4 Likes

Riding a Horse without a saddle is OK, Riding a saddle without a horse, that not my problem.

1 Like

Hello everyone,
Abhishek @abhishek has created a topic with very interesting questions.

I will give my opinion on this subject and try to answer Abhishek.

Whether we like it or not, AI has become an “indispensable” tool in our daily lives.

This is a random example, but I just want to demonstrate some technological evolution.
In the past, when “I” (I’m speaking in the abstract) didn’t know the meaning of a word, I used a dictionary; years later, I searched on Google; now I “use” AI - it’s the trend to use what is easiest, which gives us “a more elaborate answer,” and we only need to copy/paste if we need to use the answer.

Of course, we must always question the answer, just as we must question the answers we find when searching the internet.

About using AI in the forum
I believe that all of us, when mentioning research carried out on AI and posting it in the forum, for example, our friend Rosika, but I could mention any other user, are careful to mention the source of the information.

I believe that we all do this consciously only when we don’t know how to help in any other way.

I will give a concrete example that happened to me recently here on the forum, to help a user who had no audio from their webcam.

I reached a point in the topic where I didn’t know how to help anymore. I had to study, search the internet, and use AI, in other words, everything at my disposal to be able to respond.

Although I used AI, all the answers I gave, if memory serves me right, I tested on my PC and some answers on both my PC and laptop.
In other words, I tested what I learned on my PC, saw that there were no problems with the code, or changed what was actually wrong and posted it on the forum.
Although I used AI, all the answers were personal.

On the other hand, I believe we are all careful about placing links in topics.

A few days ago, we were flooded with “information” from some users.
I was suspicious because of the long texts, some of which seemed to be taken from AI, but since it is not always easy for some users to post on the forum, I genuinely thought it might be a way to post and even gave some likes. I stopped doing so when I got fed up.

In this forum, we have always been careful to provide the most reliable information possible, and when possible, we have always mentioned it, without being ashamed to say so.

If it has to become an obligation, I fully agree, so that the forum can continue with the same spirit of mutual help as always. Otherwise, we no longer need a forum, we can just use AI.

Jorge

6 Likes

Only solution for horses

I feel strongly about people using AI for their posts on a forum. The least people could do is add a disclaimer that AI was used for generating the text; otherwise the post ought to be removed.

We can’t afford to let this forum become a dump for AI drivel.

Now, when someone uses AI as part of their research for a post, that makes things entirely different.

Generating an entire post with AI on the other hand… yuck. :-1:

8 Likes

Well said!

OTOH, we all should probably learn to say “I don’t know”. It’s even better as using foreign knowledge stolen by a AI grabber.

2 Likes

One can say “I dont know”, and then proceed to look it up somewhere.

Making the whole text of a post with AI is a different thing altogether, and , as @xahodo says , totally obnoxious.

3 Likes

+1 for a disclaimer requirement from me too

I was recently in a huge “tech bridge” 30+ people - an issue with Solaris - and some “joker” kept posting AI responses related 100% to Linux only - NOT Solaris!

I kinda hate AI, sorta… But I don’t disable it - and sometimes the responses for tech queries are useful, and nicely formatted (via Google-Fu)…

3 Likes