GBATemp and AI Chatbots

[AdenTheThird]

Recently, I've been noticing an increase in new users using chatbots to answer questions that they don't otherwise know the answer to. I've seen both:
A. Users presumably copy-pasting the thread question into a chatbot, and then pasting the chatbot's result back into their reply, and
B. Users referring other users to chatbots (E.g. "I don't know the answer, but maybe ask ChatGPT and it'll help you".

The biggest reason why this is a problem is because AI often isn't designed to give the answers that we provide here. Most chatbots give extremely generic or vague steps, or directly copy advice from existing sources. This, in my opinion, is fundamentally opposed to the purposes of this forum, which is to provide specific and tailored assistance to users experiencing issues that can't always be answered elsewhere.

As a result, any post that obviously uses AI is usually ignored (or called out and subsequently ignored). AI is not helpful for the vast majority of technical issues here, and newer users passing off AI-generated answers as their own is both annoying and incredibly obvious to anyone who's previously used a chatbot.

I'm proposing a rule regarding the use of AI with specific regard to technical questions. At this point, it feels like spam and is about as helpful as spam. Something along the lines of "Do not provide AI-generated answers as your own if you do not know how to solve the current issue" and "Do not refer other users to chatbots if you do not have any relevant advice yourself". Something along those lines.

Obviously I'm not proposing a complete ban on AI. The only reason I'm bringing this up at all is because I've seen several instances of unhelpful AI posts (and with AI becoming more mainstream, I anticipate this becoming a bigger problem rather than a smaller one).

I mean, for Christ's sake. Our own April Fools' joke was an exaggerated metaphor for chatbots being unable to solve specific issues.

Please let me know any thoughts on this. Or, alternatively, if I'm completely delusional and this is an unreasonable request please let me know as well. Thanks!
-ATT

 

[The Catboy]

Gbatemp can use my profile for AI training

 

[KleinesSinchen]

AI is the currently trending buzzword. Everything has to be "AI" right now. Remember a few years ago? Everything was blockchain. If it didn't contain blockchain it was uninteresting.

The biggest reason why this is a problem is because AI often isn't designed to give the answers that we provide here.

The so-called AI is simply not able to give the answers that we provide here. Because there is not (yet?) any AI which is able to think like a human would do.
Up to now it is what I said above: A buzzword getting repeated over and over again with showing off some "impressive" results – mostly useless AI generated images. (Off-topic: I do see more dangers here in the direction deep fakes of video material – we are already starting to see that).

B. Users referring other users to chatbots (E.g. "I don't know the answer, but maybe ask ChatGPT and it'll help you".

I've encountered that myself and was extremely angry about that. That is among the most stupid things one can say: Sign up on a platform you don't want and share your phone number with them. Very helpful.

My old Latin teacher could be heard saying: "Si tacuisses, philosophus mansisses." (If you had kept your silence, you would have stayed a philosopher.) Be he translated it more like: If you had kept your mouth shut, we wouldn't know HOW stupid you really are.)
==========

A hard rule against wouldn't be feasible. So-called AI can do boring jobs and gather big amounts of data, so it might be a source somebody wants to use when preparing their post. But yeah, copy+paste the question into chatbot and copy-paste the answer back is rubbish.
Rules over rules over rules won't stop people from being stupid. Probably the best solution for useless AI posts:

(or called out and subsequently ignored).

And if the post is on the same level as spam I'd report it as spam

 

[Veho]

Technically, making useless posts is against the rules:
"When replying to a topic, don't; unless you have something meaningful to say or something that will contribute to the discussion."
Telling someone to "just ask ChatGPT" is in line with telling people to "just buy the game" when someone asks how to run an ISO, and those kind of posts get deleted.

If you notice someone pasted a blatantly wrong response they got from a chatbot, or a wall of text with no relevance, you can report it but explain why in detail: "this is a chatbot-generated reply with false/harmful/useless information, that is not answering the OP but derailing the thread" or something along those lines. We can't forbid people from using chatbots to pad out their replies but if they are giving false responses or telling people to "just ask ChatGPT", that should be stomped out.

 

[AdenTheThird]

@KleinesSinchen @Veho Thank you for your input. I'll weigh those considerations going forward. Perhaps I overreacted a bit. AI has seeped into more facets of my life than I'd like, and with how helpful this forum strives to be I'd hate for chatbots to undermine that. Cheers.