Is ChatGPT compliant with Talk Guidelines?

[Alstroemeria123]

I’ve seen a lot more posts recently that seem to be copied and pasted directly from ChatGPT and similar. I don’t think I’ve ever seen one that adds anything to the discussion, and frankly I come on here to discuss things with humans, not AI.

Could you confirm whether using obvious AI to produce a response is within the guidelines?

That’s really good question!

It's depressing isn't it? Nearly every forum I frequent seem to have these AI responses. It's particularly obvious on Reddit especially. Bloody irritating!

So much "calculated to make MN-ers froth" AI at the moment. Sometimes it looks as if every post in trending has been generated by a chat-bot. There's a samey-ness about the posts and an inevitabilty about the responses that makes MN very dull reading atm.

Hi there @Alstroemeria123. Good question - it is something a few people have asked about. It very much depends on context.

If it were being used in a way that broke guidelines, eg to troll, or to spam users with products and services etc, we would (and do!) take that down.

Outside of those obvious cases of trollery and spamming, we do see why people feel its use is not in the spirit of a genuine conversation, and that spirit of 'real' discussion is certainly what we hope to foster on the boards. However, we do know that some people who might be dyslexic, or for other reasons find getting words down in print difficult, use ChatGPT to help them formulate their thoughts, and we don't think we can really object to that, as long as it's done, as we say, in the spirit of genuine discussion.

What we'd say is that if you have concerns about a particular poster or thread, hit the report button. We're always happy to look behind the scenes and we can usually work out fairly easily if a poster is genuine or not. Hope that's helpful.

I've reported one this morning @HebeMumsnet ...

@HebeMumsnet the post I reported this morning has been allowed to stand, on the basis that:

we will see more posters using AI help to write posts, especially posters who struggle with spelling etc and we're not sure it would be sensitive to intervene

The post itself was much better written (spelling and grammar) than would have been usual, from that particular poster - who obviously I'm not going to name - which is how I realised it must have been autogenerated.

But the point is, I don't want to be on a thread having a discussion with a bot. I want to be talking to a human being - even if they can't spell or write with good grammar.

They're so obvious too. ChatGPT has got a long way to go to be relevant.

I completely agree with this. Do people not realise?

I totally get where you're coming from. It can be frustrating when you're looking for genuine human interaction, and it feels like automated responses are taking over. I value authentic conversations too, and I think it's essential to strike a balance between technology and genuine human connection. Hopefully, we can keep the real, heartfelt discussions alive here on Mumsnet!

Only if people are using it lazily.

You can set custom instructions for it to take on any kind of tone of voice or character you like. When you can tell it's ChatGPT they haven't bothered to do that

Well clearly MNHQ really don't care where their clicks come from, do they @HebeMumsnet?

In answer to my report, where I went on to explain that it isn't about spelling and grammar issues, but it's about wanting to interact/discuss things with human beings, this is the reply I got:

Thanks, but we have no reason to believe that [poster] is anything other than a human being accessing the power of AI. If that doesn't appeal, you're (currently, at least, and for as long as our robot overlords will allow us!) free to skip past.
Have a good day, BIWI!

How fucking patronising is that?

Thanks for your response @HebeMumsnet

my concern with posters using AI, is really when they copy and paste huge tranches of text. If someone finds it helpful in terms of structuring their thoughts, that’s different, but that doesn’t seem to be how the noticeable AI is being used by a number of posters. And, speaking as someone with dyslexia, it is very very difficult to read extremely long posts. When it’s the OP that’s long, I can easily close the thread and not read further, but it’s difficult to skip over a post in the middle of a thread.

I don’t think AI is going away any time soon, so it would be good to have clarity from MNHQ about what is acceptable in terms of AI posts. Because if MN becomes mainly bots talking to each other, I think a lot of people will just give up on it.

I'm quite alarmed that HQ have no thinking on AI and ChatGPT bots posting increasingly on their site and members being uncomfortable with that - other than to retort "don't like, don't look".