Posts written by Chat GPT and other AI

[CruCru]

I keep seeing posts written by Chat GPT on various threads. Sometimes the poster says something along the lines of “I got AI to summarise what I wanted to say and here it is” and sometimes they post with no reference to AI (but it is fairly obvious).

Please can this stop? This site will become unusable if the posts get written by robots. MN is populated by articulate, well educated women; we don’t need to use AI to communicate with each other. It isn’t allowed in school.

I suggest that obvious AI posts should be reported for deletion.

This. The chat GPT posts are almost invariably boring and shit.

I completely agree with this.
I'm noticing the use of what must be ChatGPT or some kind of AI tool on FB business pages more and more - they're all churning out the exact same type of stuff, and just sound like they're parroting each other.
It's so off-putting. It's like nobody has an original thought in their heads anymore.
I haven't really seen it on MN yet, and hope I don't.

I deleted that as soon as I saw it on the grounds that it’s tedious clickbait, tantamount to Isabel Oakeshott pissing about with AI.

I actually don't think the jealous of Dubai was AI. Just a person it didn't work out for and now claimed authority on how it is for everyone and bwst knowledge of the place

You may be right. I’m not sure that I want to go back onto it though.

I’d rather read a post with SPAG mistakes than anything written by AI tbh. Maybe the time will come when spelling mistakes in posts become desirable because they prove a human did the writing!

I’ve already more or less abandoned Facebook because of the sheer deluge of AI shite, and will do the same for any other platform that goes the same way.

There’s a thread under chat where the OP has put in brackets in the title (no thank you, AI). Maybe if AI is unwelcome, this is what we should all do. But it would get rather clunky.

I've just read a thread where the title is AI generated (bland, descriptive) and the OP and her follow up are also AI generated. I'm not saying she's fake at all, but by explicitly encouraging the use of AI on mumsnet with this ridiculous new suggested title feature the powers that be are making the site far more bland and boring. This really will kill mumsnet and I cannot believe they are so out of touch with the reasons people use the site as to not notice. Bringing AI into thread creation smacks of people panicking that AI is coming and feeling they need to 'get on board' and bring it in somehow when they really need to take a stand against it. I am sad to see it, but mumsnet is going to die a death if AI is allowed and encouraged the way it is now. Why would anyone bother posting their dilemma on mumsnet when they get the same quality of response from their ChatGPT app?

I just experimented with the AI thread title feature. (Didn’t post it, obvs.)

It turned a very specific, informed element of an academic subject into ‘Would anyone like to discuss aspects of [general thing]’.

It was as though I’d asked about Fermat’s Theorem and got ‘Anyone else interested in sums?’

Ffs.

Oh God. This was something I hadn’t seen (it’s been a while since I started a thread - maybe this is the last one I did).

Can posters not choose their own thread titles? I can’t imagine anyone is dithering over what to put as the subject before they post.

Why would anyone bother posting their dilemma on mumsnet when they get the same quality of response from their ChatGPT app?

I bet a lot of people are already using AI for advice because it doesn't flame them.

someone has created a thread where the OP is so obviously AI generated, but they've actually denied it. That I find annoying. If you've used AI at least own it!

Are these the ones with one deliberate spelling mistake?

I do that! But that's not what I come to mumsnet for. I love my ChatGPT and I use it a lot. But I don't want to read someone else's ChatGPT generated post. So many people are doing that now and denying it when called out. It's absolutely going to wreck the site but mumsnet don't seem to care.

I don’t think it will make any difference though.

There have been a couple of very similar threads about multi-functional clothes (going straight from work to a wedding, that sort of thing) that came across as AI generated. HQ deleted the first one but the second one developed into a lovely collection of cat photos & the OP didn’t come back.

I’d be fascinated to see what titles it came up with for some of the more outlandish threads.

No, possibly not. Perhaps it will make it easier to report AI posts if the OP has explicitly said they don’t want them.

before ai you still had half answered responses or people derailing etc thats not new at least with ai at times it does answer the ops question

Does it matter very much if you get a few half answered responses? I remember a thread where the OP was choosing a school and said she wanted to only hear from posters who had had children at both schools so she could compare and contrast.

She wasn’t enormously impressed when I said that this would mean she would only hear from people who had been unhappy with school A and moved them to school B (or vice versa). A few people told her their experience of school A / B but said they had no first hand experience of the other. These responses still had value, even if they weren’t the full responses sought.

in that context then yes i agree, but in other treads someones eg asking about cats and someone talks about dogs etc so in that sense etc

I think that might just be a facet of humanity. It’s often irritating but it isn’t as bad as getting complete responses written by robots.

true

Many posters overestimate their ability to spot AI writing

I loathe AI posts with their waffling and generalising and their flipping bullet points. It isn't even possible to tell whether an AI written post is actually making the point the person concerned wants to make, and not everyone reads them carefully to check.

As PP said, some people seem to use AI even when they are perfectly able to write clear English without it. Personally I would much rather read something with spelling mistakes and typos and an unusual turn of phrase, if that is what the poster can manage.

Many posters underestimate people's ability to spot AI writing.
I know when something is written by AI. I'm sure there are some that I miss, but I am 100% confident that when I assert something is AI, it is. I read a lot. I can recognise writing style. It is something humans are capable of doing.

Interesting. You refused the other month to delete posts from a poster who was totally derailing a thread I was on. They posted every 3-4 posts with identical points. It was a tactic to derail the thread. Contributions were neither appropriate nor helpful to the OP. Is that the reporting language i need to use?

@socialdilemmawhattodo It's always helpful to have a description of the reason for reporting and the more detail you can give the better, but ultimately we moderate against Talk guidelines rather than against the reason for reporting. It's definitely useful in helping us see what the problem is, but there are no 'right words'; it's just helpful for us to see what's bothering reporters most.