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.

Congratulations

The thing is all three of these are saying the same thing. None of them have been made up by AI. I dont find the third option that annoying and I'm not sure I'd spot it as AI particularly but then I rarely use AI so don't recognise something written by it that easily

I do know what you mean in principle but the example.you have given is actually one where I have absolutely no problem with what is written by the AI

I disagree that they’re saying the same thing.

In the AI version the son is “flat-out refusing” and “won’t engage at all”, which presents him as being actively defiant. This sounds like a clear behavioural issue and something that calls for more discipline and/or harsher consequences.
“School have said” suggests mum has been in contact with school, and it’s got to a point where they’ve threatened to suspend her son if he doesn’t “comply”. Ending on the uniform issue and the fact that her son “won’t budge”, again reiterates that the main problem is her child’s defiance, and I’d probably suggest she remove devices and any other privileges until he falls into line.
Dad’s anger is presented as being about the situation, and is a problem because she’s stressed about having to mediate between her husband and son - I’d likely suggest Dad being annoyed isn’t a bad thing, as she seems quite passive and needs to find a backbone to be able to deal with this.

In the first/second versions the son “doesn’t like to wear his school uniform”, and I’d wonder if there might be ND or sensory issues going on that could be solved with some adjustments. As there’s no mention of having any contact with school, it’s possible this is something she’s worried about rather than a threat school has actually communicated: I’d ask about what her son doesn’t like about the uniform, and would advise her to reach out to the school for support in handling this.
Much of mum’s concern around this is focused on Dad’s anger, and she’s specifically concerned about him getting angry at her son. I’d be aware of the possibility of domestic abuse, would ask about whether there’d been other times when he’d got angry at her son and what happens when he does, and would try to check that she and her son are safe.

The AI version has added information that wasn’t in the original, changed the focus of the concerns (because AI can’t read between the lines or pick up on nuance), and smoothed everything out to make it more palatable. Her anxiety is pretty much erased, her son is being a defiant little shit, and she’s being overly passive while husband’s reactions seem totally normal.
At best she’d get replies that don’t fully relate to her situation. At worst, a woman in a potentially abusive situation could find herself being given very, very bad advice.

This is why I have a problem with AI being given the basic information then having it write on behalf of people. It prioritises the ‘flow’ of the writing more than the content, adds plausible sounding information that looks right but might not be accurate, and makes everything calm and detached, removing little details that posters could pick up on and use to inform how they reply.

I'd love to see AI copypasta banned. I want to hear people's real voices, I don't mind if their SPAG isn't perfect. I also hate it when someone takes it upon themselves to "helpfully" ask AI and then spews the response onto the thread - no-one asked.
It will only increase and instead of a site with strong individual voices, it'll be tedious AI slop.

one thing thats puzzling is sometimes people ask is that ai, but then when its normal written no one says are your sources newspapers, google, the news, or a degree etc ? so why single out ai ? because people can write half a sensible idea without ai so if its also half a sensible idea with ai then in theory whats the difference ?

Posters on the boards that I frequent do regularly ask for sources (linked) and cite their sources - links to newspapers, academic articles, books, social media sources, You Tube videos, podcasts, online lectures, magazines, and/or they make clear that they're using their own knowledge of a subject, eg on the History Club, Royal Family, and Feminism: Sex & Gender Discussions (aka FWR) boards.

in that case then i can respect that

If posts are written in AI it makes me wonder whether the poster is a person or a bot.

If they come back & say yes, I used AI to write this post because reasons, it’s helpful.

thats also a fair point

From one Kitty to another 😸

Meow x

This isn’t about whether AI is boring, whether AI drafts a post in a certain way thus making it read differently.

This is about the erosion of people’s ability to communicate. Now we’re essentially saying that nobody even needs to learn to read and write, they can just dictate into an AI bot and a script will emerge. One which they won’t even have to read, because people blindly trust AI and mostly don’t bother.

It’s about creating an unemployable generation because they lack communication skills and associated inteligence.

It’s about reducing the job market because we won’t need secretaries, administrators, anyone who needs to communicate in any way because AI will be doing it for them.

And that’s before we look at the jobs which are already being performed by AI such as the manual and manufacturing jobs.

And what then? What’s going to happen to the mass unemployed? Who have 0 skills.

We need to stop this now because it’s already gone too far.

And nobody should be saying that writing general posts, text messages, emails, letters in AI is harmless. It absolutely isn’t.

You've articulated exactly what I think about AI.
People are losing the art of communicating.

Agree with a lot of posts on this thread ,basically if someone hasn’t bothered to write something why should I bother to read it? I say this as a ND writer who has published books for big 5 publishers but can find SPAG challenging sometimes.

I refuse to take the easy route and just rely on AI whether it’s mumsnet, writing an article or writing a book.

I would prefer a post with poor grammar, long and rambling (as many of my posts are!) than to read A.I.

It’s just so completely unnecessary, I was talking about writing a formal complaint to a company and a friend suggested I use chatgp and I was like…why? To save a bit of time?

It’s better and more cathartic for me to write it out myself! Especially as I’m a writer!

Recently a book was pulled from the shelves by a major publisher because an author who had originally self published her book, before signing a book deal to publish it traditionally , had apparently been caught out as having relied heavily on A.I.

I have had people in work casually suggest we use A.I. for tasks we’ve always done ourselves, with no reflection on the implications or ethics of it.

It all seems so nonsensical and dystopian.

I'm so fed up of getting AI prompts. I read a message of a few lines and the interface suggests I get an AI summary. I go on facebook and it suggests I use AI to help me write a post. For just a few lines on social media! God help us all. And the planet we are destroying with this crap.

Absolutely. I hate it when someone posts about a topic or a question, and someone replies with, "I asked ChatGPT and it said..." What the hell is the point of that? The poster wants people's opinions (or factual answers from knowledgeable people, sometimes) and that is why they posted. If they'd wanted to ask ChatGPT they could have done that themselves.

@daisychain01has said it perfectly!

but they said similar things when the computer replaced the typewriter

Exactly, it’s ridiculous.

it depends on the prompts and the custom instructions on chatgpt that can alter how a question is answered so one persons ask may give a different persons response etc

Apart from the fact the problems with forums full of AI slop, there’s also the companies behind these tools. How much power do we want them to have I would say zero but it’s probably too late for that!

Open AI (ChatGPT) is being boycotted now due to reported links with Trump/ICE/the US military. I don’t have time to look for sources for all of this but here’s one about slight changes to its agreement with the military : https://www.bbc.com/news/articles/c3rz1nd0egro

Anthropic seems to have the moral high ground at the moment (at least that’s how it’s portraying itself) but it’s not about the individual companies anyway, it’s the whole concept of giving up basic skills to tools whose main purpose is to make a few people rich.

I find it especially annoying when the OP has started with a well written, heartfelt post and then there are long ai replies by other posters. I do wish there was some sort of ban.

Why are you arguing in favour of posters using AI instead of writing for themselves, @ThatPearlkitty?

@NoraLuka Anthropic aren’t much better if at all. There was a class action against them by a bunch of authors because they ripped off their works.

https://www.bbc.co.uk/news/articles/c5y4jpg922qo

It’s not the same as computers replacing typewriters or machines replacing handmade things during the Industrial Revolution or like any other comparison that people make but I don’t know how to explain why it’s different.