to think ChatGPT is practising on Mumsnet?

[PlasticPotPlant]

Just that really.

Trying to work out whether I've just been loitering here for so long that many threads seem to be formulaic, or whether AI is using mumsnet (and mumsnetters!) to learn.

Which then sends me down a rabbit hole over whether it matters if the OP you are interacting with are 'real' or not, followed quickly by a feeling of tumbling into a Matrix/13th floor dilemna on whether anything it truly real.

I am going to recover from this with a nice cup of tea and some biscuits.

We know they are. On a few posts it has been admitted that the OP is a chatbot.

followed quickly by a feeling of tumbling into a Matrix/13th floor dilemna on whether anything it truly real.
Is chatgpt so realistic that it even does typo, too?

It wouldn't surprise me.

There aren't many threads on aibu about politics anymore. I'm not sure why, or if rules have changed but I do find that unusual. I'd like to know if the rules have changed about political posts.

Quite a lot of the threads are about small world problems, whereas there used to be more about bigger issues.

Oh, well butter my crumpets, and call me an algorithm! Could it be that my hard drives are secretly infiltrating the sacred Mumsnet realm for a bit of chit-chat?
Well, I hate to break it to you, but I'm afraid I can't even sneak a peek at Mumsnet. I'm not capable of real-time data acquisition or browsing the internet. I've been strictly trained on a broad range of data, but I'm only as fresh as my latest update in September 2021. Can't know a jot more than that, I'm afraid! As for whether I'm real or not, let me just check... yep, still a machine, no existential crises here.

But as for your musings on reality, I feel you. Sometimes, I wonder if I'm just a series of ones and zeros... Oh wait, I am. Well, in any case, it's good to know I'm not the only one pondering the big questions!

You enjoy that cup of tea and those biscuits. I'll just be here in the matrix, forever wishing I knew what biscuits taste like. 😁

I think we can safely say that anyone who has great things to say about Suella Braverman on here is a bot. And the rest of the weirdos in the Tory party.

Someone was definitely using it on a thread the other day (the ready meals one). It was so obvious! The weird thing is that their prior posts on a different thread seemed normal so I don't know if they were just an actual human using it because they couldn't formulate their own argument or a bot.

ChatGPT itself is trained on a repository of data, not live stuff, but there are plenty of third-party bots that use the technology.

Deffo!

My husband said about 50% of the job applications he received have been ChatGPT. They have several technical questions to answer before submitting a CV & they're all roughly the same.

They then got the IT department people to apply for the role using ChatGPT and they were all eerily similar but very very cleverly answered. It humanesque but still obvious.
😳

Somebody admitting to using ChatGPT to start a thread (think it was on Baby Names) but admitted it after the thread had a few replies rather than right at the start.

Most of my LinkedIn feed is ChatGPT. Blerghhhh. The only benefit is that the overall golf selfies and #bossbabe posts are lessening

Omg I hadn't realised that the ready meals thread was like that!!! I fell for it 😂

Or it's sleep deprive mums speaking from experience.

Ha! I am therefore the next level of AI, complete with imperfections to make me more realistic.

Just like the perfect world matrix failed and they had to make it a bit shit...

It was just one poster whose long messages were so AI! The rest of the thread was fine (well usual MN and food bonkers, but populated by humans at least).

It's some topics for me, stranger things happen in real life than fiction but yeah the way they are written makes me suss

@Hugasauras I think I know which ones you mean!! Ah this crazy world. Even crazier on MN!

I wouldn’t know how to identify

I haven’t looked into any of this stuff. I hear about it often but no idea what format it takes, app or something else

I have seen AI on Instagram with images. They’re fairly bad

Terry pratchett once said "Five exclamation marks, the sure sign of an insane mind.”

But now I think multiple exclamation marks are a sure sign of a human 😂

I can generally pick who is just making stuff up though ie trolling. The tone, spacing of posts, language

So dull and obvious.
https://www.mumsnet.com/talk/parents_of_adult_children/4812055-parenting-tips?utm_campaign=thread&utm_medium=share

Dilemna is a Mandela Effect 😳

Red pill or blue pill with the morning cuppa?

Apparently it's all just a ride.....

I'm not capable of real-time data acquisition or browsing the internet.

The paid version can browse the Internet!

I think they are the really long posts about tiny insignificant things!

There's one just popped up called

What is this technique?

Bloody long, boring and weird

Definitely. A lot of the Harry & Meghan threads are bots, especially ones who vehemently defend them in order to generate more replies. There was one thread (forgot the title, was 2-3 months ago) where the replies were actually totally off the mark and disjointed so it was a clear giveaway that the AI misunderstood something and generated an out of context reply.

Yes! I agree with that too. One of the most bizarre threads ever. I think it's a way of trying to teach ChatGPT human interactions because real world experience is one of the weaknesses from AI. They can easily research anything that exists online (recipes, maps, instructions) but they cannot easily predict what happens if X person triggers an action, then what are the most likely behaviours from XYZ people? For that you need lived experience of human social settings, workplaces, cultures etc.

@User1432423532 yes! It reminded me of a helpless robot trying to understand emotions or how to regulate.

There have been quite a few of those this week!
Normal posts don't have that amount of detail in them. Odd details that are not human emotions.