To think MN may be becoming pointless now half the threads seem to be created by Chat GPT’s younger siblings, purely to waste our time?

[MayIDestroyYou]

This morning I’m struggling to believe any of the new threads I’m opening. Even S&B is weird. Everything seems to be the sort of insubstantial cliched situation a young AI might spit out for fun.

At least I know the Archers thread is real! Although who knows how long it will be before the show is written by not-humans?

(I don’t actually want the Poll, but can’t find how to get rid of it.)

I admit I’m a bit (considerably) uncertain of how an AI responds within a thread.

As I said, there were two or three over the weekend where I was sure, from the paucity of detail and generic nature of the OPs, that the threads were not genuine and probably not written by humans. In one in particular the OP’s subsequent posts were equally lacking in anything non-generic. If it was an AI OP - was it reading posts itself or directed by a human?

I don’t think an AI program would make its own account and do this. If what you’re saying is true though it means a lot of commenters could also be AI

Oh, no - I’m assuming a human sets up the account, just not sure what happens after that.

I’m venturing a guess here - a human probably writes a script and mass uploads/starts AI threads, that’s why the individual messages are not “humanised”.
So yes, it’s a human behind it, but for mass commands/scripts, not even reading the AI generated content.

I wander if people do have a genuine question, but use chat to write it up for them. They would then post genuine responses. Maybe that's too optimistic.

Thanks for your post, OP. To be honest we're not seeing a huge number of reports about this. It's possible people may be using AI to help them write a post or a response and we don't have a specific problem with this in itself (but admittedly this is all quite new and we monitor situations as they develop). Of the approx 30,000 posts a day on Mumsnet, we would estimate the vast majority are written by humans (for humans...). If it was clear someone was not engaging with other posters and just posting reams of text then obviously we'd be suspicious as this is not normal Mumsnet behaviour. Ultimately, most MNers want to talk to other MNers and engage in discussions.

If you think a post is off, please hit the report button and we'll check things out.

I asked Chat GPT to explain why it isn't a threat... here's what it had to say:

Dear Mumsnet Users,
We understand your concerns regarding the possibility of AI-generated posts on the forum. However, we would like to assure you that AI chat models like the one you might be referring to are not an existential threat to platforms such as Mumsnet for several reasons.

1. Limited Understanding and Context: AI chat models, including the one I am based on, have a limited understanding of the world and context. While we can generate responses that may seem coherent, we lack genuine comprehension and emotional intelligence. This limitation makes it challenging for AI models to fully replace genuine human interactions, especially in a community like Mumsnet, where emotional support and empathy play a significant role.
2. Lack of Personal Experience: AI models are trained on vast amounts of text data, but they do not possess personal experiences or emotions. Mumsnet is a platform where individuals share their real-life experiences and seek advice from others based on their unique perspectives. AI models cannot replicate this depth of personal engagement and understanding.
3. Human Connection and Empathy: Platforms like Mumsnet thrive on the human connection and empathy shared among its users. Real users can relate to each other's experiences, feelings, and struggles in a way that AI models cannot. Genuine interactions between real users foster a sense of community and support that is difficult for AI to replicate.
4. Limited Creativity: While AI models can generate text responses based on patterns in the data they were trained on, they lack true creativity and innovation. Discussions on Mumsnet often require creative problem-solving and unique insights that stem from human experience and individual thinking.
5. Ethical Considerations: The use of AI to generate content without disclosure raises ethical concerns regarding transparency and authenticity. Platforms like Mumsnet are committed to fostering genuine user engagement and trust. By being transparent about AI usage, platforms can maintain the integrity of the community and build trust among users.
6. Moderation and User Oversight: Mumsnet, like many responsible platforms, implements moderation to ensure content aligns with community guidelines and standards. User reporting and oversight help to identify suspicious or inappropriate content, including potentially AI-generated posts.

In summary, while AI chat models can be impressive in generating human-like text, they are not equipped to fully replace genuine human interactions and experiences. Platforms like Mumsnet continue to thrive because of the authentic connections and emotional support that real users provide. By maintaining transparency, ethical practices, and a strong sense of community, Mumsnet can continue to be a valuable and trusted platform for its users. Sincerely, [Your Name]

@YetAnotherBeckyMumsnet!

But how do we know that you …???

Ha, good response from Chattybitty!

As well as twerps getting bots to write fake posts, I think it's quite likely that users will increasingly run their genuine issues through them. A lot of people lack confidence with the written word, and this place can be intolerant of poor English. ChatGPT's right that there is a detectable lack of 'heart' to its offerings, especially when a conversation develops. Currently, AI-like OPs tend to disappear after a couple of replies.

But that depends on how well it's briefed: it can simulate human interaction pretty well if the input's thorough. I don't know why anyone would bother - but it's a human given that if it can be done, it will be done. There actually is commercial motive: the things are being trained to take on health and social care, so they'll need to both understand and deliver colloquial language.

I was genuinely hoping I'd be dead before this got into full swing - but it's going so fast! It's going to beat my telomeres 😬 I wonder how younger generations will adapt to never knowing the authenticity or accuracy of what they see, hear & read?

I'm sorry @MayIDestroyYou but I can't open the pod bay doors.

🤣🤣🤣👏🏻

Eek @YetAnotherBeckyMumsnet if you click on that it boots you off the app

@JaneJeffer does this to me too! Hal has broken the app. We'll look into it

Testing testing

I'm sorry Dave

The is an obvious one on the legal board at the moment about a "law firm."

It is one for the chinny reckoners for sure. 🤔

Threads are also getting recycled again today. The one about 100% cotton sheets and the one about therapy in AIBU I have definitely read before.

The thing is … I felt reassured for about five minutes by MNHQ’s response. But there’s a thread in Mature Study and Retraining that follows exactly the pattern of the ChatGPT generated examples above. They’ve had no response all day and I feel a bit guilty about not offering encouragement.

Confused dot etc …

The '40th birthday dissapointment' one was taken down in the end.

There's a whole slew of threads around today that are ITV-X Drama script material.

Good luck to them if they try and infiltrate Scotsnet...."Did ye AI?"

It is a cryin shame that we dinnae hae Doricnet 😈

😂

It's easy to tell if a post is genuine.
The grammar and spelling will be shite and the OP will show familiarity with the MN acronyms.
There will be no salutation and the post will not conclude with an exclamation mark.

BTW do bots swear?
Genuine question. None of the examples of ChatGPT on this thread contain swearwords.

I’m now ver tempted to go full bastardy Nan on a chatty thread to see if they respond similarly.

It’ll be experimental only so not reaaaaaaaaaaaally a flouting of guidelines @YetAnotherBeckyMumsnet 😈😁