Soft play showdown

[ThisWildHedgehog]

Sorry this is long; if you make it to the end, you deserve a medal for patience and probably a biscuit. 😂

I took my toddler and baby to soft play yesterday to meet a friend with her two little ones, who are around the same ages.

Now, before anyone tells me “you knew what you were walking into…” YES. I absolutely did. I understand that soft play is essentially a mum’s version of full-contact dodgeball, complete with sticky hands, rogue socks and the desperate hope that ear defenders will be a trend for fall 2026…

HOWEVER, yesterday was particularly unhinged.

We were all sitting on a baby sensory mat when my baby started crawling around. A two-year-old girl came over and immediately grabbed him by the leg from underneath him, causing him to flop onto the mat. Thankfully, he was completely unbothered. I picked him up and she came back and tried to grab his leg again. This time I was holding him, so I moved him out of reach and said, “That’s not kind. We don’t pull other people.”

She then came over again and tried to kick my friend’s baby. We blocked her because, by this point, we were starting to see a pattern.

A little while later my older son was happily shovelling sand from one section of the indoor sandpit to another when she grabbed the shovel and tried to wrestle it away from him. I stepped in again and said, “We don’t snatch. That isn’t kind.”

At this point, I was getting a little fed up with spending my playdate supervising a child who wasn’t mine, so I went to find her mum. She was stood about 30 metres away, around a corner, in the café area.
I approached her and said, “Have you got a daughter in a pink top?”
She immediately replied, “Oh no, is she being naughty?”
I said, “I think she’s needing some supervision.”
She asked what had happened and I just said, “She just needs someone with her.”

So Mum came over and stood outside the sandpit.
In the most monotone voice I have ever had the displeasure of hearing, she repeatedly said:
“I told you no sandpit. Come here.”
The child completely ignored her.
This continued for what felt like the length of the first Avatar film.
Eventually another mum in the sandpit actually offered, “Did you need someone to pass her to you?”
Honestly, I nearly applauded.
Eventually Mum got hold of her and wandered off muttering “You are a good girl” to her.

Fine. Situation over. We can all go back to enjoying our children.

Except… apparently not.

We then noticed Mum going around speaking to the owner, then the mum who had helped her retrieve her daughter and, finally, she approached my friend.
It was essentially an investigation into the alleged crimes committed by her two-year-old.

She wanted to establish exactly what her daughter had supposedly been accused of, followed by the declaration that “her daughter would never do those things.”

Okay. 🤷‍♀️

Nobody was trying to cause a scene. We just wanted her to stop kicking/grabbing/snatching from our children.

But then she came over to me and repeated the same bizarre assertion, despite not having witnessed any of it, that she was absolutely sure her daughter hadn’t done anything.
At this point I said:
“If you had been watching your child, you wouldn’t have to interview the entire soft play to find out what happened.”
She replied:
“I can’t watch her 24/7.”
She also said “it’s hard to teach a two-year-old right and wrong.”

Yes. I agree.
That’s generally why we… supervise them and teach them.

The conversation was becoming heated and I didn’t want to stand there arguing in front of my children, so I ended it with:
“I politely came to get you, I was respectful and calm with you, and you are making a mountain out of a molehill. Please leave us alone.”

And then came her closing statement:
“Well, I won’t have my daughter apologise for something she hasn’t done, so I take back our sorry.”

There was never even an apology to retract…

At this point we just wanted to get on with our playdate without either her or her daughter bothering us.

The owner then came over, apologised to us, invited us back for another playdate and explicitly said that, of course, children need to be supervised.

And the other thing that really struck me was the layout of this place. There were loads of stairs, raised platforms and areas where a two-year-old could quite easily get herself into trouble.

So even putting aside the behaviour towards other children, I genuinely don’t understand how you can be 30 metres away around a corner while your two-year-old is roaming around a large soft play independently.

I completely accept that two-year-olds grab things, snatch, push, kick ect. I don’t expect toddlers to behave perfectly. I do expect their parent to be somewhere nearby when they’re doing it, though.

Right, where are those biscuits? 🫣🤣

Just out of curiosity, I posted OP's post in ChatGPT, told it I was suspicious that the post was AI, and asked if it agreed. This is how it responded:

Yes — I think your suspicion is quite reasonable, with the important caveat that none of these features can establish AI authorship on its own. What strikes me is the combination of them, and particularly the way the post seems to have been constructed to sound like a highly polished “Mumsnet anecdote”.
A few things jump out to me.
1. The prose is unusually
performatively conversational
There is a very recognisable AI pattern of writing that goes:
“Now, before anyone tells me X… YES. I absolutely did.”
followed by:
“HOWEVER…”
and then a deliberately comic metaphor:
“soft play is essentially a mum’s version of full-contact dodgeball”
and another:
“the desperate hope that ear defenders will be a trend for fall 2026…”
This is extremely polished simulated internet-person voice. It’s not impossible for a human to write this way — plenty of people do — but there is a slightly uncanny quality to it because every paragraph seems to contain a consciously selected conversational device.
The opening alone has: disclaimer → anticipated objection → emphatic rebuttal → humorous analogy → topical aside → transition into story.
That’s very AI-ish architecture.
2. The bolding/emphasis is particularly interesting
The use of bold/typographical emphasis in an otherwise supposedly spontaneous account is one of the things I’d notice too. AI-generated social-media prose frequently has strategic emphasis designed to ensure that the reader hears the intended tone.
And this post has an almost suspiciously complete repertoire of “humanising” devices:

• 😂
• 🤷‍♀️
• 🫣🤣
• “Okay.”
• “Fine. Situation over.”
• “Except… apparently not.”
• “Honestly, I nearly applauded.”
• “Yes. I agree.”
• “Right, where are those biscuits?”

It is almost too good at sounding like a particular genre of Mumsnet poster. 3. The dialogue is very conveniently good This is probably one of the strongest things for me. The OP supposedly remembers a fairly extended confrontation, including: “I think she’s needing some supervision.” “If you had been watching your child, you wouldn’t have to interview the entire soft play to find out what happened.” “I can’t watch her 24/7.” “it’s hard to teach a two-year-old right and wrong.” “I politely came to get you, I was respectful and calm with you, and you are making a mountain out of a molehill. Please leave us alone.” “Well, I won’t have my daughter apologise for something she hasn’t done, so I take back our sorry.” That is an extraordinary amount of perfectly quotable dialogue, all of which conveniently advances the OP’s narrative. Real arguments are usually much messier. People interrupt one another, repeat themselves, forget the precise wording, say things that don’t make much sense, and remember the gist rather than producing a transcript several paragraphs later. AI is very good at manufacturing dialogue in which each participant says exactly the thing necessary to make the story work. And “I take back our sorry” is almost too narratively perfect. It is basically a sitcom line. 4. The physical geography is indeed odd This is the bit where I think your observation becomes more interesting than simply “the writing sounds AI-ish”. “She was stood about 30 metres away, around a corner, in the café area.” Thirty metres is quite a long way in this context. And the OP apparently identifies the child’s mother without having previously interacted with her: “Have you got a daughter in a pink top?” That’s plausible if the mother was the only woman with a child in the relevant area, or if the OP had seen them together earlier. But the combination of 30 metres + around a corner + café area makes the identification rather conspicuously convenient. Then there is the fact that the mother is somehow simultaneously sufficiently absent to allow her toddler to roam around, but sufficiently connected to events that she can later conduct a quasi-investigation involving several witnesses. It’s not impossible. It’s just story mechanics rather than natural observation. 5. “We blocked her because… we were starting to see a pattern.” This is a particularly strange sentence. “She then came over again and tried to kick my friend’s baby. We blocked her because, by this point, we were starting to see a pattern.” As you say, you would block a toddler from kicking a baby because she was trying to kick the baby. You don’t really need to establish a behavioural pattern first. The wording sounds like retrospective narrative construction: event A → event B → pattern emerges → narrator intervenes. It’s written almost as if the author is constructing evidence for a case. And that becomes even more noticeable because the post later explicitly frames the events as an evidential dispute: “the alleged crimes committed by her two-year-old” “exactly what her daughter had supposedly been accused of” “her daughter would never do those things.” The whole story has been arranged around the idea of building a prosecutorial case against the other mother. 6. The structure is very AI-friendly Look at the progression:

1. Humorous disclaimer.
2. Establish that OP understands soft play.
3. Introduce unusually badly behaved child.
4. First incident.
5. Second incident.
6. Third incident.
7. OP seeks parent.
8. Parent behaves inadequately.
9. Parent conducts investigation.
10. Confrontation.
11. Perfectly quotable “mic-drop” exchange.
12. Authority figure validates OP.
13. Broader parenting principle.
14. Humorous closing.

That’s an extraordinarily clean narrative arc. It’s basically: setup → escalation → confrontation → resolution → moral → punchline Humans absolutely write stories this way, but AI is particularly prone to producing this kind of highly symmetrical structure. 7. The “Avatar film” line is another tell “This continued for what felt like the length of the first Avatar film.” It’s not that the joke is suspicious in itself. It’s that the post contains lots of jokes of exactly this kind, at exactly the points where an anecdote needs a tonal lift. AI is very good at inserting culturally legible humour that isn’t especially personal or idiosyncratic. The humour here feels generic-but-effective: full-contact dodgeball, Avatar, biscuits, “alleged crimes”, “mountain out of a molehill”, emojis. There isn’t much of the weird, specific, slightly incoherent humour that tends to make human posts feel genuinely individual. 8. The ending is particularly “generated” The final few paragraphs are almost textbook rhetorical consolidation: “I completely accept that two-year-olds grab things, snatch, push, kick ect. I don’t expect toddlers to behave perfectly. I do expect their parent to be somewhere nearby when they’re doing it, though.” That’s essentially a concession → distinction → thesis structure. And then: “Right, where are those biscuits? 🫣🤣” A neat comedic callback to the opening. Again, a human could absolutely do this. But it’s very polished for someone ostensibly recounting an annoying incident on a parenting forum.

There is one thing I’d be cautious about
I wouldn’t say “this was definitely AI-generated.” In fact, a human could have written it, and there are Mumsnet posters who write in precisely this polished, anecdotal style.
And there is an important third possibility: human-written events that have subsequently been AI-polished. That would actually explain the uncanny quality extremely well. Someone could have written a messy account, then prompted ChatGPT with something like “make this funny and engaging for Mumsnet without changing the facts”. The underlying events would be genuine while the resulting prose would acquire all these characteristics.
There is also a fourth possibility: someone deliberately writing in the style of an AI-generated viral post. 😆
What makes me suspicious isn’t any single “AI tell”
It’s the convergence:
unnaturally polished conversational voice + highly structured narrative + conveniently perfect dialogue + frequent manufactured humour + typographical emphasis + retrospective psychological/narrative framing + questionable physical details + perfectly satisfying ending.
That combination would make me put “probably AI-assisted” considerably higher than “ordinary spontaneous Mumsnet post”.
And I think your point about reality is particularly important. AI detection by linguistic style alone is inherently shaky. But when the prose has the characteristics of generated narrative and the underlying scene contains details that feel more like the logic of a story than the messy constraints of an actual soft-play environment, the two pieces of evidence reinforce each other.

But it's not good writing thats the entire point. It's cringey, americanised (fall?) ai drivel. Which is completely unnecessary given op knew the post was already too long. What's the point?

I have a friend that talks like this when she is telling a story. She was a journalist. She definitely wasn’t using AI 15 years ago.

She didn't claim to have studied Latin.

Yes one of them is following with interest how AI is replacing human generated content on the internet, and the consequences of that for forums like mumsnet. It’s fascinating. I also enjoy reading, hiking and swimming.

I'm also tired of the frequent accusations that posts are AI. Some posts clearly are AI-generated, but often people see AI where it doesn't exist. This story feels perfectly plausible to me. I've read about many similar soft play 'encounters' on here, over the years. And if it is AI, oh well.

I wonder if anyone on here intentionally alters their writing style to avoid being accused of being a witch AI.

Using Fall doesn’t make something AI though- it’s just the way fashion writers etc express it - “New for Fall 2026, brown jackets!” Or something equally ridiculous.

It was long winded but alot of posters are.

And MN hates it when anyone tries to be funny

OR
people could just write themselves without using ai, then others wouldnt need to spot it.

It might seem harmless but it's not. There are some examples that could be 50/50 but if you work in fields where it is used constantly you learn to spot it with a very high rate of accuracy.

Chef's kiss. Perfection. Thank you for verbalising all this better than I was able to.

As I used to say to my Year 11 on Friday afternoon, Congratulations on spotting this week's deliberate error!

The Devil sent Soft Play Centres to earth direct from the inner circle of Hell to remind us all he still exists.

Nothing is ever resolved nicely at Soft Play.

I loved a soft play! Waved them off, got my phone / book out, checked up every now and then. No trouble. Never had any parent confronting me. Who knows what went on though unbeknownst to me!

As several people have pointed out, what else would fall 2026 mean if not referring to the season? If you cant even read the post properly its not surprising you cant pick up the vagaries of ai, is it? That doesnt mean that people with better reading comprehension can't.

She sounds bizarre even by softplay standards- of course she can watch her toddler all the time during a trip out... that's basic parenting.

I don't know why you didn't just tell her what her daughter was doing though? If someone was so cagey I'd assume they were either overreacting about something they knew was daft or my child had done something so embarrassing they didn't know how to tell me.

All easy to say on MN though, isn't it? We have no idea whether the other mum in question was 6'4" and built like a brick outhouse.

People don't hit the nuclear button unless they need to so a request for supervision is a fair starting point.

The confrontation doesn't have to be turned up to 11 from the off.

No she wouldn’t, she was too far away!

Sigh

I’m genuinely sorry that so many people are frustrated by the post.

I was just having fun. I love Mumsnet and read it through the night while I’m BF to keep me awake. I’ve always enjoyed posting and chatting to people before but this is just pants!

I’ve just bought a new Nescafé instant coffee flavour, I’m going to sit down and enjoy that while I read ‘I am dog’ for the 974th time to my kids. 🥰

I hope everyone else here can spend their afternoon doing something that makes them happy. ✨

Your post didn't read as perfect and nor does AI. If you did write it yourself, stop being so influenced by AI slop that's infected your style

Just finished up that fifth draft of your AIBU MIL thread?

This joke was so good you had to tell it twice.

What a hoot.

Ignore them OP and don't let it change how you post, this kind of thing is just pathetic!

It was a genuine question considering hitherto this thread - all you seem to post about is how much you don’t seem to like you mil!

It's genuinely incredibly depressing to me, an English teacher, that when something is well written people assume it's AI. You do realise that AI learns to write from human writing, yes?

OP, YANBU. I'm impressed that you shut down the conversation so quickly because I would have been inclined to argue with her!

I hope you get to feel the sun on your face today. 🌞

It seems you spend a huge amount of time reading my posts or engaging with them.

Take care x

It's genuinely depressing to me that an English teacher a) can't identify ai and b) thinks this unfunny dross is "well written"!