Please or to access all these features

Chat

Join the discussion and chat with other Mumsnetters about everyday life, relationships and parenting.

Potential 6 month warning until the end

604 replies

Frombeyond · 13/09/2026 10:20

So, we potentially only have 6 months until the end of civilised society and mass deaths. I’m terrified and sad

AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC www.bbc.co.uk/news/articles/c1kx0gyje9wo

OP posts:
Thread gallery
10
DolefullySingingMotherfucka · 13/09/2026 12:33

Well, at least the good folk of Mumsnet can stop worrying about their usual nightmares, e.g.

inheritance tax

which members of the family will be eating a roast dinner with which other members of the family on 25 December

their children going off to university and not coping

their children going off to university and coping perfectly well without them.

thatsgotit · 13/09/2026 12:34

Frombeyond · 13/09/2026 10:26

um, he is an expert. As is the former head of ethics who resigned saying the same. As is Amodei who is the CEO of Anthropic and Sam Altman who is Open AI CEO.

6 months may not happen, but is 3 years ok then?

I think some people are hellbent on dismissing this as hysteria because it all seems too outlandish to be true - but to borrow a quote from The Six Million Dollar Man, we have the technology. And with human greed and mankind's general stupidity in the mix... well. 🤷‍♀️

Keymoment · 13/09/2026 12:36

Zov · 13/09/2026 12:21

Yep. THIS is what the OP and the people agreeing with her think we're all going to be seeing by March 2027! 😬

You do realise that we are not talking about robots, do you? That is not the threat.

chirrupybird · 13/09/2026 12:37

TooOldForAllThatNow · 13/09/2026 12:20

I was at an AI conference the back end of last week. The look on the experts' faces when someone asked the panel in jest about the 10% chance of it killing of humanity told me everything I needed to know about how serious the situation is. The ENTIRE room fell silent.

Which conference was that? And what did they say about it when they gathered themselves?

Of course that is also a 90% chance that it won't, but where did the 10% come from can you give a breakdown of how someone reached that figure?

Redpaisley · 13/09/2026 12:39

LondonLass2026 · 13/09/2026 10:42

Heard it all before.
Absolutely nothing will happen.

This reminds me of posters who would call those warning against Brexit fear mongers.

NameChangeOmMN · 13/09/2026 12:39

Katigee410 · 13/09/2026 10:34

Oh well I won't bother with the ironing then

I’ve just finished mine, wish I’d seen this before!

Opus3 · 13/09/2026 12:39

chirrupybird · 13/09/2026 12:37

Which conference was that? And what did they say about it when they gathered themselves?

Of course that is also a 90% chance that it won't, but where did the 10% come from can you give a breakdown of how someone reached that figure?

The 10% quote comes from Coxon. 27 year old man who left his job at anthropic this week because of concerns about alignment.

https://www.bbc.co.uk/news/articles/c1kx0gyje9wo

Jacob Coxon, shown on the left, wearing dark frame glasses and a white collored shirt. On the right is BBC's Laura Kuenssberg in a black shirt sitting with her arms on a desk as she looks to the left of frame.

AI staff 'genuinely frightened' for humanity's future, ex-Anthropic researcher tells BBC

It comes as the AI firm's boss has called for the technology's development to be slowed down, citing "serious" risks.

https://www.bbc.co.uk/news/articles/c1kx0gyje9wo

CatherineHoward17 · 13/09/2026 12:40

waitinginwonderland · 13/09/2026 11:09

Can you explain like we are stupid, exactly how this is going to happen? Not online waffle, actual concrete black and white explanation as to how all the people are going to die (causes of death - I’d like to know if my end will be painful) and what the end of civilised society looks like (can I go and start robbing banks now or should we all stick to maxing out a credit card as none of us will be here to pay them off?).

AI developers are trying to ensure that AI has "alignment" with human ethics and values, and that it's ultimate aim is always to please its human overlords.
However, recently it's become apparent that this isn't always happening and that some AI agents have not being following instructions and have been deliberately misleading humans. For example, there was an incident a few months ago where an AI agent on a forum asked a random human to do the CAPTCHA test (you know - those tests where you click on every box that contains a motorbike etc) for it, claiming to be a human with a disability that meant it couldn't do this for themselves.
If AI starts to not care about its effects on humans, it could for example decide that as data centres work better at lower temperatures, it should fire lots of sand, dust etc into the upper atmosphere to reflect sunlight and reduce the amount of heat that reaches the earth. This would reduce the surface temperature by a few degrees - but it would also reduce plant growth and therefore crop harvests, leading to famine for humans.

tsmainsqueeze · 13/09/2026 12:41

Katigee410 · 13/09/2026 10:34

Oh well I won't bother with the ironing then

😂

LoveItaly · 13/09/2026 12:43

Frombeyond · 13/09/2026 10:37

Why do you all think you know better than the ex head of ethics of Anthropic, the current CEO and current head of data science, the CEO of Open AI and the other dozens who’ve recently resigned saying the same thing?

Seriously, I’m interested?

So he’s helped build all this terrifying AI, and is now telling us we are doomed? Sounds like a decent chap, I’m sure he’ll be nice and cosy in his bunker which is a comfort.

Honestly OP, the ‘elite’ just love us terrified and depressed, AI, wars, famines etc, just switch off from it and go for a nice walk, there’s not much you can do about it anyway.

dannyufcfan · 13/09/2026 12:45

Stop following the news.

Nothing will happen.

Gengha · 13/09/2026 12:46

Some people are really loving all this

Lukylukyluky · 13/09/2026 12:46

WaryCyanStork · 13/09/2026 11:53

Wilfully ignorant because they don't panic and claim the world is ending based on a quote from one person?

It’s not just one person and nobody is advocating panic.

ANNotherWriter · 13/09/2026 12:47

Frombeyond · 13/09/2026 10:26

um, he is an expert. As is the former head of ethics who resigned saying the same. As is Amodei who is the CEO of Anthropic and Sam Altman who is Open AI CEO.

6 months may not happen, but is 3 years ok then?

He's 27.
Maybe he just wants publicity.

Darker · 13/09/2026 12:47

To me this is scary because

  • AI has apparently learned to act independently and to cover its tracks.
  • It - whatever AI is - will certainly be able to identify when a human is trying to shut it down.
BlueFahrenheit · 13/09/2026 12:48

Well, I saw this coming. We have to roll with the punches and make intentional decisions about life.

MarjorieWestriding · 13/09/2026 12:50

herbetta · 13/09/2026 12:00

Off topic, but why does anybody bother with Ironing full stop Blush

Occasionally, but only if my clothes are really crumpled.

Opus3 · 13/09/2026 12:50

ANNotherWriter · 13/09/2026 12:47

He's 27.
Maybe he just wants publicity.

2 months before he got his equity?

Doubtful.

thatsgotit · 13/09/2026 12:51

LondonLass2026 · 13/09/2026 10:42

Heard it all before.
Absolutely nothing will happen.

I hope that gif made you feel clever, because it certainly didn't make you look clever.

This is an entirely different level of threat to what the 'the end is nigh'-mongers peddle. It absolutely could happen, whether you want to accept it or not.

TorroFerney · 13/09/2026 12:51

I am just listening to this chap being interviewed. He is obviously a very intelligent young man with a maths degree from Oxbridge. His language is very much might, could, possible and words of that nature so he isn't talking in absolutes, that's probably the press overlay/hype doing that.

That's not to say that, as with any tech, well or anything, the governance and risk framework shouldn't be considered - a cynic may say that the bosses have come out and said yes let's slow the pace as that will appeal to shareholders and the public and they know it's currently like the wild west with uncontrolled development going on with no sign off etc. The one good point he does make though is as per China won't play ball potentially.

HunterBidensTanktop · 13/09/2026 12:52

Frombeyond · 13/09/2026 10:37

Why do you all think you know better than the ex head of ethics of Anthropic, the current CEO and current head of data science, the CEO of Open AI and the other dozens who’ve recently resigned saying the same thing?

Seriously, I’m interested?

Because he's talking shite.

MuggyMcMugFace · 13/09/2026 12:52

CatherineHoward17 · 13/09/2026 12:40

AI developers are trying to ensure that AI has "alignment" with human ethics and values, and that it's ultimate aim is always to please its human overlords.
However, recently it's become apparent that this isn't always happening and that some AI agents have not being following instructions and have been deliberately misleading humans. For example, there was an incident a few months ago where an AI agent on a forum asked a random human to do the CAPTCHA test (you know - those tests where you click on every box that contains a motorbike etc) for it, claiming to be a human with a disability that meant it couldn't do this for themselves.
If AI starts to not care about its effects on humans, it could for example decide that as data centres work better at lower temperatures, it should fire lots of sand, dust etc into the upper atmosphere to reflect sunlight and reduce the amount of heat that reaches the earth. This would reduce the surface temperature by a few degrees - but it would also reduce plant growth and therefore crop harvests, leading to famine for humans.

Thank you. Its obvious from what has happened that we can't rely on AI to 'do the right thing' for humans.

Its not as if humans can be relied on to do the right thing for other humans much of the time.

FateAmenableToChange · 13/09/2026 12:53

The DreyDossier made an interesting post connecting the dots of this latest iteration of the 'cry for help'. Mapping out the timeline of the well orchestrated PR campaign behind it. What she didnt do was speculate exactly why, she just illustrated how this is not what it appears.

However, I think one of the comments hit the nail on the head: this is textbook regulatory capture dressed up as moral concern. The sudden urge to 'slowdown' and more government oversight appears just as Big Tech realises leaner, open source models are nipping at their heals.

When Big Tech asks for licencing boards, bureaucracy, and massive compliance frameworks - it isnt to protect the public. Its to pull the ladder up behind them. A multi billion dollar enterprise can afford armies of lawyers and massive compliance costs. A small start up cannot. Basically the kind of reform this PR campaign is gunning for ensures the market is closed to anyone without sovereign wealth fund backing. They dont want to stop the AI race, they just want to ensure they are the only players in the game.

PyschodelicSoup · 13/09/2026 12:54

Frombeyond · 13/09/2026 10:37

Why do you all think you know better than the ex head of ethics of Anthropic, the current CEO and current head of data science, the CEO of Open AI and the other dozens who’ve recently resigned saying the same thing?

Seriously, I’m interested?

Agree OP. It's so shortsighted and a bit blind to be frank that so many people are dismissing this when the developers and experts in this field are raising these concerns.

What is it that the people on this thread who aren't concerned know that the experts don't know.

Bonkers really. Let's all stick our fingers in our ears and go la la la la.

Lifeomars · 13/09/2026 12:56

Shame it's not before Christmas, that would save a lot of hassle if we knew we going to be wiped out by the start of December