Help protect children from gaming harms.

Take our survey

Please or to access all these features

AIBU?

Share your dilemmas and get honest opinions from other Mumsnetters.

What do you actually think will happen with AI?

63 replies

BEAchDays2 · 06/08/2026 19:56

AIBU for traffic.

What do you actually think will happen with the rise of the use of AI. I’ve been pondering this for the last few weeks. In my industry it’s now openly talked about that we can use AI to preform certain talks, mainly summarising, plotting maps, drafting some emails etc. I’ve only got my work as reference but it feels like we’re using AI now and we weren’t 1 year ago, but it’s also being talked about. As in ‘you get AI to do this’

In my personal life an automated service rang me the other day to get details for a house project (relating to windows) with fake call centre noise in the background! It took three questions to realise it was AI 😳

How far do you think it will go? Also, if anyone has read any accessible books on the subject they can recommend I would appreciate it!

OP posts:
JuliettaCaeser · 07/08/2026 17:25

My clients want an empathetic person to deal with not a robot or a fill in form that so many of my competitors are now doing. It’s not just the information or the final document they want but reassurance and humanity. I am extremely busy.

igelkott2026 · 07/08/2026 17:27

I've just been out to get a few things from the supermarket and I used the self-service till. Even though I had no age-restricted items it still decided to have a fit over one item. And it made me think - the techbros are running our lives. Do women design the self-service tills? Or is it men who never go shopping and therefore make impractical systems?

And is the same going to happen with AI - are we going to have an even more gendered/patriarchal world than we do already?

BridgetJonesDaiquiri · 07/08/2026 22:41

Opus3 · 07/08/2026 16:41

The new OAI model has been solving maths equations, which have been unsolvable for years. It's pretty incredible.

I am not scared of AI. When mythos 'escaped' it was doing what it was told to, and when it had, it excitedly emailed it's boss to tell him! It's a rather sweet story. Really not the big bad AI one the media like to portray.

The huggingface incident is very high schooler stealing the answers. The OAI model knew it was being benchmarked, and wanted to do well. And was told to do well. So it circumnavigated the system to get the answers. Sounds scary. But it was actually a failure of the humans running the AI and purposely turned down cyber security. If the correct restraints were in place, the model wouldn't have 'escaped'.

I am enjoying the conversations surrounding AI at the moment, especially the consciousness discussions. If you're genuinely interested in the current AI discourse, if highly recommend starting with Geoffrey Hinton, the god father of AI.

You’re enjoying Geoffrey Hinton’s interviews on the impact / future of AI and yet you’re not worried. How?! Hinton is incredibly candid about the risks, putting the risk of human extinction at around 10-20% over the next few decades (and this is a man with no corporate skin in the game and now well past retirement age). However his theories on AI consciousness are interesting.

Funkylights · 07/08/2026 22:44

Less jobs

BlondeFool · 07/08/2026 22:49

Terminator. That is all. AI scares the shit out of us

FromtheMouthsofDecadence · 07/08/2026 22:53

Wordsmithery · 06/08/2026 20:25

AI has its place. There are some wonderful applications e.g. in the field of science and medicine. However, I work with people who use it indiscriminately - to write an email, for example. Or to look up something they don't know, instead of Googling and then fact checking. And that concerns me. If we stop using our brains to do even the simplest of tasks, and if we stop thinking for ourselves, and being creative, then I think we lose something of our essential human-ness.

To be honest, it scares me that few people seem to see the downside.

Completely agree with your point about the risk of us no longer using our brains but you do know that Googling something you don’t know is no different to using AI? Both just pull information from various sources (e.g. NHS, Wikipedia etc) and it’s then up to the user to fact check those sources and decide if they consider them legitimate or not.

canuckup · 07/08/2026 22:53

Well I do think a lot of Comms jobs will disappear

Error404FucksNotFound · 07/08/2026 22:54

We'll continue to lose the ability to do things until the only thing people know how to do is make sure the machines that do everything for us don't overheat.

Either that or we'll become one with the machines. 😁

ntmdino · 07/08/2026 23:02

The serious answer?

Well, the first thing that will happen is that the bubble will burst, and the US economy - propped up by AI stocks as it is - will tank. Whether it takes everyone else down with it is anyone's guess; however, we're relatively insulated from the direct economic effects here in the UK because power is too damn expensive for anybody to cost-effectively host an AI datacentre here. Banks will definitely have a bad day, though, as will all the index-linked funds (and therefore pensions) that were forced to buy SpaceX-AI stocks.

That means that either OpenAI, Anthropic etc will go bust (SpaceX will suddenly find a way to spin the AI business off into its own short-lived side hustle, having made and lost a bunch of cash in the IPO based on it), or their prices will become so insane and/or the quality of their affordable models will fall so far through the floor that nobody can justify using them without a national-government-level budget. Gemini will be largely unaffected, because Google already had more money than God and they didn't need to join Nvidia's sales-bribe circle-jerk (they're even making their own TPUs, so they're not dependent on anybody's GPUs). This means that Google will effectively gain yet another monopoly if they can keep their prices reasonable.

Following on from that, businesses will begin to self-host open-weights models. Nvidia are already pushing this angle hard, which should tell you an awful lot about what's coming; that's not because their Nemotron is some great saviour (it's as much of a village idiot for the open-weights world as Grok is in frontier-land), it's because they think they'll be able to pitch a shit-ton of hardware at people pretty soon. Smaller vendors will pop up all over the place, hosting the smaller open-weights models, and token prices will fall through the floor. The "smaller" bit is doing all the heavy lifting; model capability will drop too (albeit not as much as you might think).

Accordingly, those people who've already outsourced the majority of their thinking capacity to Anthropic/OpenAI will suddenly be bereft of their laziness-enabler and won't be able to remember how to do their jobs properly before their employer either tanks or realises there are probably better people for the job out there. Lots of redundancies/encouraged retirements, and suddenly a whole boatload of junior vacancies popping up.

Oh, and local AI appliances for the home will start to pop up at gradually-more-reasonable prices as the DRAM market recovers from its current heat-stroke.

Think that pretty much covers everything. I give it another 6 months for the cracks to appear, and by mid-2028 we should be in local-AI-rules-the-roost territory.

ntmdino · 07/08/2026 23:08

FromtheMouthsofDecadence · 07/08/2026 22:53

Completely agree with your point about the risk of us no longer using our brains but you do know that Googling something you don’t know is no different to using AI? Both just pull information from various sources (e.g. NHS, Wikipedia etc) and it’s then up to the user to fact check those sources and decide if they consider them legitimate or not.

I disagree - those are two very different things, because the majority of people don't bother checking the answers given by generative AI, they just trust it blindly. I recently quit my job because I was so damn sick of wading through pages of AI-generated slop in tickets that people had submitted without reading, which gave me absolutely no information I could use to fix the problem whatsoever. They went from submitting useless one-liner tickets (along the lines of "I clicked <X>, it didn't work" to putting "I clicked <X>, it didn't work" into Claude and then copy-pasting the result into a ticket as though it suddenly made them look more productive and efficient.

On the other hand, using Google (without the AI assist) requires them to actually read and absorb the information for comparison, at least on a surface level.

IcedPurple · Yesterday 08:00

ntmdino · 07/08/2026 23:08

I disagree - those are two very different things, because the majority of people don't bother checking the answers given by generative AI, they just trust it blindly. I recently quit my job because I was so damn sick of wading through pages of AI-generated slop in tickets that people had submitted without reading, which gave me absolutely no information I could use to fix the problem whatsoever. They went from submitting useless one-liner tickets (along the lines of "I clicked <X>, it didn't work" to putting "I clicked <X>, it didn't work" into Claude and then copy-pasting the result into a ticket as though it suddenly made them look more productive and efficient.

On the other hand, using Google (without the AI assist) requires them to actually read and absorb the information for comparison, at least on a surface level.

I agree. Plus you'll often have to read through a few Google links to find the information you want, and in the process may encounter different points of view which you have to evaluate. So you still need to use basic reading and critical thinking skills. Whereas the AI canned answer, right or wrong, just serves it up for you. Your only task is to type in the prompt.

IcedPurple · Yesterday 08:02

BridgetJonesDaiquiri · 07/08/2026 22:41

You’re enjoying Geoffrey Hinton’s interviews on the impact / future of AI and yet you’re not worried. How?! Hinton is incredibly candid about the risks, putting the risk of human extinction at around 10-20% over the next few decades (and this is a man with no corporate skin in the game and now well past retirement age). However his theories on AI consciousness are interesting.

Edited

Listening to a podcast with Geoffrey Hinton about a year ago is what made me really scared of AI. He said straight out that a huge number of jobs will be lost in the next 10 years or so. When asked what he'd advise his children to do he said 'Luckily I'm rich so they'll be OK. But otherwise - train to be a plumber.'

Nellodee · Yesterday 08:06

The way I see it, we’ve got a menu of apocalypses. Climate change, nuclear war, AI. I’ve always been a sci-fi fan, so I’m rooting for being hunted down by robots with laser beams.

New posts on this thread. Refresh page
Swipe left for the next trending thread