I have begun relying on AI and I don’t know how I feel about it.

I asked if it wasn't put under democratic control what could be the outcomes by 2030;

Great question. Let’s break that down based on population estimates and potential unemployment scenarios in the UK by 2030.
🇬🇧 UK Population Projection (2030)

• Estimated total population: ~70 million
• Estimated working-age population (16–64): ~42–45 million

• The last time UK unemployment was >10% was during the early 1980s recession.
• If AI displaces jobs faster than society adapts, we could return to or even exceed that level.

But Excel HAS replaced jobs.

I can do a complex spreadsheet in a few hours that used to take days using paper, pencil and calculator and then needed to be typed up by a typist.

Same with accounts - used to have to be drafted by pen and paper, then typed up by a typist, and "called over" by two people to check the typing and additions. Now it just pops out of the book-keeping software in seconds.

I run my own "one man band" accountancy practice on my own. I do everything from book-keeping, sorting paperwork, drafting returns, drafting accounts, printing and posting, etc. Before desktop computing, similar sized firms (in terms of number of clients, income levels, etc) would need 2 or 3 "support" staff for the qualified accountant owner. They used to be "pyramid" shaped organisations with lots of lower level staff supporting the qualified at the top - now it's a lot flatter, qualified accountants can do most things themselves because the "admin/grunt" work has gone.

So that's 2 or 3 job losses per qualified accountant! Just because all the "grunt" work has gone, mainly because of spreadsheets like Excel (although I preferred Lotus 1-2-3!).

Going back before computers, firms employed "comptometer" operators who spend all day every day just adding columns of figures on old fashioned adding machines. It was literally all they did. Spreadsheets have wiped that kind of job out at a stroke, likewise typing pools, etc.

Excel has not replaced people's jobs or is being used to replace whole industries of people. The law is not being written to keep it limited and pandora's box cannot be closed on this thing. Glad I'm middle aged already.

Thanks, @Surf2Live , that may be very useful indeed.

I'd use it almost exactly the same way as you do – to organise what I already have. It's information prose, not a great novel.

I have ethical concerns as @MistressoftheDarkSide outlines above. In fact I used to make my living as a writer. But now I don't have the cognitive powers for this, and me refusing to ever use AI even in carefully controlled circumstances will have as much impact on the rise of AI as Canute had on the tide.

I actually hate where we're going with this, and if you gave me a magic wand to take away the likely harms of AI, at the cost of not being able to use it for my document, I'd wave that wand in an instant.

I'd be surprised if, in the long run, the benefits (except to a tiny power-hoovering techno elite and their clients of choice) will outweigh the harms.

But as no such wand exists, I'll be paying the cost of the harms of everyone else using it anyway, so I might as well have some benefit from it.

It was stark, wasn't it? In China, the state will control it and harness it for the advancement of society, as it trims the wings of its billionaires. It becomes just another tool in the authoritarian toolbox one that could actually make the CCP more powerful and even more popular, as society reaps the productivity benefits.

In the capitalist, free West where capital flows freely and automation drives a competitive arms race the masses risk being left behind: redundant, useless. That’s why effective democratic or, failing that, autocratic control is absolutely necessary.

Very interesting video.

It does address the fact that in some ways we're almost engineering our own obsolescence.

It makes the point that those able to can currently take advantage of this growth phase to feather their own nests, but on a personal level even if I could, I'd be wary of doing so as inequality of wealth and opportunity is already growing at an exponential rate and society is becoming more and more divided. I don't see that ending up on a "virtual" and isolated conclave with masses of displaced and disenfranchised "others" circling the walls and being picked off by drones is a desirable life, and if managed badly, and purely for profit, that is where this could go.

The psychological and practical ramifications, and as pointed out the speed of this progress is something we have no frame of reference for.

Again, on a personal level, my one solid skill is writing. I get compliments on it and was considering monetising it possibly in the future, but now it's de rigeur to hand it off to AI, so I feel pretty stumped, and according to said video and some of the posts here, my one skill is no longer even a skill because AI is the preferred choice in such roles.

I really do think that there's an element of "Emperors new clothes" here as well.

I suppose we'll just have to ride it out - my money is currently slightly more on Dystopia over Utopia, but I acknowledge I'm an ancient cynic, so I hope I'm proved wrong.

Your job won’t exist either.

"If your job is as routine as it comes it's gone in the next 2 years" - having started to research this briefly the outcomes are potentially horrendous for our society and they are coming at us like a runaway express train.

I would never read a book mostly written by ChatGPT even if the author had thrown their jumbled memories in it somewhere

All well and good but if it can do your job in 30 seconds do you really think your job will exist soon. Re the email surely you could just click on the sender email address to look at the address and see if it’s fake? These emails often have massive giveaways that their fake even the better ones.

Today I asked chat gpt to translate a load of documents for me - they were in tiny type business speak and very dense and I could not work out what was being asked of me. I told it to please explain in very simple clear language and tell me what steps I should take next. It helped me navigate it all in thirty seconds when before I would have been peering at the screen for days and probably have given up. It also verified that a fed ex email that kept appearing in my inbox box was real and even worked out where the parcel was coming from.

Did Claude do the colours thing?

ChatGPT, paid account, which means my information on it is private and doesn't become part of the web.

It's a travel book. First, I wrote a blurb, then created a chapter outline.

Then I started writing where the memories are strongest. Writing anecdotes and observations. Telling it to put each piece into a specific chapter.

It's kinda like it's doing my filing for me. I can jump all over the place as the mood takes me and as memories surface. But the book will be organised.

I have source material to work with, I kept a vlog for years now.

@changedusernameforthis1 could you tell me which app you user to test new hairstyles on a photo of you please?

You don’t need a brain anymore. It’s now just an extension of your hands

"Im a Japanese-to-English technical translator. Ive been watching AI become more and more prevalent in my industry for the past few years until this time, last year. In January, 2024, I noted that none of my regular clients was contacting me. It went on for several months, so I started to call them to find out what the issue is. They all told me they had switched completely to machine translation - essentially AI - and didn`t need me any longer. I was put out of business overnight. Luckily, I was ready to quit and was planning my exit in a year or two - so this came sooner than I was hoping. I was able to kick freelance translation to the curb and go off and do something else in retirement, but it came and stole my lunch overnight."

I know there are perfectly sensible ways that AI can help with paint colours, but this one always cracks me up.

www.aiweirdness.com/new-ai-paint-colors/

WTAF !

en.wikipedia.org/wiki/P(doom)

Catty as I call it (ChatGBT) is a fantastic tool when used properly. It's like anything. Use it sensibly. Check your usage on your weekly online calculator of different apps you use if you are worried about usage.

I’m not sure I’d ever use it for work, but ChatGPT has been great for recipes, today I asked it for a gnocchi recipe that wasn’t tomato based and didn’t have mushrooms. It then offered a few ways to zing it up….everyone really liked it 😆.

https://www.bbc.com/news/articles/cpqeng9d20go

Interesting.....

The increase in power needed for AI & larger data centres is most definitely a real thing.

Microsoft seem to be going ahead with their leasing of one of the Three Mile Island reactors to have an exclusive source for their new AI data centre.

Small (nuclear) Modular Reactor (SMR) development & research is being bouyed by investment by the big AI & cloud service providers, with Alphabet (Google) and Amazon (for their AWS system) at the forefront together with Microsoft looking beyond just leasing a single reactor in Harrisburg.

I’m pro nuke (reactor, not weapon) so the development of more power plants with big money for zero to low carbon energy investment is win win, but don’t naïve to think that this is just like a fun addendum to helping plan this week’s meals or choosing your colours (whatever that means).

I’m definitely not anti AI but I am cautious. For every piece of information you feed that little AI prompt, you are arming it with that information you have given it. Every question, every fact, every image, every small, almost inconsequential prompt you type is providing a point for that AI to gather data, bank & learn from it.

Obviously, there is a huge potential for security breaching (make sure you have your company’s consent before you start chatting with Claude or ChatGPT at your desk) and a lot of what AI fodder is given back to you is currently utter bollocks.

Until there is a Utopian, god-like AI that is infallible (which is never), use critical thinking before you employ that answer in your work.

And in time there will be deskilling & job losses. I don’t see why or how there won’t be.

No matter what your view though, the humongous AI rhino is barreling towards us, so in our industry, it’s either be trampled by it or train to become the bloody zoo keepers who will wrangle the thing.

Here’s an older article on the whole nuke reactor thing.

https://www.bbc.co.uk/news/articles/cx25v2d7zexo.amp

I get where you’re coming from — some AI-generated messages do feel generic or lack a human touch. But I think it depends a lot on how they’re used. AI can be a great tool for speeding things up or getting past writer’s block, but ideally people should personalise the output so it actually reflects their voice. Like any tool, it’s only as good as the effort someone puts into using it well.

I have seen the term techno feudalism bandied around - the owners of the tech will continue to be as rich as Croesus whilst the mass of society will be impoverished. We need democratic control of society and the implications. Reform will just throw us on the pyre of billionaires.

Yes I did it with chatGPT, just uploaded a photo of my garden and asked it to show me what I can do with the space.