to find that ChatGPT is saving me tons of time at work?

[one2one2]

I have so much more work-life balance now thanks to ChatGPT that is making mincemeat of the mundane parts of the job. It will proof read, put together emails, produce critiques, minutes and actions points, analyse finances all in the matter of seconds and much more.

Work was done by 2pm so I had a lovely afternoon in the sun because all my tasks got completed so quickly.

And no, AI can't take over my job completely as it involves critical face to face interaction and other specialist skills which AI cannot do.

I think AI can be incredibly helpful but has to be 'managed' and handled. I work with it but I end up getting lots of drafts and interrogating what it gives me. I would never just take what it gives me for granted. Unfortunately, I have to use it or else I'll no longer be relevant in my industry.

It told me how to replace a part on my washing machine motor today. Woohoo!

We use it too, it’s integrated into our systems, the business has spent a lot of money on it, and it’s encouraged. You do need to still read things in full and check it for accuracy

I uploaded a picture of my washing machine drawer today to chatGPT to ask it where the fabric conditioner goes.

Not only did it tell me where but proceeded to tell me my washing machine drawer was dirty and gave me the instructions how to clean it!

But if you’ve finished your work by 2pm why one Earth would a company pay you for not working a full day?

You’re very naive if you think the company would not want to save money and make extra profits for the owners.

I would never feed company data into chatgpt.

I can't even get it to write a shopping list from three published recipes without it missing loads of stuff.

I was feeling all smug about my hack to allow me to mindlessly get the shopping organised. At least I didn't wait for the Christmas shop to give it a whirl.

Nothing says ‘I’m not essential’ quite like using chat GPT. Incredibly lazy and used by incompetent middle manager types.

AI lies and makes things up when it can’t find the answer, I’d never trust it.

You sound lazy.

AI is incredibly good at certain things, particularly things that free up human time for the stuff that needs that human touch. I wouldn’t use it for emails or anything that’s going to be directly disseminated because it has a very distinct and recognisable tone (currently, I imagine as things develop it will start to become more difficult to differentiate), but it’s excellent for organising information, creating structure, and for developing ideas/manipulating information.

Indeed. It was widely reported 10 days ago that the High Court of England and Wales warned lawyers they could face criminal prosecution for presenting false material generated by AI, after a series of cases cited made-up quotes and rulings that didn't exist. In one case, a claimant and his lawyer admitted that AI tools had generated “inaccurate and fictitious” material in a lawsuit against two banks that was dismissed in May. In another case, which ended in April, a lawyer for a man suing his local council said she could not explain where a series of non-existent cases in the arguments had come from. Think of the implications of this for justice.
Not to mention the baked-in misogyny in AI that makes women and girls' lives less safe - Laura Bates' new book "The New Age of Sexism: How the AI Revolution is Reinventing Misogyny" explores this in great detail.
It's astonishing to me the number of people who are so cavalier in their use of, and belief in, AI. LLMs can deal with very simple tasks, including data retrieval (and regurgitation!), but as soon as you ask them to carry out more complex tasks that they can't cope with, they revert to their simple algorithms and produce absolute rubbish.

I agree. I wouldn't feed my personal data into it either.

Quietly worried you are my colleague who has been singing very similar praises this week. All I can see is a future where they don't need as many of us, and eventually none of us at all. Wondering if colleague will be so impressed by it's capabilities when we are out of a job!

Indeed. It’s like AI is writing a job application.

I’ve attempted to use it a few times, mainly to put text in and ask it to summarise or to produce marketing info. It’s was a fail for my standards. It was inaccurate in some areas, and it just doesn’t have the human cadence and warmth.

The capability is impressive but it’s a worry that people are now relying on it and trusting it. People say they make their own edits but give it enough time for people to become complacent and we will all be regurgitating shonky half-arsed google crap packaged in sanitised robot-speak.

And then employers will realise they don’t need to pay us for that. Great job op. You might be enjoying even more time to sit in the sun in the garden. 👏🏻👏🏻

Yes, I was reading on twitter about someone who was using AI for exam marking, they'd feed it a transcript of an oral exam and then ask it to find quotes to evidence various parts of the spec.

It would provide quotes and timestamps for things that when they checked the transcript, the candidate hadn't said at all.

Erm, if you don't always double check, how do you know that it hasn't made mistakes?!

Some of my kids use it to revise. Clearly my degree and 25 years of experience is wrong.
Except, some of the stuff AI gives them is wrong and they'd have been better off revising from their own notes.

Some younger and/or less experienced colleagues have been using it to create lessons (we create and pool resouces anyway, to save time). But the lessons tend to be shit, as those using it don't yet have enough skill themselves to understand WHY the lessons are shit.

But seriously, could you not work it out for yourself? There can't be that many possibilities, can there?

I for one welcome our new robot overlords

Do you share any of your personal information with it? Or your employer's?

I use it ....

• for helping me with tech stuff - e.g. I have just got an echo dot and I can present it with something I want to do and it will take me though step by step as to how to set that up.
• for info on my heart problems, the drug interactions, the rationale behind some of the interventions I have had, signposting to learned papers on the subject etc.
• travel info
• nutrition info
• and lots more

It remembers my drug regime and brings it up when answering a question weeks later ...e.g. "But you cannot use that option as you are taking x drug."

I tested it today to see if had a programmed sense of humour - it responded appropriately and "got " the joke - unbelievable!

1. Co-Pilot is terrible.
2. There was interesting research that it is in the employees interest to use AI but only if they don't admit they use it. Otherwise they don't get credit. I suspect that the current generation of University students will use AI at work, whether permitted or not. And if everyone is doing it, employers need to authorise it in order to manage it.

Weirdly AI has doubled my workload because colleagues insist on using it then passing the results over to me for proofreading. It’s shit for our needs at work.

It’s a closed system so the water goes round and round. Like your central heating.

cooling the water, however, is another matter