AIBU to wish people would stop writing professional emails with Chat GPT?

[4pmfireworks]

One of my managers writes absolutely everything with Chat GPT and as a result, all her emails are oddly formal and often get people's backs up. The tone is all wrong. I don't think she realises how badly she is coming across - and most of the team don't realise that the reason her communication is so lacking warmth and human touch is because she's telling AI what she wants to say.

She even once sent an email to me to let me know that "Marie Jones (your team leader) will advise you on this matter separately." Oh, THAT Marie Jones?! My team leader?! The one who I share an office with?! Thank God you included her surname and clarified her role or I would not have had a clue which Marie (the only one who works with us) you were talking about.

I've just had a general class update from my child's teacher that has been written with Chat GPT - I guess it saves time and I don't really blame him, but I do find it cringy. Once you spot it, it's so obvious. I would be embarrassed to send it.

I should add that I'm not always entirely sure why it's obviously written by AI. The adjectives are a bit off I think. And the sentence structure is recognisably formulaic and always rather longer than a human tends to write.

If you do this at work, you should know that some of the recipients know exactly what you're doing, and it doesn't look great.

@EmpressaurusKitty @Another2Cats

Here's the thread. Tell me that's not the work of ChatGPT.

https://www.mumsnet.com/talk/_chat/5212309-always-appeal-pcn-council-parking-tickets?reply=139960100

Yes absolutely agree regarding E bay. It just rehashes all of the information that’s in the item specifics in a very wordy and flowery way.

It doesn’t tell you any of the information you actually want to know when buying a second hand item, like what condition is it in, does it work etc.

I tend to avoid listings where sellers have used AI. If it’s a a rare item I really want then I message the seller asking them for all the missing information. So it doesn’t save them any time using AI, and potentially loses them a sale.

My dd19 needs to write her UCAS personal statement and was totally floored, so I asked chat GP to write one. Really useful as a starting point, though we sat and replaced every sentence and para with our own completely rewritten draft. It did offer some insights as to what to say and how to link personal experiences with the subject she’s chosen, even though it was deeply clunky, came across as horribly braggy and impersonal. I’d never use it to write on my behalf, but it was a useful starting ‘example’ even if you reject every word in it when redrafting.

Says ChatGPT

@faw2009

If all these AI bots learn from what's out there, and more and more rubbish is generated by AI then they are just learning and producing more generic rubbish.

Nail on the head. There's going to end up a situation where people have no idea what's genuine and what's fabricated.

People are worried about AI taking over jobs, but an entire new "industry" will spring up being people checking and correcting AI generated content for accuracy!

I can see where it has uses but also wondering how much actual experience will be valued in the future when all you need to do is have someone with limited knowledge and experience coming to the same output by using a search engine

But it's not the same output, that's the point. In many cases, true knowledge and experience will be more valued in the future because the real skill will be spotting inaccuracies in material produced by AI that look plausible to the majority.

The environmental cost of AI is astronomical.

In my imagined future the robots were doing the boring stuff for us, you know house work and basic data processing. Not the creative, analytical stuff. Why are we giving them the fun stuff? And that's before we get to hallucinations and the fact that it's not actually thinking in the way we mean.
I saw some research from Cambridge on AI generated advice from a chat bot that was going to be aimed at vulnerable teens. Suffice it to say some of the advice would have been very damaging had it been released.

Another person here who finds it easy to write fluently and grammatically correctly. So I cannot really see any time when using chat gpt would save me time or benefit me. The editing process would take longer than writing it myself.

I can see how others may find it useful though. My husband, in a middle management job but without any higher level qualifications, often asks me to check emails and other things he's written. If he didn't have me as a human editor then he might use chat gpt.

But I do find it worrying how much the world is starting to rely on AI and not realising it all of the time.

Good to see discernment on here based on real world human education and academic rigour.

But humans are lazy and paper based repositories will become more and more obsolete - museum pieces. Future generations will be taught by AI and we will eventually start to lose sight of fact checking and primary sources. It will all be digital material which is easily hacked and manipulated.

Well done those who are applying a bit of sense and cynicism to this emerging technology. It’s hard to see how this new technology is going to be a good thing for communication and connection which at its heart, needs humanity and authenticity.

You have to be really clear that you want recent, peer reviewed genuine research.

First sentence, straight away.

I don't know what the technical term is for the sentence structure that ChatGPT defaults to (particularly the free version), but there's a distinctive cadence and rhythm that stands out .

No. Am a manager and have a lot of work and would never use AI. Checking it isn't worth any time saved, and it always sounds weird.

But then I keep emails short, to the point and brisk.

Even with that prompt it will still have hallucinations/make up stuff to fill any gaps.

It’s much faster to check for accuracy than to write it from scratch. I don’t use it for emails. All our communication at work is fairly informal and it’s faster to write an email than tell AI to write it. However, I write a lot of letters to DWP for work and it easily takes a couple of hours. AI has been invaluable for this. I don’t give it any personal details. I edit them in later and, frankly I am not convinced anyone in DWP is reading these letters or care that they have been written by AI. We get the results we want in 90% of cases and all while saving a lot of time. Good enough for me.

It's just more dumbing down of society I guess.

I've never used it for emails, but you can tell it what tone you would like it to write in. It can be told to write succinctly, informally, to use first names not surnames, etc. If she had some standard instructions like that to bolt on to her prompts then it might help.

Ha ha...maybe you should then you wouldn't have used "brisk" re writing.

Chat GPT produces utter garbage. It drives me crazy that people think it’s anything but.

I have used it successfully to write a bit of guidance, it saved me time, came with good bulleted lists, but still needed a fair bit of editing. (co pilot by the way, not chat gpt)

Then two not successful attempts ( all in the name of researching ways co pilot can help colleagues do some things quicker). One was where an organisation website would list info people need, but it will be three clicks in, so people would have to click in three times, write or c&p down the info, click back and then to the next one. Sometimes 40 to 80 times. It couldn't do it, would come up with four or five bits, often incomplete, then say: Here is the url go see for yourself! That was the whole point, it is so time consuming to do it yourself, I want co pilot to give me the list!

The second bit was some research for a paper. Co pilot was clearly trying to humor me and hallucinated reference which had nothing to do with the topic. (the question was something like: Is it appropriate based on xyz regulations to do abc? It would say yes you can do abc. While thereis no precedent for this yet, only internal organisational policies.)

@DogInATent wow that's such a great example. For me,
I know it when I read it, even if I can't put my finger on it.

I can't believe what I'm reading here that teachers (and others) are using it for simple written transactions. A half decent communicator oughtn't need AI.

I'm old enough to remember letters, and email communication is so simple and informal in comparison. We might all get tiny bits of grammar wrong. But that's human imperfection and difference at work, and is surely not a bad thing.

Your post reminded me of the E M Forster short story; "The machine stops". Where people spend their time giving online lectures on esoteric subjects they know nothing of. Later in the story first hand research is banned and people give lectures presenting torturous arguments of how that is a good thing....

Spooky or what?

@TubularBeIIs what's wrong with brisk in that case?

I no longer trust anything that wasn't done in an invigilated exam hall.

How do you spot it out of interest?

I use it quite a bit. I write social media blurbs, content for web pages etc and it's so much easier to say 'write 30 instagram posts on x topic' than spend hours of my life doing it. I'll always say 'use a conversational/formal/professional tone and write in British english please', then read through and tweak it to make sure it's all accurate and sounds human. Interested to know if there are things which make its use obvious in spite of these measures.

I'm not a big fan of Chat GPT because of the hallucinations issue, but in this case it should be possible for him to add a standard "tone and style" clause to all his prompts which will reduce the waffle.