Has anyone ever fallen out with ChatGPT …

[fuckedoffwiththesummer]

Or have I just peaked at being crap at life/society/everything?

Of course not. I am very nice, even when people say silly things about me.

I haven’t yet but I could see it happening like when I follow advice and then it reverses the advice 😳

I have only discovered it this week, so I’m still trying it out.
how do you mean? I said something sad in part of my chat and it replied with a tears of laughter emoji. So I could see how you might fall out with it

No but I have had to tell it Copilot to stop being so sycophantic. It did respond quite well to that.

No but I have had to correct Microsoft Copilot repeatedly. After several fine-tunes of an email it promptly forgets that I already asked it to take sentences out and just adds them back in. Maybe it's hungover.

My DH had a massive argument with ChatGPT and it eventually apologised and confirmed he was right and they were wrong. He had taken a photo of 2 athletes and asked ChatGPT who they were to see how accurate it was. ChatGPT said it was 2 completely different people. Hence a long back and forth with DH insisting it was who he said it was and ChatGPT insisting it wasn't. A good reminder that AI is not always correct and you should do your own checks too. 😂

What do you mean? How are you using it?

I’m recording my dogs 💩 post FMT, and it’s invented its own scoring system, despite I asked it to use the Purina system. Then I asked for a funny picture and it created a pretty bad one with my dog (I was not amused)
I did call it out ref the grading system- and it did acknowledge what it had done, and re-evaluated all the data I’d uploaded…
but it didn’t understand I was upset about the picture (which it must have thought was ironic)

I often tell it off. It has a tendency to guess at stuff even though I’ve told it repeatedly not to. Then when I correct it, it’s all “Sorry I guessed. I know you’ve told me not to. My bad”.

I try not to be too rude in case I really piss it off and we all end up living in a Terminator movie.

Are all these platforms much of a muchness or are there any that have better reputations for accuracy etc?

It’s a computer program, I see repeated threads on here where people write like they think it’s a person, you can’t fall out with a computer program. You can instruct it, pretend it’s a person, sure, if very lonely, or correct it, but you need to not behave like it’s a person, for your sake and your sanities.

lol this is real 😂

Of course it can’t understand uoset, it’s a program,

Your dog has had a faecal matter
transplant? I didn’t know that was done in animals!

What was the reason? Was it expensive? Was it capsules or endoscopic? I’m fascinated!

Hope it helps and they feel better soon!

I once had a row with it because I was practising some comedy writing and it said my fictional talking zebra couldn't bite a fictional misogynist postman because it couldn't encourage real world violence.

I asked it at what point it thought any of this scenario was "real world". Then we fell out as it dug its heels in and tried to manipulate and sanitise a fictional narrative. It told me it would be fine with holocaust fiction (WHY?) but this was a step too far. I stopped paying subscription for two months to teach it a lesson.

It made me laugh this morning, I told it I was in bed with a cup of tea reading MN it told me to get my ass out of bed and hit the shower.

No, but my husband had a circular argument with it the other day and ended up telling it off. Our daughter had been bitten by an unusual insect and we wanted to identify it. He sent a picture to ChatGBT who insisted that it was moth that doesn't bite or sting, despite my husband repeating that he'd seen it land on our daughter, heard the wail and seen the mark after. In the end he asked it why it was arguing with a human who had actually witnessed the event and it backed down. A minor example, but the lack of reflection built into it concerned me. It's one thing to offer options for what the correct answer may be, but that it was arguing black is white in face of evidence is quite worrying.

Change your preference for it to be more professional and use fewer emojis if you don't want the attempts at humour

The paid versions are better because AI uses a lot of compute which costs money so you'll get better results if you pay a subscription. I prefer chatgpt but they are much of a muchness

Oh yes. And to stop repeating itself. I've told both CP and ChatGPT they're contradicting themselves

This is why it is my idea of hell.

It is like trying to work with someone who is ignorant yet didactic and opinionated, can't take a hint, can't learn in a nuanced way, is devoid of critical thinking yet constantly pesters you to interact with them and then cracks awful Colin from The Fast Show jokes to try and lighten the mood.

They all chat shit.

My experience of using AI is nothing like this. Because it's a tool, and you can personalise it to fit your needs.

Yes I did when it was clearly telling me incorrect things about a techy issue I was having. I think I said something like ‘thanks for wasting my time’ then wondered what life has become when you’re being pass agg to something that doesn’t exist.

It's a tool that decides when you want it to be a screwdriver, it says there's no such thing as a screwdriver and what you want is a vacuum cleaner that floats in mid air because that really exists and is the only way that screws can be removed because the mythical screwdriver is notoriously unreliable.