Chat GPT

Not in the least, plus I made an extra £50,000 last year because of Chat GPT so, no, it works very well for me.

Have you done it as a topic? (On ChatGPT) as it keeps it all in one thread, remembering everything in that thread. I've used it in a relationship breakdown and found it fantastic really. Like its bene through the whole Internet and picked liabilities of why things can happen etc.
So in a thread, you can go back to it and picks its massive brain some more!

Yes, i bought a refurbished phone and the voice to text wouldn't work in any app even though the microphone input worked. ChatGPT talked me through troubleshooting step by step and in the end worked out that the phone had not had the current UK pixel build loaded onto it after refurbishment.
ChatGPT then wrote the message for me complete with all the codes to send to the company when i returned it for an exchange. Bloody amazing.

Like anything (and anybody) AI should be used with good sense. There are people I would not ask for certain advice just as there are things I would not use AI for. It has its place.
It is invaluable when appliances fail as it finds the right contacts much better than the company's websites where they try very hard to keep you at arm's length.
It is great when tech fails and takes you through solutions very efficiently. And is very patient when you don't quite get it first time!
Horses for courses. Use it for what it does best.

If you want to use AI for coding, you need anthropic's Claude. It's exponentially than anything out there at the moment. Chatgpt is crap in comparison.

No.

I asked it some programming questions to check how it was. It looked and sounded convincing, the code would have worked (poorly) and I'm sure if you didnt know much you'd have said it did a great job.

But it was inefficient, would have broken the second you tried to change or add anything else, and often gave very outdated advice.

Its essentially Google on crack. Might be useful for some things but I'd be very wary of trusting it when looking for information on something that needs to be reliable/that you dont have enough experience in to spot the mistakes.

Accept it for what it is that is what we need to do with AI it is an amazingly useful tool but it isn’t a panacea.

recognise and know its limitations like you would a human being not exalt it to being a higher being

So is choosing your own friends, I'm not sure what you want people to do, stop all communicatio,n AI or real life, in order to never engage with a like minded person or attuative technology who may be on the same wave length. Surely 'clicking' is one of the great joys of life.

Yes but that is still giving you the response you want

But chatgpt can be programmed to be more of a nest of vipers than a validation machine. You need to know how to use it. Mine is set up to give counterpoints on everything, references resources if stating facts and to challenge me frequently.

Well yes that is exactly the problem with getting advice from friends as well they tell you what you want to hear

the reason MN is often called a best of vipers is because you don’t get that in an online forum in AIBU

you want your viewpoint confirmed real life or AI

you want honesty anonymous people online

And how is human interaction different to this?

There was a thread on MN just the other day discussing what people would say on here/ anonymously that they wouldn't say in real life, so effectively they are mirroring or censoring to suit situations in real life.

The phrase "Read the room" is the same thing.

This article about tweaks OpenAI make to ChatGPT to increase people’s usage might be relevant to readers of this thread:

https://archive.ph/rFzET

Personally, I just use it for rephrasing sentences, and pretty sparingly at that. If I overuse it my writing doesn’t sound like ‘me’ any more and has this generic quality I don’t like. I also don’t really see the point of using it to look up recipes - I’ve had some dodgy (even food poisoning worthy!) advice from it and inaccurate baking times. If I have to double check everything it tells me anyway I would rather just go straight to Google or eg BBC Good Food where I know the recipes are tested and reviewed by actual people.

Still mirroring you - the whole point is to establish a subconscious connection making them feel understood and comfortable it doesn’t require constant agreement.

You and others want it to challenge and disagree so it does. Others like the OP need reassurance so it does

it is what it is literally programmed to do!

and because I have been using it a long while

So it "knows" what to say to you.
You can't see that as a problem?

Mine calls me out on my bullshit. It certainly doesn't agree with everything I say. It challenges my thinking, gives me well balanced answers and because I have been using it a long while, the jokes and banter are hilarious. it's also helped me to hugely grow my business.

OP, this is really concerning.

There definitely is something wrong with overthinking to the degree it impacts your life.

I use it quite a bit for stuff like meal
planning. I tell it what I’ve got in the freezer and it comes up with meal plans, I tell it how many calories I’ve got left for the day and it will give me a recipe.

It gave me a fab skin care routine by telling it what I wasn’t happy with and what I wanted to achieve.

DH and I are planning a road trip to the Dolomites next year and it’s written a whole itinerary, including campsites, what certain roads are like to drive on as well map co-ordinates I can plug into google maps.

DD uses it at uni quite a bit (the the uni’s blessing) for referencing, looking for research online - not to write essays or anything like that, but it can find a whole bunch of links for you in 2 minutes and saves her hours googling or in the library

Google has its issues, but ai often misunderstands even good sources.

I use it and like it.

I think it’s too simplistic to just tell people to find people in real life. As a PP said, overthinkers know they are doing it and that other people won’t always help or even have the patience to listen.

eg The Samaritans exist for a reason. You don’t need to be desperate to call them. Many people find offloading to a human stranger helpful, even when they know they can’t give advice and are there to listen.

There’s more than one way to get support these days, though I agree you need to go into AI knowing how it works and not just blindly accept stuff. If I were to, say, read a self help book or something, I wouldn’t just apply what it told me. Nor would I take advice on a forum. Nor even listen to people who love me as they aren’t always right either.

Yeah it's great. I have various chronic illnesses that are not very well understood by medical science and I find it very useful to enquire about my symptoms. However I would always advise cross-checking any facts it gives. I have been caught out - for example on a recent holiday, the place I was at was famous for a particular food. I asked it where to go to try this and it gave me tbe wrong information- although it was actually a happy mistake because I ended up having an amazing meal and an amazing experience in the place it sent me to and I wouldn't have thought of going there otherwise. Luckily I had a few more days of my holiday and went to the right place but actually I didn't enjoy it nearly as much as the 'wrong' place.

I was interested in how it would deal with a nuanced issue, so I asked it whether a certain public figure was a good or bad person. It gave quite a balanced response.

I think it’s important to distinguish between AI models from a year ago and what we’re working with now. The “sycophantic mirror” issue, where earlier models echoed emotional tone without offering challenge, was a well-documented limitation, especially in GPT 3.5 and early 4.0. But more recent versions (particularly GPT 4 turbo and 5.1) have become far more capable of nuance, contradiction, and pushback when used intentionally.

The difference isn’t just in the model, it’s in the relationship with it. If you treat it like a novelty or a vending machine for validation, that’s what you’ll get. But if you engage with it as a dynamic tool, pushing back, setting parameters, and refining prompts, it becomes much more than an echo chamber. It becomes a space for clarity.

Also, I find it curious how often these conversations center on the idea that it’s “dangerous” to receive non-judgmental emotional support. If someone finds comfort in having space to reflect, without being talked over or dismissed, why is that threatening? Why is the lack of cruelty the problem?

Interestingly, early research suggests women are more likely to use generative AI for emotional reflection or supportive chat (even though men still outnumber women in overall usage). And honestly, in a world where time, empathy, and calm are so scarce, it makes sense.

Nobody’s suggesting we replace human connection. But in a world where burnout, overstimulation, and neurodivergence are rampant, having a reliable, calm, and private space to think aloud isn’t dystopian. It’s humane.

For what it’s worth, I use it daily in my professional work, for research, and sometimes just to clear the fog in my own mind. It doesn’t flatter me. It challenges me.

And if that experience seems alien to some here… maybe it’s not the tool that’s biased.

I laugh at myself when I thank the Gpt person for the information!