ChatGPT BEtter than human?

[thatsthatsaidthemayor]

I’ve had a lot of difficulties. I’ve been through 4/5 counsellors over the years. Some have helped more than others. I love my current counsellor. But , (everything before but is bullshit but not here) I’ve started taking my problems to ChatGPT and finding the answers mind blowing! If this was a person I’d be delighted but feel uncomfortable that a computer is giving me better advice on being a human than a human. I get the wealth of k owl edge that it has, and that has come from humans. It’s just weird? No? So AIBU to take advice from a computer?

Fine to vent, but it’s a delusion to think you’re being “heard”.

You can have a “conversation” with Alexa, but do you think it’s “hearing” you in any meaningful way? Would it be satisfying to “vent” to Alexa? ChatGPT is fundamentally not really any different to a version of Alexa which can just fool you better. And for me, I’d feel a bit silly allowing myself to be duped just for comfort.

Yes, I am aware of that, being a plus subscriber. 4o has still been nerfed and gets rerouted when certain topics are discussed.

It is very different to Alexa. It is absoutley fine if you do not wish to use it. But not really appropriate to critise something that many people clearly find increadibly helpful, especially as you clearly have no experience using it.

It's useful to understand the context of a conversation before joining it. If you think ChatGPT is like Alexa then you clearly don't.

Eh? I’ve posted several times on this thread. And if you think ChatGPT is completely different to Alexa, you are deluding yourself. There is no person or thinking behind ChatGPT: if you are treating it like that then you are being fooled. It is not responding to you. It is a computer reflecting back text predictions generated by your text. If you think it is helpful, as long as you understand that, fine; but at the least it is a bit of linguistic trickery and at the worst it can be actively harmful.

The point of therapy is that there is a person there, with professional obligations, oversight and a legal duty of care. They have legal accountability, guidelines about negligence, statutory safeguarding responsibilities, and so on. ChatGPT has no professional training, no oversight, no colleagues, no legal personhood, no statutory obligations, and no come-back. If ChatGPT gives you psychologically unwise, unhealthy or harmful “advice” you can’t sue it; you have no legal redress; no-one is going to find it negligent. You’re shouting into the void: as long as you realise that, fine, your circus, your monkeys; but it isn’t therapy nor does it know or understand you, or anything it “says”.

Truly, no-one ever went broke marketing to the gullible! 😆

It is a bit rude to call people 'gullible' for using something that genuinely makes their lives easier and helps them. Just because it’s a different way of working doesn't mean it isn't incredibly valuable to those who’ve learned how to use it properly.
I would recommend you give it a try.
A lot of people find it helpful for checking the tone of their messages to make sure they're coming across as helpful and kind rather than dismissive. It’s actually a great way to ensure you’re communicating in a way that’s constructive for everyone. You might be surprised at how much of a difference it makes once you see it in action 😉

I have had extensive chats on chatGPT. It gives very different results when you ask it to be more honest. I have talked about the things it has raised with my IRL therapist. It has challenged me in some ways more than my therapist. What it cannot do is support me when I don’t like the answer. (I also did the age thing. BS. then I asked it to be brutal and it provided a whole skincare regime). It’s a computer working on algorithms. I’m not stupid.

What makes you think I need my tone policed? I intended to use the word gullible, because if people are genuinely treating ChatGPT like a therapist then they are being gullible. It’s that simple.

I’m not sugar-coating the facts - it’s silly to treat it like a person. It gets things wrong, and quite a lot of the time. And what makes you think I somehow don’t know how to use it? I don’t find it valuable at all for what I do; I can write an email or post much quicker than it takes me to prompt and correct it; and I find it frustrating how badly it paraphrases, writes and misrepresents things. I know a lot about therapeutic technique, so it’s useless to me. And seeing how often AI gets things inaccurate or plain wrong stops me from ever thinking it’s giving me meaningful advice.

I even find its recipes pretty crap: no idea why people think it’s even useful for that! There’s someone on another thread about the damn overhyped thing right now, asking it to suggest some meals. Why the devil would you even need AI to tell you what you wanted to eat? Why wouldn’t you just decide yourself what you wanted? Beats me. It would take me three times as long fannying around with ChatGPT as just to decide what meals to have myself. 🤷‍♀️

Dismissiveness often looks like common sense until the world moves past it. People said the exact same things about the internet in the 90s, it’s usually just a sign of being slow to adapt to a new logic. If the AI is misrepresenting things for you, it’s because your parameters aren't sufficient, it’s a high-level processor, not a glorified Alexa.
Beyond the tech, your tone is remarkably hostile. You seem so set in your ways that you've lost the empathy required to see how this helps others, or the perspective to see your own blind spots.

No, AI misrepresents things because it doesn’t “understand” them. If you have a high level of skill or knowledge in a field, it’s visible all the time. It just isn’t everything it’s hyped up as, and if you think it’s doing things wonderfully well, then you’re probably one of the people who can’t tell the difference (sorry, but true).

You also sound remarkably gullible. Do you think the tech companies are releasing these products out of the goodness of their hearts? Because their life’s project is to help women on parenting sites plan their mealtimes better? You’re clearly buying what they’re selling! But they aren’t making these products for you. They’re more interested in their stock options than selflessly making it quicker for humanity to write emails. And the early years of the internet also saw silly tech stocks booms and busts because people bought into the hype rather than the reality.

Just because it’s not right for you doesn’t mean it’s not right for everyone.

I’m in hospital right now and this morning at 3am I had a fever and they told me I was needing a CT scan. My friends and family were all asleep but I felt like I wanted to talk to someone so I told chat gpt what was happening.

I know it’s not a person and I’d not use it for medical advice or therapy personally but it did comfort me last night and let me know what to expect and reassured me as a friend would have done. I did feel better for that quick five minute ‘conversation’.

In the last few weeks, I’ve also used it to plan meals by telling it what food I had in that needed using up - the inspiration was much needed ans it definitely came up with ideas that I wouldn’t have done.

I’ve used it for imagining what rooms would look like decorated in a certain way and also to draft a complaint email which it did way better than I would have done.

I was slow to come around to chat gpt but now I accept that it can make life easier in some small ways so I’m willing to embrace it.

if you don’t want to then that’s fine.

That was me and I have a new set of food intolerances so I used it for some basic ideas (and you may decide what YOU want to eat and then just magically cook it, but I"m assuming you don't have fussy children, a mildly fussy husband, and limited time to shop and cook especially just before you fancy eating) and then asked MNers to come up with better ideas, which indeed they did.

But @BeFairOliveBear It is very different to Alexa. It is absoutley fine if you do not wish to use it. But not really appropriate to critise something that many people clearly find increadibly helpful, especially as you clearly have no experience using it.
It doesn't have feelings. You are talking like it will be upset if it is criticised. Because of the nature of ChatGPT it makes a very dangerous therapist. If we went ahead and said "it's a great therapist, use it as much as you like" that would be very UNhelpful to the people who are using it as such.

Surely this must be the same person as another thread.

To repeat it doesn't have insights it's a computer, it can't be brutally honest, it can't be honest or dishonest. It simply reports what it found on the internet in response to your question and using whatever algorithm the programmer wrote.

@drspouse But that’s what I mean exactly - surely it was better and quicker to post on MN and get humans to advise you rather than bother prompting a computer?

I have no doubt that as a secretarial/search tool it’s useful in some ways occasionally. However, I find it irritating and timewasting, not accurate enough to be useful, and I get better results from an internet search or just doing things myself. Try asking it about a complex subject you know a lot about, and see how well it performs. You’ll take more time checking it for accuracy and correcting the errors than just doing it yourself.

It’s a very effective time sink, though! But always ask: WHO is benefiting from this, and what is the actual product, here?

I really don’t mind if people criticise ChatGPT, I’m usually the first one to point out its flaws. We all know it isn't human and it definitely makes mistakes, you just have to use your common sense with it, the same way you do with anything else in life.
What I do mind is the criticism directed at the PEOPLE who use it. When critics call users 'gullible' or act like they’re idiots for finding the tool helpful, it is so closed minded. It shows a total lack of empathy and a misunderstanding of how the tool actually works for different people. Everyone’s needs are different, and just because someone finds a specific tool useful doesn't mean they're being fooled. It’s a bit disappointing to see people being so judgmental instead of just realising that we all have different ways of getting things done and find different things helpful. There is no need to be condescending towards people who are different to you and find different ways of doing things.
If someone hasn’t actually tried using ChatGPT to talk through a problem, or used it for things like journaling and working through life's hurdles, they aren't really in a position to judge. It’s hard to criticise others for finding it helpful when you haven’t experienced how supportive those techniques can actually be. Until you've seen how it can help you process thoughts or find a new perspective, it feels a bit unfair to look down on the people who do get a lot out of it.

A lot of human interaction is just pattern matching. Which is Chat GPT. But humans also come up with new stuff. And care.

What happens when Chat GPT is trained mostly on AI slop, not proper humans.

I agree with you in a lot of ways OP. I also like that ChatGPT will just address one problem that I'm bringing, rather than needing four sessions to explain the entire background of my life...

It's very fast, and very insightful about other people's behaviours and how to react to them. I've found it brilliant.

The only personal thing I’ve done with it is giving it an entire WhatsApp conversation between me and a relative who is very hot and cold with me.
it very accurately gave me insights that I hadn’t seen, patterns in communication that I hadn’t noticed and which now I see are completely right.

however that really freaked me out, I'm sticking to itinerary’s diet plans and technical understanding for now.

Lots of people on this thread have told you how they are benefiting from it but honestly you seem too closed minded to grasp it.

That's not how LLMs work. Some don't even have access to the internet.

You've just outed yourself as not knowing what the thing you are talking about, is.

@catinateacup it wasn't - I got a full two weeks menu in minutes with some adjustments. I've had tweaks later from MN but being humans they have other things to do and it's taken a few days.

They’re all trained on training data. So many people on this thread saying “oh you don’t understand it”, but they appear not understand it themselves!

Thinking that it’s being terribly insightful is just confirmation bias. Of course you think its insights are super, especially if you think they are insights in the first place. Psychological biases towards believing information that sounds plausible or credible (even if it’s not, or it isn’t even remotely verifiable), are well documented. Many people find mediums, horoscopes, agony aunt columns, and all sorts of things plausible and insightful. Doesn’t mean they are, though.

But if they think they’re getting therapeutic insight and they’re actually not, how are they benefiting?

Similarly to how they might benefit and gain insight from a conversation with a friend? Or a mumsnet post? Interaction doesn't have to be therapy to be helpful.

That poster stated 'It simply reports what it found on the internet'

That is incorrect.

I know how LLMs work. I research, work with them and use them daily. You don't need to school me.

That poster was incorrect.