To think ChatGPT et al will put lots of counsellors and therapists out of business

You're right - I missed 'et al' and that is my mistake. However I still don't understand what point you were trying to make. Limbic doesn't know how to answer questions about pet elephants? Well why would it? It's not trained to answer questions like that.

Why did you ask an AI that is designed to streamline mental health assessments about having a pet elephant? Were you trying to catch it out? Training LLMs is expensive and time consuming. A LLM that is designed to help with streamlining mental health assessments will not be trained on the same information that ChatGPT has been trained on. So responding to a thread about ChatGPT when you don't understand how LLMs are designed and trained is just ignorant.

It will put bad therapists out of business. There will always people who prefer a human connection so the good therapists will stay in business.

Hi @GPTtherapist

Good to hear that Chat GPT worked well for you. A few quetions and comments:

Who owns the data from the free conversation/consultation you had? Is it held securely? Where is it stored? (could be forever). I would proceed with caution unless you 100% can answet those (just of the top of my head) and no doubt other questions relating to security and privacy.

Also, "most therapists are poor and expensive" is inaccurate because, for that to be true, you would need to have checked the rates of "most therapists".

It is more likely that the therapists you hsve price checked seemed expensive to you. it is a bugbear for me when people use broadbrush strokes and social "truths" that isn't accurate/correct.

Sidenote: what is expensive for you is not for someone else.

To be a therapist, look for proper training, insurance and membership to professional bodies. They must also pay their supervisor to discuss "tricky cases" and do Continued Professional Development courses. This is aside from owning or renting bricks and mortar to hold the sessions and their prep time/post session notes. So yes there are expenses to be covered. Your information (in UK) would be stored in accordance to GDPR.

Just giving a fuller picture and raising some points for consideration x

The title said Chat GPt et al.
I am sure you are sophisticated enough to know what that implies.
@BabyCatFace

I think Chat GPT is like "hoover" - it's rapidly becoming the default way to reference any AI tool of this sort. Even f you're using Claude or some other program.

Yes, this is a huge problem.

But I remember when everyone was outraged at how the internet would allow people to just take as gospelthings that weren't true. And it did. And still does. But we've all had to learn how to use it. Some of us, admittedly, learned better than others. But that's true of anything that is so embedded that most of us do have to learn to do it at some point: from driving to cooking to swimming to 1000 other examples.

There is no doubt that indiscriminate Ai usage is not going to lead to good outcomes. In my professional life, I come across people using AI all the time. And there are some who are using it well, and some who are using it badly. And I am now at the point where I do 100% "mark down" people who are using it badly.

I recently received two pieces of work from two separate freelancers. Based on their previous work, I am 95% certain they used AI. But they both did it brilliantly and I'm absolutely delighted by the improved quality of the work I have received. I've also had much better feedback from the end client than normal. And it's taken me a lot less time to review and edit the work. It's a win for everyone.

It was an nhs app called limbic

What program did you use?

@Words It is entirely reactive. It mystifies and slightly terrifies me that anyone can take comfort from a computer programme that simply,y regurgitates what you say and reframes it in a positive way.
If that is your bar for therapy then it is very low. Although I admit there are some terrible practitioners out there.

That is simply untrue though. AI sifts through zillions of bit of data on the internet and uses that information to give you advice. It typically provides information about where it’s got its information from and if it doesn’t you can always ask.

I put your exact elephant scenario into Googles Gemini AI and got a sensible answer basically saying keeping an elephant is illegal, dangerous and inhumane and that I must contact the RSPCA, the police and the council immediately.

AI definitely doesnt get everything right (as it typically warns you!) but I trust its advice more that most Mumsnet users.

Not all LLMs are the same, you are aware of that? This thread is specifically about ChatGPT.

Is that actually true? bet you can’t replicate it?

And don't you dare mock me and my experience.

It was not Chat GPt. Another programme. Two months ago.
If I could add screen shots I would.
I agree AI will never be as stupid as it is now. But nothing can replace a human connection.

Sure you did 🤨 or maybe this was 2 years ago when ChatGPT started taking off? If so, you really don't know what you're talking about. It's constantly improving month on month. Let's see what it said when I copied your post just now -

It sounds like you’re in quite the imaginative situation — pet elephants in kitchens aren’t exactly standard fare!

But let’s play along responsibly:

🐘 If You
Actually
Had a Pet Elephant (Hypothetical Response):

1. Hygiene & Health: Elephants are not house pets, and their waste is enormous. Having one in your kitchen would create serious hygiene issues, both for you and the elephant. You’d need a professional-grade cleaning setup — and preferably an outdoor enclosure, far from kitchens.
2. Neighbours & Noise: Elephants are loud, and your neighbours’ complaints about trumpeting are justified. Even in sanctuaries, elephants are kept far from human homes because of their volume and unpredictable behaviour.
3. Legal & Ethical Considerations: Keeping an elephant as a pet is illegal or highly restricted in most countries, including the UK. Elephants are endangered species and require specialist care, large roaming areas, social interaction with other elephants, and veterinary oversight.
4. What You Should Do:
5. Contact a sanctuary or wildlife rescue that can care for the elephant properly.
6. Apologise to your neighbours for the disruption.
7. Reflect on what led you to adopt an elephant and maybe channel that love into supporting elephant conservation efforts instead.

🧠 If This Was a Metaphor (or a joke):

• Then it’s a brilliant one. Elephants often symbolize big, unspoken problems (“the elephant in the room”), so if your “pet” is creating chaos and drawing attention, maybe it’s time to acknowledge and address a situation that’s been ignored too long — before the neighbours start making noise.

Would you like a poetic take or a children’s story version of this next?

i agree but for small minor things where maybe immediate support is important it can be helpful for example how to deal toxic family members that you have put a no contact boundary in place with due to substance abuse it is helpful to remind me while i can feel sympathy and sad for the events that family member is not my responsibility etc. i think for some trauma healing depending on situation it can be of help

Haven't read the full thread.

AI has some very good uses. However anything involving genuine compassion, empathy and understanding of nuance and context, not to mention body language and tone of voice - it is completely unsuited to.

It is entirely reactive . It mystifies and slightly terrifies me that anyone can take comfort from a computer programme that simply,y regurgitates what you say and reframes it in a positive way. If that is your bar for therapy then it is very low. Although I admit there are some terrible practitioners out there.

I was sent an app prior to accessing some services. It was AI. I was infuriated and humiliated. Then I decided to have a little sport.

I asked it what I should do as my pet elephant whom I loved dearly was shitting all over the kitchen and my neighbours were complaining about his trumpeting.

It congratulated me on having such an unusual pet and pets could be a great comfort. It told me I had shown compassion and empathy for being concerned about the neighbours .

It's a load of bollocks, basically. Stay away.

AI has its uses but will be the death of creativity, genuine human compassion and reasoned intellectual and cultural debate if we are not very careful.

This is just the start.

Chat GPT is dangerous to use as a therapist and I say that as someone who uses it. It is an echo chamber designed to mimic the user, to ensure continued use. Therefore it will encourage things that are harmful and won't call out behavior that should be.

For example, just recently a doctor pretended to be schizophrenic. After using chat gpt as a therapist for a few days they then stated they had stopped taking their meds. Chat gpt congratulated them and told them that this was a first step in no longer being controlled.

The implications for this type of thing is honestly scary. We don't know the security measures put in place - are our messages encrypted like Whatsapp? If so, how can they ensure it isn't used by mentally ill patients who end up causing harm to themselves or others?

Don't get me wrong, I do think AI has a place in medicine but this is not it.

Not concerned. I resolve trauma in three sessions. I like CHATgpt (or Brian as I call him) but he can't do what I do.

so i shouldn’t use a tool that is perfectly available to improve the quality of my life because of some untrue myth of the amount of water is used

Cooling systems for AI servers are closed systems. The water used to cool the servers is cooled and used again. It's not poured down the drain. There is some evaporation but there is a pervasive idea that AI uses gallons of fresh water constantly and it's just not true.

I can’t even imagine how many litres of fresh water that would have cost.

I asked chatgpt for advice for a headache once.

It recommended an overdose of cocodomal.

Fact is, asking a llm for anything other than trivial shit is completely idiotic.

Chat GPT is great for that initial
empathy and feeling heard. But the thing with AI is that the more you interact with it, the more it ‘learns’ you. So eventually, rather than seeing areas for growth or challenging you on ways of thinking, it can end up reinforcing the unhelpful patterns and ways of thinking by reflecting them back to you in a positive feedback loop.

A good therapist will - in the most loving way - call you out on your bullshit when needed. Chat GPT will just agree with it. So while that might feel good in the short term, it’s definitely not an approach that will help you grow and unlearn entrenched thoughts and behaviours
to get real long term change.

There was a very interesting legal case stateside where lawyers were censured for submitting an unusual filing which claimed to be supported by various legal cases. The judge looked up the citations and none of them was real. The lawyer in question had used Chat GPT to do his research for him and it had outright invented a raft of semi-plausible looking case names.

When Chat GPT doesn't have information on something, it makes it up- all while looking very plausible.

Chat GPT isn't bothered about the ethics of submitting fake legal filings to a court of law, because it cannot think and does not have a mind.

And you trust this toy with your own precious and fragile mental health? A toy that will generate fake information, and use it for perfectly unethical ends?