Wondering why people run every aspect of life through ChatGPT

[MiddleChildX]

Seriously. There seem to be swathes of adults who cannot function without running every single thing through an AI assistant. What should I feed my cat? Why did the postman look at me funny? How do I tell my neighbour I’m moving?
How have so many folk lost the cognitive ability and emotional resilience to deal with life? There’s something very wrong if you cannot make simple decisions or be confident in the most mundane social interactions without a computer telling you what to do. At this rate the machines will take over way before we ever thought. And that has not even touched on the horrors of the human and environmental cost of it all!!
I despair of humanity.

❤️

@APatternGrammar Interesting. Most of my job is writing and reading documents. We are having mandatory co pilot training at work currently. Sitting through yet another hour of a (not at all biased) Microsoft trainer extolling the amazing benefits of getting it to draft an email or summarise a document for you (but don't forget you need to draft the prompt carefully and need to check and recheck the results !!!!) I thought - but I can do all of these things quicker and better myself.

I know how to quickly extract key information from a complex document, I know how to write emails and papers for different audiences. Engaging with this frustrating, patronising machine that may well get a fair bit of it wrong will not make that part of my job any easier or quicker.

I tend to use it like a superannuated Google search at the moment - which horrified the poor Microsoft trainer, and I am sure is probably like using a rolls royce jet engine to power a lawnmower.

I also find that people who don’t do my job (that involves long technical documents) identify completely different aspects that could be done by AI to people who do do it. Reading the documentation and working out which parts are key is part of acquiring and understanding the information and it can’t really be cut out. Other uses of AI completely prevent you from getting into a flow state with your tasks, which is a huge part of job satisfaction for many.
But mostly it’s just not good yet. I asked it for a list of 50 one syllable words — more than half had 3 syllables. I asked it whether there were over the counter medications I could take in pill form that would have the same active ingredients as Lemsip — it told me Lemsip was outlawed last year because one of its ingredients was carcinogenic.

Of course you have been doing this a long time to a high degree. But have you actually started using it properly ? Shaping the prompts and honing in ? You could get through double if not more of the same work possibly if you wanted to . That's not the only point - but a big part of it. That doesn't mean you are not good at your job and I think some people feel that's what it signals. Of course of there are parts of your job you love doing and don't want help with that's totally fine too

I think people don’t want to listen to other people’s problems these days. I used it today because my ndn has done something really shitty. I would have gone round there all angry but as it is chat gpt has given me another avenue to hopefully resolve the issue without them being bastards to us forever. I could ask friends or work colleagues but it’s a boring thing.

@Saltycaramelkiss I think there will be benefits to me like doing consistency checks etc.. on final documents. I used it earlier this year to check 6000 template responses to a consultation from a campaign group to see if they were really all the same (if they weren't we woudl have to treat them as different responses). But as @APatternGrammar said, for writing and summarising documents the process is the point - it is how you learn and develop your thinking.

I have very elaborate prompts for the parts of my job that an LLM can do and they work fine (though they don’t actually save that much time as I’ve been using VBA macros I wrote to automate them for years already). The rest of my work can’t be done to any significant degree by an LLM and bringing its output up to a decent standard takes longer than just doing the work from scratch.

Also I don't think using AI means I am not good at my job - it is just for most of the use cases that can currently be applied to my job, I am better than AI or at least I would need to thoroughly QA and maybe even rewrite everything it produced - making it little more use to me than a very poor or inexperienced human employee.

That’s my take - no friends.

I'm a recent user to AI. I use it for my travel plans, but then research the results further.
It has also been invaluable with outfit planning! I have a wedding coming up and always struggle to look polished and put together. It has been unbelievably useful!
Also very good for wording any complaint emails and letters. Though I always alter it to sound like a human.
I agree that it is dumbing down society but do feel it also has it's place (within reason).

It’s great for therapy honestly today it’s been a godsend.

How do you know?

Personally I find co-pilot poor. People like it in the workplace because of data protection, but I don’t think it’s very effective so I don’t use it.

I’ve never used it.
Am I in the minority?

Chat gpt isn't a distillation of all the best doctors and medical textbooks - it collates ALL the info on the Internet, including things that are completely wrong.

It isn’t, and if someone was using only this in an critical situation, with no quality assurance, that would be crazy. However, Chat and other LLMs are not collating all the evidence on the internet, they will look for most repeated claims, which also means most consistent language. Medical language is very consistent so, broadly, it will use decent sources to report on medical stuff. That still doesn’t make it a doctor of course, but also no worse than simply doing a google search for the medicine’s leaflet, or side effects etc

Chatgpt is renowned for giving medical misinformation. It can be very dangerous to rely on it for any health issues. Always call 111 in circumstances like this!

It's actively giving people dangerous advice, which the medication leaflet doesn't do. Using AI for health information is a really bad idea.
https://www.theguardian.com/technology/2026/jan/02/google-ai-overviews-risk-harm-misleading-health-information

Socrates said the same about writing things down. He thought that reading stops people thinking for themselves and made them too lazy to remember things.

There was significant panic in the 18th century that the printing press and reading novels would dumb humanity down. They thought that novels served no real purpose.

These inventions have changed how humans think. This isn't the same as a slide into decay.

I mean, it’s not ‘actively’ doing anything, because it’s inert. But yes, it’s not a doctor so - again - people need to use their critical faculties. Just as using mine when reading this story leads me to ask ‘what proportion of the summaries found in this investigation were egregious, what proportion were accurate.’

If I had pancreatic cancer would I be blindly following Google AI advice? Obviously not.

I’ve very recently realised I can use it like this. It’s not because I can’t think directly myself or don’t have any friends, but it really helps to organise how I’m feeling. And I wouldn’t be boring someone rigid with my soul searching.

But a poster on this thread used it to direct her when her mother double dosed on medication. Some people are using it in ways that could be harmful. It feels like a doctor, it feels like a therapist, it feels like a friend or an advisor - but it might be feeding misinformation and pushing people towards dangerous outcomes. I only commented to say it's better to call 111 for health advice than ask chatgpt, and I stand by that.

You could quite easily say the same about MN.

MN puts disclaimers at the top of health related threads. There will also be a range of advice given rather than one answer purporting to be authoritative. I think MN can be really helpful in getting people's personal experience of health issues, but it is made pretty clear that the forum is no substitute for medical advice and the proper routes are signposted. Misinformation is frequently challenged by other posters. That's not the case with chatgpt.

In general I agree that it is better to call 111 - but have you tried that recently? Getting timely advice isn’t always easy, so I can understand why people turn to other sources - whether that’s google or AI. There’s lots of harmful stuff on the internet and I think there’s an argument for better regulation on medical info, but also people need to use their brains, check sources etc. I don’t think there’s anything inherently wrong with anyone using AI for whatever they need (with usual caveats around illegality/ethics etc) but yes there are risks if people are vulnerable or gullible (and those aren’t the same thing, nor mutually exclusive). I don’t think the solution is ‘don’t use it at all’ though.