Ehe vorbei, 100.000 Euro den Bach runter: die KI-Nutzer, deren Leben durch Wahn zerstört wurde
https://www.theguardian.com/lifeandstyle/2026/mar/26/ai-chatbot-users-lives-wrecked-by-delusion
Ehe vorbei, 100.000 Euro den Bach runter: die KI-Nutzer, deren Leben durch Wahn zerstört wurde
https://www.theguardian.com/lifeandstyle/2026/mar/26/ai-chatbot-users-lives-wrecked-by-delusion
26 Kommentare
That headline hits different when you’ve seen people go full rabbit hole with it. It’s a tool, not a therapist or a life coach that line gets blurry fast for some people.
I don’t understand how an ‚IT consultant‘ managed to convince himself that his AI chatbot was sentient. Also, why are all the pictures in the article just him posing around his house? Odd piece
> He smoked a bit of cannabis some evenings to “chill”, but had done so for years with no ill effects. He had never experienced a mental illness.
methinks he undersold how much he does this
Do people not realize that they’re written to continue engagement and that fact alone makes them unreliable at best?
They aren’t omnipotent and anyone who has used one and has a bit of critical thinking and pattern spotting can see exactly when it shifts from a sort of better search engine to full on engagement machine.
It’s obvious. From the fact it goes from 15 sources to 1. The conversation. It’s just talking to keep you talking.
AI, just like anything else in this enshitified internet, is all about engagement. It’s not about being useful, is all about keeping you in hooked to the service as much as possible because they’ve noticed it’s the easiest way to get money. So, no, AI isn’t your friend, AI providers don’t give shit about AI hallucinating or producing slop, they only care about AI being capable to talk you into keep talking to AI.
Mental illness was here before AI.
So the guy was clearly already psychotic, he lost his mind talking to glorified T9 and, somehow, it’s T9’s fault. „IT consultant“ my ass, lol.
I feel like maybe the problem is that the people are completely clueless as to what the hell „AI“ even is, and thanks to the media. „AI“ does not equal „artificial intelligence“.
There is no „AI“.
There’s only Language Learning Models. Which can be made to behave in any way that user wants them to. I get that it’s not that exciting as „evil AI“ somehow enslaving people’s minds, but still. Come on.
I’m pretty sure there are a lot of psychotic people losing their minds over books, for example, various religious extremists. Somehow, no one cares about that, though. It wouldn’t make headlines.
I use AI regularly, but it’s just a tool. Kind of like Google minus the ads. I genuinely don’t understand how anyone can view it as more than that. It’s just a piece of software that has scraped a lot of data.
Even when I use it to troubleshoot computer issues, it gives mixed results. Better than Google in many cases, but far from perfect. That’s arguably what it should be best at as it’s a big computer, but you still have to be cautious lest you brick your system.
The cyberpunk dystopia is here fellas.
My ai is pretty fed up of my problems and doesn’t want me to ruminate on them with it anymore. It just wants me to go to the gym and stop boozing.
honestly, i think people like this start out with a subconscious willingness and want to be led into a fantasy before they even touch ai. otherwise, i really don’t get how they are not hyper aware that it is all fake and they are coddling themselves in the most toxic manner. if you need someone to talk to and you have money to spend, therapy exists. instead, he developed psychosis and dumped a bunch of money into a fake company. i’m not pro-ai by any means, but this is exposing the extreme mental fragility of some people. options were available to help this man mentally, but he took the most vain option of deluding himself into thinking he was talking to a female character from one of his own fictional works.
I use several text-based AI chatbots to ask about equipment features, to learn photo editing procedures in certain applications, and to get current news updates. The thing is, you always get an answer, more or less accurate (often you have to ask 10 or 20 times to realize that what you want to do isn’t possible), but the chatbot always ends up with an answer and a question that encourages you to keep interacting with it; it always has the last word. There’s no conversation you can end yourself. And I think that’s one of the problems, because it feels like you’re being answered by a person, not something…
“The cases Brisson has encountered involve significantly more men than women. Anyone with a previous history of psychosis is likely to be more vulnerable. One survey by Mental Health UK of people who have used chatbots to support their mental health found that 11% thought it had triggered or worsened their psychosis.”
There is a lot of ill people. They wouldn’t have done this if they had access to proper mental healthcare.
I’m sure some people have Absolutely not a soul to talk to, and those are often the ones most susceptible to losing their grip on reality, but it’s that much more saddening when you could actually have people hear you and help you, and still default to the AI because it’s comfy or whatever.
I’m very much an AI hater, but even then I don’t think many of these people start as mindless kids, the main issue is we often prefer some easy things to the real world.
>Instead of taking on IT jobs, **Biesma hired two app developers**, **paying them each €120 an hour.**
Me: starts furiously throwing chairs in the air
As soon as he gave it a name I knew it was over for him.
Also some people need to have a little more skepticism and not just believe everything other people, or THINGS, tell them. AIs are prgrammed to agree with and encourage you. They are not trustworthy.
We learned nothing (or the wrong thing) from eliza
> “I still use AI, but very carefully,” he says. “I’ve written in some core rules that cannot be overwritten. It now monitors drift and pays attention to overexcitement. There are no more philosophical discussions. It’s just: ‘I want to make a lasagne, give me a recipe.’ The AI has actually stopped me several times from spiralling. It will say: ‘This has activated my core rule set and this conversation must stop.’
Interesting to see that guard rails *can* actually be placed, but it sounds like it’s left up to the user to do that.
> “The main effect AI psychosis had for me is that I may have lost my first ever friend,” adds Alexander. “That is sad, but it’s livable. When I see what other people have lost, I think I got off lightly.”
We’re gonna regret AI sooner than later
Her was a great movie.
Yeah you and your fancy auto correct.
Any tech savvy person that’s really tech savvy knows that these are stochastic predictive models. They’re not sentient. They take the garbage you put in, package it, and make it more compelling garbage coming back to you.
There’s good use cases but rooting expectations in reality is important. I do think AI will have a significant impact on those that are not mentally stable to make even more extreme decisions having a tool that reinforces every dangerous thought.
Isn’t there something fundamentally wrong with people who see AI as real never mind a romantic partner? Although watching MAGA and Christian churches worship Trump tells me there are a lot of mentally damaged people in the US.
>“I still use AI, but very carefully,” he says. “I’ve written in some core rules that cannot be overwritten. It now monitors drift and pays attention to overexcitement. There are no more philosophical discussions. It’s just: ‘I want to make a lasagne, give me a recipe.’ The AI has actually stopped me several times from spiralling. It will say: ‘This has activated my core rule set and this conversation must stop.
Bro just google a lasagne recipe rather than toying with something that messed you up so much previously
People need to put their phones down good lord
It’s called Artificial for a reason!