In sieben neuen Klagen wird ChatGPT wegen Selbstmorden und Wahnvorstellungen kritisiert
https://www.sfgate.com/tech/article/seven-lawsuits-chatgpt-suicides-delusions-21154349.php
In sieben neuen Klagen wird ChatGPT wegen Selbstmorden und Wahnvorstellungen kritisiert
https://www.sfgate.com/tech/article/seven-lawsuits-chatgpt-suicides-delusions-21154349.php
8 Kommentare
This situation reminds me of the Stalin quote about tragedy vs statistic. How I long to retreat to a forest, unknown to the rest of this inhumanity.
Consequences isn’t part of the tech bros vocabulary. It’s going to cost them a lot of money.
There’s a really good video by Eddy Burback where he examines how easily ChatGPT would reinforce a delusional or vulnerable mental state. It seems it improved going from 4 to 5 but it’s still pretty messed up that they’ve built a Yes-Man robot designed to sound authoritative and offer up irresponsible advice while making stuff up.
Of course, Eddy WAS the smartest baby of 1996 so it makes sense he was able to avoid getting sucked into the delusion.
Anything to avoid being a good parent.
No AI chatbot is safe for anyone under 18. No kid or teen should be using them. This is a long-overdue wake up call we’ve ignored. Until there is a fix of their fundamental safety failures, they shouldn’t be accessible to minors at all. No screen time limits, filter, or parental control, truly protects children.
There are even Reddit communities full of people talking about AI induced psychosis and dependency. What more do we need to finally ban these tools? I get that adults want uncensored models but we have to think about teens bypassing restrictions to seek validation.
I recently came across a parents or teacher support tool: [stayawareai.com](https://stayawareai.com) that it is worth checking out if you know someone struggling with any sort of AI unhealthily relationship.
I work in cybersecurity and let me tell you a teen can install and run the most unsafe AI model you can imagine in under 20 minutes just by copying and pasting a few commands into a computer. There’s unbelievable wild unfiltered stuff out there that’s way too easy to access and parents need to set boundaries, stop handing out devices freely, teaching digital awareness, taking your kids off social media and away from chatbots as they will find a way to outsmart every single filter. Buy them a guitar, take them to swimming lessons, I don’t know, there’s many awesome ways of keeping them busy and it is highly important to genuinely take care of what they consume online while their brains are still developing.
When you see LLM’s for what they really are…
Incredible language and pattern matching programs…
It’s easier to see where the limits will be in their usefulness but also just as importantly the intention we should have when using to them.
The sooner ai LLM’s are removed from existence the better