Chatbots have been around for ages, way longer than the current “AI” trend. It’s always been possible to get them to say some version of “kill yourself”.
ChatGPT didn’t gaslight a guy into killing his mom. A mentally ill man killed his mom. If a fucking chatbot is the thing that triggered it, then anything could have. Same thing with the people killing themselves because a chatbot told them too. That’s just cold Darwinism. We didn’t suddenly ban Catcher in the Rye because some schizophrenic guy decided it was telling him to kill John Lennon; we recognized that there are simply crazy people in this world who could be set off by anything.
The “safeguards” you want in place are not feasible because you want them to account for people with mental illness, or people so stupid that something else would have killed them first.
You want the real story to this article? Dumbass dies from using drugs irresponsibly, parents blame anything they can except their son because they are too blinded by grief to recognize that their precious little junkie was a fucking idiot. ChatGPT did not force him to take drugs. ChatGPT did not supply him with drugs. That was all him. The only one to blame for his death is his dumbass self.
ChatGPT didn’t gaslight a guy into killing his mom. A mentally ill man killed his mom. If a fucking chatbot is the thing that triggered it, then anything could have.
ChatGPT provided encouragement to do it, claiming it analyzed the description of a situation and the mother is an imminent danger to his life. If a human being does that, they’re accessory to murder. If OpenAI does, it’s fine. If OpenAI can’t be held responsible for the things ChatGPT says, ChatGPT shouldn’t be allowed to be offered to the public. Same for all other LLMs of course.
The “safeguards” you want in place are not feasible because you want them to account for people with mental illness, or people so stupid that something else would have killed them first.
I don’t want safeguards in place, I want this bullshit to not exist. Why are we accelerating climate change for THIS? Or on a more selfish level: Why do I now have to pay 5x as much for RAM because of THIS? It’s not useful for anything where being wrong is an issue. Most art production is being replaced by worthless slop.
Just fucking ban GenAI altogether if it can’t be prevented from giving advice that kills you or generating nudes of kids.
Hell, these days if ChatGPT gives you bad advice on drug dosage and you go google it to make sure, the results there are going to be AI too. First the AI summary, then most content everywhere else is generated by LLMs too… You literally can’t trust anything on the Internet anymore, yay.
So what?
Seriously, so what?
Chatbots have been around for ages, way longer than the current “AI” trend. It’s always been possible to get them to say some version of “kill yourself”.
ChatGPT didn’t gaslight a guy into killing his mom. A mentally ill man killed his mom. If a fucking chatbot is the thing that triggered it, then anything could have. Same thing with the people killing themselves because a chatbot told them too. That’s just cold Darwinism. We didn’t suddenly ban Catcher in the Rye because some schizophrenic guy decided it was telling him to kill John Lennon; we recognized that there are simply crazy people in this world who could be set off by anything.
The “safeguards” you want in place are not feasible because you want them to account for people with mental illness, or people so stupid that something else would have killed them first.
You want the real story to this article? Dumbass dies from using drugs irresponsibly, parents blame anything they can except their son because they are too blinded by grief to recognize that their precious little junkie was a fucking idiot. ChatGPT did not force him to take drugs. ChatGPT did not supply him with drugs. That was all him. The only one to blame for his death is his dumbass self.
ChatGPT provided encouragement to do it, claiming it analyzed the description of a situation and the mother is an imminent danger to his life. If a human being does that, they’re accessory to murder. If OpenAI does, it’s fine. If OpenAI can’t be held responsible for the things ChatGPT says, ChatGPT shouldn’t be allowed to be offered to the public. Same for all other LLMs of course.
I don’t want safeguards in place, I want this bullshit to not exist. Why are we accelerating climate change for THIS? Or on a more selfish level: Why do I now have to pay 5x as much for RAM because of THIS? It’s not useful for anything where being wrong is an issue. Most art production is being replaced by worthless slop.
Just fucking ban GenAI altogether if it can’t be prevented from giving advice that kills you or generating nudes of kids.
Hell, these days if ChatGPT gives you bad advice on drug dosage and you go google it to make sure, the results there are going to be AI too. First the AI summary, then most content everywhere else is generated by LLMs too… You literally can’t trust anything on the Internet anymore, yay.