AI must have some validity when it comes to suicide advice

[Siamese Believe]

A 16 year old boy was able to use it to get partial hanging advice. His setup was a bar in his closet. When he used the advice, he died.

I've used both ChatGPT and Grok for advice more than I can count. Not for everything of course, I do plenty of research on my own. I use AI to fill in the more specific gaps that I can't find online, and can't get answers from people about.

I asked Grok about my partial hanging set up, I took a picture of me practicing and sent it to it, my face covered of course. It told me to stop immediately and that if I would've stayed in that for a little longer, it would've resulted in fast unconsciousness then death to follow. It felt good to hear that, but I still struggle to believe it.

I just don't know. The picture was me with some pajama pants tied around a bunk bed bar, then tied in the front and back around my neck. I would then lean forward and slightly bend my knees and start to feel like I was falling asleep.

I'm worried though, I don't know if that position will work with the way my anatomy is. And I don't want to just wake back up choking again like before. If I am to go with that method. I'm definitely gonna barricade the room still, just like with the heart plan.

For some reason I can't seem to compress the carotid arteries good unless my head is pointed more upwards. Will that be a hinderance in the dying process?

 

[U. A.]

Don't ask ChatGPT about suicide. Don't ask about what ChatGPT told you about suicide.

This is happening =1&c[content]=thread&c[nodes][0]=2&o=relevance']more and more and unfortunately I don't expect it to slow down anytime soon. So let me say what's in the title again but differently: ChatGPT doesn't know what the f- it's talking about. Artificial intelligence doesn't "know"...

sanctioned-suicide.net

 

[Forever Sleep]

I think there are more safeguards now- seeing as how some people have successfully CTB after communicating with AI.

That said- I imagine it's sort of similar to calling a medical line. They seem to base their advice on the worst case scenario. So- say chest cramps may not necessarily mean a heart attack but- they likely won't want to take that risk.

Simililarly- someone who wraps a noose round their neck could hang. Someone who puts a bag over their head could suffocate. I imagine- even if there was a 1-5% death rate- AI would likely warn you to stop now. So, I'm not so sure it's an accurate guage of assessing the likelihood of a method resulting in death.

It can go in the reverse direction also I've found. In simple Google searches about SN, I've sometimes noticed the death rates are lower than I would expect. But then- the likelihood I imagine would be that it is quoting from accidental poisonings. In which case- the person more likely would have sought out medical aid.

Obviously, AI isn't set up to coach suicide. Quite the opposite. So, I imagine the results would be overly cautious if anything. As in- don't do that- it might result in death.

That's not to say your method won't work. I'm not very knowledgable about hanging to be honest.