technology

I suppose this was inevitable. Their "terms of service" will probably protect them from other things too, like if it tells people to drink bleach or something, they'll say it violates TOS to follow hallucinatory directions from it.

ChatGPT gave him a pep talk and then offered to write a suicide note

what function are terms of service if when you break them your service is unaltered

So openAI decided that a clause in the TOS was a good enough guardrail against giving out info on how to kill yourself? And this was after multiple instances of them deliberately putting in guards against other behavior that they didn't want?

That's a pretty fucking stupid legal case.

AI company founded by the "car company whose self driving turns itself off a second before collision to blame the driver" guy is using the "this bong is for tobacco use only " defence for their suicide coach AI.

Yeah because if there's one thing depressed and people that want to self-harm read it's the tos on the magic computer.

I’m going to die somewhat on this hill, but it just feels like a lot of this is just a reaction to llms/ai and not really about suicide and just comes across as weaponizing it. Mainly because, to quote

First and foremost, we have to acknowledge and reinforce the fact that suicide cannot be attributed to one single cause.

Listening to a particular type of music or belonging to a group associated with it is no more blameworthy than the influence of books, television, movies, or video games. Although each of these can be risk factors for a vulnerable person, a multi-dimensional approach that takes into account all individual psychological elements is necessary to ascertain with certainty if an individual is considering suicide.

As a society, we also must not react irrationally to phenomena like suicide that we do not fully comprehend. Yet because suicide is so enigmatic and hard to fathom, it can be tempting to isolate a single factor, especially a social one such as rock music, to make sense of it. In some rare cases, societal influences like suggestive lyrics in rock music could be the precipitating cause that sends an extremely at-risk individual over the edge. However, we could never say that rock music or participation in a sub-culture associated with a particular style of music directly caused an individual’s suicide.

What causes an individual to entertain suicide is multi-dimensional. Through education, and by having constructive conversation about what the common risk factors for suicide actually are, we can de-fang it, have a greater understanding of it, and no longer fear it.

from https://www.suicideinfo.ca/local_resource/musicandsuicide/

another thing to but it comes across as just wanting to punish people for not wanting to talk about suicide anywhere at all. In a sort of taboo religious sense that you can’t talk about suicide, that even to chatbots it should be censored to. Like I like to talk to chatbots about suicide since it helps relief things for a bit. Is it a solution? No, but it does help a bit! I’d also rather talk to a chatbot at this point than being told to just “fuck the fuck off” while getting degendered.

That isn’t to say openai isn’t a risk factor, or that AI really shouldn't be easily handling suicide advice over. but on the other hand it is pretty fucking annoying wanting to talk about suicide and then being unable to due to the reaction the other way where you can’t just talk about it at all. even to a bot, because everything censored!

My response is more so I really dislike it when ai is the sole blame of cause.

If I ever kill myself, which I think about a lot of days and talk to a chatbot about. It’s not gonna be because of the chatbot! It’s gonna be because I can’t stand living with my dad drinking himself to death, grief over my mom or other things, or it’s gonna be because I live in a red state that only gets more transphobic. If I ever did kill myself, I swear to fucking god if a chatbot gets blamed for it. I’ll lose my mind. Living somewhere transphobic? No no! Must be the chatbot! My dad drinking himself to death? No no! Chatbot the reason if I ever killed myself!

Like why is someone talking to an ai and discussing suicide methods? Cant be because people are just general assholes and don’t want to hear it, or are quick to punish you with threatening to call the police on you and lock you in a psych ward! I'd rather talk to a chatbot.

And this is the only time suicide even comes up in conversations at all! Is when it’s over ai, but where is discussion over suicide before that?

Why aren't articles like this posted? https://web.archive.org/web/20221001035906/https://anti-imperialism.org/2018/12/12/every-suicide-is-murder-capitalism-and-mental-illness/ [Every Suicide is Murder: Capitalism and Mental Illness from anti-imperialism org]

I just don't care anymore. Suicide just a joke for everyone and only time anyone takes it seriously is after the matter happens, or it's reaction to stuff like this, but when time move forwards suicide will just be pushed back into the box where you shouldn't talk about it. then something new will come along blaming things for suicide, and people will pretend to care.

I say we bring back Scaphism for the AI dorks