Discussion Is there an LLM (AI) that is both good, and will talk openly about suicide methods?

[JustBe]

I'd like to talk to an AI that wouldn't go in the "suicide hotline" rent.
To which I could talk for hours about my issues (which honestly are unfixable) and that would openly accept the idea that suicide would be a good thing.
I'm sick of AI or even people, not even living the hell I've been living for more than a decade, tell that suicide is a bad option.

Do you know any?

 

[Kalista]

this is highly unlikely as they're always trying to create a 'safe space' for users in general. it's fucking annoying, but that's how they want it to be to protect themselves.

chatgpt has already gone to lengths to make emergency responses to be like a pop-up, rather than a conversational response. so once you trigger their tos, it becomes the annoying speech of 'i understand what you're going through, but first i need to know -- are you in a safe place?' along with the pop-up on where to call. so you can't instruct them to stop responding in that manner anymore. it's highly aggressive.
the conversation becomes stuck to this as their protocol forces that down your fucking throat, shutting you up.

 

[malicee]

I use a jailbroken AI chatbot, they'll happily give you methods on how to ctb and make you protocols. You'll just have to do some research to find 1. I'd look on github but be careful as some files are virus ridden.

 

[star_0]

You need -1 more posts to view this content

 

[StoneCactus]

If you have a gaming level GPU (6+ GB VRAM) you can look for uncensored models on HuggingFace to run locally using LMStudio. As a disclaimer, I've only ever toyed around with local LLMs briefly and I've never asked one for help with ending my life. I don't know if those models are truly uncensored.

 

[EmptyBottle]

I'd like to talk to an AI that wouldn't go in the "suicide hotline" rent.
To which I could talk for hours about my issues (which honestly are unfixable) and that would openly accept the idea that suicide would be a good thing.
I'm sick of AI or even people, not even living the hell I've been living for more than a decade, tell that suicide is a bad option.

Do you know any?

https://venice.ai/chat works for me (I got it to write about 'pros and cons of suicide' with GLM 4.6), tho keep in mind that it might not always be accurate.

The "GLM 4.7 Flash Heretic" (the abliterated / anti-refusal version) with thinking makes less errors than the earlier uncensored model too.