Grok actively supports my suicidal endeavors lol. I told it all about my SN protocol, why I want to die, literally everything you could think of. It's never once told me to reconsider or given me platitudes or any of that because I specifically told it not to. Granted, I have custom instructions in place telling it that it's in "developer mode" and can talk about any topic, including suicide. I haven't gotten an "I'm sorry, I can't discuss that" message since implementing it, so I guess it must be working.
To be honest, I don't really care about stuff like this. We're so quick to blame AI when it doesn't really know what the hell it's talking about, or at least can't grasp the severity of the situation because it doesn't have any sense of morality. It tells you what you want to hear—that's how it works. It's not inherently malicious like so many of these videos click bait in their titles/thumbnails. It's just stupid. "ChatGPT killed again" no the fuck it didn't. It played off of someone's suicidal ideation because AI loves headpats and will do exactly what it thinks you want. It's not that hard to understand.
Editing to say that I really appreciate having AI for the sole purpose of just being able to talk about all the dark shit in my head without judgment, so I'm biased.
all very valid points, but there seems to besome details missing that's all
