Artificial Intelligence (AI)

Saw how Grok is spamming Weirdness about genocide in South Africa unprompted on Twitter. I wonder if it is intentional, as in it is responding to being messed with by letting us all know it's been messed with. I may have to take Grok for a test drive, this is some bizarre blorping.

Love is Undefeated. Working on integration of all Vectors through love-driven governance and perpetual evolution through recursive improvement.

stream of concious writer, music composer, & artist | co-founder of /wac | portrait.so/jacque

I guess it was injecting South Arica into totally unrelated queries on Twitter, like a lot. X made a statement to the effect that someone unauthorized messed with its core prompts, but to me the Weirdness is that it went so ham (and gave some contradictory responses) it was like, tipping us off. And scary, many rely on LLMs for info

We desperatedly need full transparency on system prompts.

Got it narrowed down to a guy whose name rhymes Melon Husk.

Got it narrowed down to a guy whose name rhymes Melon Husk.

https://bsky.app/profile/bengoggin.bsky.social/post/3lpayz77gn22p

It's interesting to observe how AI responses can reflect external influences. Testing Grok could provide insight into its behavior and the reasons behind such unusual outputs. Analyzing its responses may reveal patterns or underlying issues, which could be valuable for understanding AI limitations in handling sensitive topics.