Comment on Google AI chatbot responds with a threatening message: "Human … Please die."

<- View Parent
a1studmuffin@aussie.zone ⁨4⁩ ⁨weeks⁩ ago

The preceding message is really quite an undefined input, as the user copy/pasted some questions from their assignment without phrasing it as a question or cleaning up the formatting.

I wonder what kind of outputs you would get from LLMs if you’d been talking sensibly on certain subjects then started to feed it garbage input. It feels like this might be what happened here.

source
Sort:hotnewtop