Comment by idle_zealot
Comment by idle_zealot 2 days ago
It means the agent should try to be intentional, I think? The way ideas are phrased in prompts changes how LLMs respond, and equating the instructions to life itself might make it stick to them better?
I feel like you’re trying to assign meaning where none exists. This is why AI psychosis is a thing - LLMs are good at making you feel like they’re saying something profound when there really isn’t anything behind the curtain. It’s a language model, not a life form.