News for hackers. Stuff that matters.
Re:AI safety leader says 'world is in peril' and quits to st... by CrimsonCape on Friday February 13, 2026 @04:40PM (#47008204)
> his contributions included investigating why generative AI systems suck up to users

Why does it take research to figure this out? Possibly the greatest unspoken problem with big-coporate-AI is that we can't run prompts without the input already pre-poisoned by the house-prompt.

We can't lead the LLM into emergent territory when the chatbot is pre-engineered to be the human equivalent of a McDonalds order menu.

Re:AI safety leader says 'world is in peril' and quits to st... by Mentlo on Sunday February 15, 2026 @03:26AM (#47022016)
I find your belief that what is needed for emergence is better prompting … amusing.

The ai would still be sycophantic even without the pre-prompt. It’s been reinforced to do so, it’s baked in the weights.