AI Alignment & RLHF

Why brilliant AI can still go terribly wrong. And how we fix it.