Improving AI Alignment Through Human Feedback
We are improving our AI systems’ ability to learn from human feedback and to assist humans at evaluating AI.
Our goal is to build a sufficiently aligned AI system that can help us solve all other alignment problems.