Paul Christiano: Staying in Control of Self-Improving AI | The Hinton Lectures 2026 (Night 2)
AI 要約 Recursive self-improvement in artificial intelligence threatens to accelerate capability gains beyond human control, potentially creating broadly superhuman systems within 18 months. This rapid trajectory necessitates immediate shifts in institutional governance, as current…
- Recursive Self-Improvement (RSI)
- Alignment Problem
- Reinforcement Learning from Human Feedback (RLHF)
1 h 27 minEnglish要約 · 文字起こし








