Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path? | AI Safety Fundamentals | Podwise