World-R1: Reinforcing 3D Constraints for Text-to-Video Generation - Daily Paper Cast | Wave AI Podcast Notes