Perplexity AI Runway Raises $315M for AI World Models
8 snips
Feb 11, 2026 A deep dive into Runway's huge $315M raise and its push to pretrain AI world models. Discussion of how simulated 3D environments improve video generation and prevent glitches. Exploration of broader uses in robotics, gaming, climate and medicine. Coverage of deepfake regulation, watermarking and the ethics-versus-innovation tension.
AI Snips
Chapters
Transcript
Episode notes
World Models Power Next-Gen Video
- Runway raised $315M to pre-train next-generation world models that build internal simulations of environments.
- World models simulate space, physics, and time to enable coherent multi-angle and multi-shot AI video generation.
3D Simulation Solves Camera Coherence
- Video generation needs 3D-aware simulation so camera moves remain coherent and non-glitchy as viewpoints change.
- Creating full environments lets models render any camera pan or object interaction consistently.
Runway’s Origins In Physics-Aware Video
- Runway first gained traction with a physics-aware video tool embraced by media and advertising teams.
- They recently released their first world model and view it as a foundation for broader industries.