You Can't Tell What's Real Anymore. Here's What Comes Next - Cristóbal Valenzuela, Runway

Why watch

Runway co-founder Cristóbal Valienza traces the evolution of AI video generation from 128x128 pixel outputs in 2019 to photorealistic video indistinguishable from real footage. He explains how next-frame prediction at scale enables models to learn physics, object permanence, and world dynamics. The talk explores world interface models, robotics applications, and Continuum—a prototype enabling any screen element to become a live interactive video. For AI/ML engineers and researchers exploring generative video, world models, and human-computer interaction.