The discussion traces the evolution of AI from the early days of deep learning and the creation of ImageNet to today’s generative models and spatial intelligence. Fei-Fei Li and Justin Johnson reflect on key algorithmic and computational advances—such as Transformers, diffusion models, and neural radiance fields (like NeRF)—that have enabled machines to move beyond 2D image analysis. They emphasize the limitations of language-centric models in capturing 3D reality and advocate for AI systems that perceive, reason, and generate dynamic spatial environments. World Labs’ mission centers on building Large World Models that simulate real-world geometry, physics, and semantics. With applications in AR, VR, robotics, and creative tools like Marble, their vision aims to bridge virtual and physical worlds. Success is measured not just by technical milestones but by enabling intuitive, immersive interactions with AI-generated 3D spaces.