World Labs co-founder Justin Johnson on Atlas world models and the future of 3D AI
a16z Podcast · rss · 2026-09-13
In an a16z podcast, World Labs co-founder Justin Johnson discusses Atlas, the startup's latest world model, and the case for AI that understands and acts in the physical world.
- Atlas's three core tasks: generating new worlds, reconstructing real environments from images, and simulating how objects or robots behave in them.
- The thesis: just as LLMs became general-purpose engines for text, world models could become a horizontal layer for visual and physical intelligence across entertainment, gaming, construction, and robotics.
- Games and creative tools: world models could reshape both, with precise spatial control being key.
- Real-to-sim-to-real robotics: a few photos of a physical space may eventually suffice to build a simulation and adapt a robot to it.
More from Embodied
- Proxy Policy Steering adapts frozen VLA models to new tasks at inference time — weichiuma · 2026-09-14
- Tesla China to exhibit Cybercab in Shanghai and Beijing from September 17 — yunta_tsai · 2026-09-14
- MIT's HardFlow enforces hard constraints on generative AI without retraining — MIT News AI · 2026-09-14
- MIT spinout Atlas turns plastic waste into building materials with AI robotics — MIT News AI · 2026-09-14
- UBTech Opens World's First 10,000-Unit Humanoid Robot Factory; SK's AI Data Center Hits 900MW — 创业邦 · 2026-09-14
- Chinese Robot Dog Maker Vbot Lands 5,000 Orders at IFA, Tops 13,000 Units Sold Globally — 创业邦 · 2026-09-14