Fei-Fei Li's Atlas world model rebuilds 3D spaces from 3 photos, 50-100x cheaper
Chris_Armstrong · x · 2026-09-05
World Labs co-founders Fei-Fei Li, Justin Johnson, and Ben Mildenhall join a16z's Martin Casado to discuss Atlas, their newly released world model for spatial intelligence.
- Core idea: LLMs use next token prediction and video models use next frame prediction; Atlas is built on next view prediction, the first model to unify pixel generation and reconstruction — two problems computer vision has kept separate for half a century.
- Practical result: digitally capturing a 3D representation of a space becomes 50-100x cheaper — previously 100-300 photos of a room were needed; Atlas works from just three.
- The conversation also covers technical details like the slow-motion shot.
More from Embodied
- YC-backed Antropi Robotics emerges with autonomous CNC factories, already profitable in 6 months — ycombinator · 2026-10-21
- Robotics startup launches Deployments to serve fine-tuned HF checkpoints to robots in real time — arnie_hacker · 2026-10-21
- Amazon FAR's FIBER uses action-aligned latent planning to let Unitree G1 open doors and carry boxes — ChongZzZhang · 2026-10-21
- Sentdex trains robots to go from lying down to a half kneel, eyeing a full stand — Sentdex · 2026-10-21
- NVIDIA open-sources MotionBricks: 350K motion skills at 15,000 FPS, 2ms latency — aftahi_ai · 2026-10-21
- NVIDIA Cosmos3-H-Dreams: Generative Simulation Brings Clinical Surgery to Life — Hugging Face Blog · 2026-10-21