Humanoid Robot Funding Tops $3B as Brian Potter Explains the VLA Models Driving Robots
scottleibrand · x · 2026-10-18
Brian Potter's Construction Physics essay explains how vision-language-action (VLA) models—the "policies" controlling robots—actually work. Context: NEURA Robotics raised $1.4B, Figure AI $1B, Apptronik $520M, Unitree went public raising $900M, and Agility Robotics plans a SPAC listing. Demos include Figure sorting packages and unloading dishwashers, Physical Intelligence making coffee and folding boxes, and Generalist AI's GEN-1.5 learning tasks from few examples. His core point: most robotics progress comes from AI models rather than hardware, and while robot AI capabilities remain limited today, they could accelerate rapidly as general AI improves.
More from Embodied
- Fishing in VR while watching your real TV: passthrough makes mixed reality living-room setups work — Scobleizer · 2026-10-18
- Snabbit could make huge money selling household robotics training data — AashaySachdeva · 2026-10-18
- Apple's immersive horror short Remnants wows Vision Pro users with 3D presence — Scobleizer · 2026-10-18
- TRACE Open-Sourced: Migrating G1 Controller to Full-Size Humanoid H2 in 8 Hours on One GPU — zhengyiluo · 2026-10-18
- MIT's SANDO Drone Planner Avoids Every Moving Obstacle Across 12 Real Flights — Olivier__OG · 2026-10-18
- SimInfinity: open-source robotic simulation benchmark with 10K real-world tasks — joemeno · 2026-10-18