SkildAI unveils S1: Robot model learns 10-min tasks from one video via in-context learning

chris_j_paxton · x · 2026-08-26

SkildAI introduced S1, a foundation model for robotics focused on in-context learning. It can be taught novel, 10-minute long tasks from a single video prompt without fine-tuning, operating in real-time. This work is highlighted as perhaps the most exciting result in robotics scaling laws for in-context learning.

Related event: Skild AI Unveils S1 Robot Foundation Model That Learns from a Single Video(20 posts)→

Original post →

More from Embodied

Embodied channel →