Skild's S1 robot watches one wheel-mount demo, then self-corrects when the wheel misaligns
heyshrutimishra · x · 2026-09-10
Skild's S1 robot showed human-like task generalization: after watching a single video of a human attaching a wheel, it re-aligned the wheel on its own during deployment when it didn't fit — the prompt video contained no recovery demonstration.
Per Deepak Pathak, S1 watches the prompt to grasp the goal and improvises to reach it, retrying when it makes mistakes — a line between scripted motion and autonomous intent.
Related event: Skild S1 Robot Self-Corrects After Misaligning Wheel(4 posts)→
More from Embodied
- StackChan robot performs an AI-generated skit, orchestrated by Hermes Agent via JSON script — alexcovo_eth · 2026-10-27
- DONIC hits 86.79% on bimanual dexterous tracking, crushing prior SOTA's 9.43% — liuziwei7 · 2026-10-27
- Bedrock Robotics raises $600M as its fully autonomous excavators do paid work on live job sites — Sethwinterroth · 2026-10-27
- AW-0 hits 80.8% on RoboTwin tasks, beating strongest baseline by 16 points — trevordarrell · 2026-10-27
- AW-0's code stream fixes world model prediction: picks mango where baseline grabs apple — trevordarrell · 2026-10-27
- AW-0 runs at 230ms on a single RTX 5090, 2.4x faster than autoregressive code planning — trevordarrell · 2026-10-27