Skild AI Unveils S1: Robotic In-Context Learning with One Demo
CyberRobooo · x · 2026-08-26
Skild AI has released S1, a model enabling in-context learning for robotics.
- Zero-Fine-tuning, Long Tasks: Using just a single video demo and no fine-tuning, it executes novel 10-minute+ tasks (e.g., pancake flipping, potting plants, pour-over coffee) unseen during pre-training.
- Efficiency & Scaling: One example reportedly matches hundreds of post-training samples for conventional VLAs; the performance gap on unseen tasks widens exponentially with scale.
- Paradigm Shift: The author compares this to the BERT-to-GPT-3 transition, suggesting that pre-training's real job is teaching the model how to learn from context, as robotics develops its own scaling laws.
More from Embodied
- Robotics LLMs tested on 16 tasks: GPT-6 Astra hits 17.2% success vs Fable's 9.1% — pabbeel · 2026-10-16
- WEDGE paper generates training environments conditioned on navigation agent failures — _akhaliq · 2026-10-16
- Viral meme nails AI's irony: it can write symphonies but can't do the dishes — briannekimmel · 2026-10-16
- Knödel Watch: animated vocabulary cards on ESP32-S3, open-sourced 4 days after unboxing — alexcovo_eth · 2026-10-16
- NVIDIA RTX Spark PC launches tomorrow with global events and strong early reviews — NVIDIA Blog · 2026-10-16
- SlipSense fuses pressure and vibration for 23ms zero-shot robotic slip detection — micoolcho · 2026-10-16