Robot One-Shot Imitation: From Specialized Architectures to Generalist Models
animesh_garg · x · 2026-08-30
A robotics expert compares technical paths from 2020 to today: achieving one-shot imitation previously required carefully designed specialized neural architectures (like Neural Task Graphs). Now, generalist foundation models (e.g., GEN-1.5, Skild S1) enable physical in-context learning simply by inputting video prompts directly into context. While scaling has improved convenience, the author notes a surprising fact: performance rates for one-shot imitation haven't drastically improved compared to the specialized methods achieved in 2020.
Related event: Robot One-Shot Imitation: Bigger Models, Little Real Progress(2 posts)→
More from Embodied
- How Microsoft optimized Windows for Nvidia's Arm-based RTX Spark laptops — The Verge AI · 2026-10-15
- John Deere's second-gen autonomous tillage: 16 cross-checking cameras beat sensor drift — lukas_m_ziegler · 2026-10-15
- Europe's Application-Layer Bet: Can the Rotterdam Humanoid Center Deliver? — ingliguori · 2026-10-15
- VC thesis 'Data at the Edge': cheap sensors plus AI are rebuilding the internet's data flywheel offline — rebeccakaden · 2026-10-15
- PingpongDuo shows humanoid robots playing doubles table tennis — arankomatsuzaki · 2026-10-15
- Robotics data gap: just 300K hours of manipulation data globally vs 100M+ needed, $3B spend projected — 0xSammy · 2026-10-15