The default is brute-force synthetic data — and its interpolation assumption
KyleMorgenstein · x · 2026-11-17
The author summarizes the common approach in robotics for contact-interaction problems: synthetic data generation over a huge distribution of objects, betting that all real contact interactions lie within that set so the policy can interpolate. He contrasts this brute-force data route with methods that find better representations first, which he sees as more promising for sample efficiency and safety.
More from Embodied
- Robotics researcher: edge cases are literally the entire problem for robot deployment — chris_j_paxton · 2026-11-17
- Better representations first: where does data scaling break down in robot learning? — KyleMorgenstein · 2026-11-17
- Policies use only a sliver of image information — extracting it matters for safety — KyleMorgenstein · 2026-11-17
- Enactra shows agentic robot simulation with controllable worlds and per-frame annotations — Lianhuiq · 2026-11-17
- South Korea's robot density hits 1,066 per 10k workers, 7x the world average — randal_olson · 2026-11-17
- Open-source Human Models MCP connects AI agents to MuJoCo musculoskeletal simulation — usefulwrong · 2026-11-17