Chelsea Finn says robotics is bottlenecked by rollout cost and memory
chris_j_paxton · x · 2026-07-28
Chelsea Finn: robot scaling is bottlenecked by rollout cost and missing memory
The post highlights two key constraints in robotics foundation models:
- RL scaling is expensive in the physical world. A million 1-minute trajectories would take roughly 700 robot-days, so scaling PPO/GRPO-style post-training is fundamentally different from LLM training.
- Memory is still missing. Most SOTA robot systems have effectively no memory, and naively extending video history is too costly: 10 seconds at 50 Hz across 4 cameras is about 500K tokens.
The cited approach uses short-term video memory plus long-term text summaries of the prior 10–15 minutes, which helped enable a fully autonomous multi-step kitchen workflow.
Related event: Chelsea Finn Identifies Physical Rollout Costs as RL Bottleneck(2 posts)→
More from Embodied
- GPT-6 tested on LIBERO robot task: turns on stove, fails to grasp moka pot — YuXiang_IRVL · 2026-09-23
- OpenRoboto Shift launches: decentralized egocentric video data network for robot brains — markjeffrey · 2026-09-23
- OpenRoboto launches decentralized rival to Figure's robot data network on Bittensor — markjeffrey · 2026-09-23
- Typesafe's decision model Jev drives a Go2 robot inside NVIDIA Isaac Sim via OM1 — paigeinsf · 2026-09-23
- Tesla pitches Optimus as the first general-purpose humanoid robot — XFreeze · 2026-09-23
- XPENG's IRON robot demos full-duplex speech with 9-mic array and lip reading — ChrisGPT · 2026-09-23