Latent Space·· 2 d ago
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
AI summary
Our first robotics pod with Dyna’s Chief Scientist on their big bet on World Action Models & the Next Frontier of Physical AI
Selection record
Threshold 76Media and individualsFirst 78Second 78
AdmittedSum of both 156 ≥ twice the threshold 152
- Source tier
- Media and individuals; this tier's threshold is 76
- Pre-filter
- passed:世界动作模型、机器人智能体与训练数据等实质AI内容
- Why it was chosen
- Dyna's co-founder breaks down the data recipe and deployment trade-offs behind world-action models, showing the concrete path from demos to production lines for general-purpose robots.
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Latent Space · latent.space