Stop Simulating, Start Experiencing - by Paolo Di Prodi
paoloai.substack.com · 6,528 words · saved by 1 readers
The empirical and developmental case for sandbox-trained robots
Nowadays most if not all research in robotics has been focusing on simulation first, reality second: you take a high-fidelity replica of a robot in a simulator (Newton, MuJoCo, Gazebo, Isaac), you run all your RL magic on it until it can walk, jump or whatever martial art you fancy, and then you apply the learned policy or VLA into the real robot. Landmark examples include OpenAI’s Dactyl (Andrychowicz et al., 2020), which learned dexterous in-hand manipulation entirely in simulation, and ANYmal (Hwangbo et al., 2019), which transferred agile locomotion policies to a physical quadruped. More…
saved by
related reading
- State of Robot Learning, December 2025vedder.io
- Robots Must Be Ephemeralizedevjang.com
- Building Worlds That Train Robotsworldlabs.ai
- Many Small Steps for Robots, One Giant Leap for Mankindnotboring.co
- Eric Jang: Robots Must Be Ephemeralizedblog.evjang.com
- How Claude Performs on Robotics Tasks \ Anthropicanthropic.com
- How Can We Make Robotics More like Generative Modeling? | Eric Jangevjang.com
- All Roads Lead to Robotics | Eric Jangevjang.com
- Frontier of General-Purpose Robotics (2025)adampatni.com
- Making Robots Useful with RLyoutube.com
- Learning Sim-to-Real Humanoid Locomotion in 15 Minutesyounggyo.me
- The Reality Gap in Roboticsrobotics-reality-gap.github.io