@lord.luke5: Comment “QUACK” 🦆 and I’ll send you the GitHub repo. This tiny robot is being trained in simulation, then those skills are transferred into the real world. Pollen Robotics just open-sourced the reinforcement learning stack behind Microduck, showing how sim2real robotics can let developers train behaviors virtually and deploy them onto physical robots. This is the kind of open-source robotics infrastructure that could make building with AI + robots way more accessible. Would you actually build one of these? [Microduck, Pollen Robotics, open source robotics, reinforcement learning, sim2real, AI robotics, robot training, GitHub robotics]

Luke

Open In TikTok:

Region: US

Friday 28 August 2026 04:12:15 GMT