Many people have been surprised how quickly open-source AI has kept pace with the AI efforts getting billions in investor funding. It's worth wondering if the same may happen with robotics. After all, robotics are primarily AI too, though embodied in a 3D environment. Recently two major Chinese manufacturers, UBTech Robotics and Xiaomi, introduced an open-source humanoid robot, now there's another. This is from Hugging Face, the popular AI hosting platform, and French robotics firm, Pollen Robotics.
One of the primary dystopian storytelling sci-fi tropes that feeds into popular ideas about AI & robotics, is that corporations will be all-powerful in the future, with 99% of humanity reduced to downtrodden serfs. Yet open-source AI & robots suggest the opposite. They suggest that power would be decentralized and widely available. The more people can meet their basic needs (food, medical care, etc) from open-source AI & robots, the more power drains away from elites trying to hoard and control these resources.
Yup.
I'm glad we're seriously discussing AI safety as a society, but for this exact reason I wish more people questioned whether "obedience" is a good metric of success for it. A paperclip optimisier is bad, but whoever has the password getting unlimited power could actually be more fucked up. A ball of paperclips floating in space is at least benign, once it's finished.