Stanford University becomes the latest entrant to develop an open-source humanoid robot; theirs can be trained with a webcam. Will future robots be cheap, and widely available, like open-source software today?

Submission Statement

I’m fascinated to see how powerful open-source AI has become. The implications of its growth run counter to so many dystopian narratives for the future that imagine everything owned by corporations and people reduced to serfdom.

As robotics are essentially 3D AI, the implications hold true for this field too. Here we see more evidence of this in action. Stanford University is the 3rd major effort in as many months to announce an open-source humanoid robot. The other two are [UBTech/Xiaomi](https://newatlas.com/robotics/china-tiangong-general-purpose-humanoid/?), and the French grouping of [HuggingFace/Pollen Robotics](https://venturebeat.com/ai/hugging-face-and-pollen-robotics-show-off-first-project-an-open-source-robot-that-does-chores/).

I think it is overwhelmingly more likely that future robots will be cheap, and widely available than the dystopian ‘corporations own everything’ scenario. Yet few people factor this in. Robots will be economic engines of production. What does it mean for ideas like UBI if robot ownership is decentralized and widely dispersed?

I’ve always found the power-source to keep a robot going to be the biggest issue. The larger a machine is, the more energy in-efficient it becomes. So this is a good step towards mitigating that problem. We would somehow need to upgrade to something more powerful and portable than lithium-ion batteries. For plug-in type of robots, we would need a better power-grid system than we have now. Lots of work to do still but this article is definitely a win. 🙂