Artificial intelligence startup Hugging Face Inc., best known for operating a hub for hosting open-source machine learning and AI models, is launching a new project focused on robotics under the direction of former Tesla Inc. staff scientist Remi Cadene.

The new project, first spotted by VentureBeat, was announced on X, formerly Twitter, by Cadene today in a post where he said that he was joining Hugging Face to start the project and that the company was now seeking engineers to “build real robots in Paris.”

After 3 years @tesla and Optimus, I am thrilled to announce that I joined Hugging Face to start an ambitious open robotics project! (open as in open-source, not as in Open AI) Looking for engineers to build real robots in Paris 🇫🇷 https://t.co/cFuNL4PVI4 🤖🤗 pic.twitter.com/7IkunPXOpX

— Remi Cadene (@RemiCadene) March 7, 2024

In his post, Cadene noted his background of working for three years at Tesla and the Optimus project, which is a bipedal humanoid robot concept under development by the company. Tesla demonstrated the prototype’s second generation in December, showing the robot walking and manipulating objects with its hands.

Until now, Hugging Face has focused entirely on software innovations and AI models, so the pivot to hardware is a vast change in paradigm. However, robotics and AI have always been two industries that intertwine tightly thanks to the interaction among computer vision, contextual understanding, learning systems and collaborative intelligence.

Although there are no details as to exactly where the company is going with the new robotics project, Cadene included a job listing in his post on X for an “Embodied Robotics Engineer.”

“At Hugging Face, we believe ML doesn’t have to be constrained to computers and servers, and that’s why we’re expanding our team with a new opportunity for a Robotics Engineer focusing on Machine Learning/AI,” the company wrote in the job listing.

According to the listing, the new engineer would be responsible for working with machine learning engineers, researchers and product teams to develop new solutions to design and maintain open-source and low-cost robotic systems that integrate AI technologies. Specifically, the company emphasized, that work would be aimed at deep learning and embodied AI.

AI agents that can perceive, interact with and learn from their physical environment such as robots, virtual assistants and other intelligent systems – those with some sort of “body” and senses – are referred to as embodied AI. This is where self-teaching AI robots come into play. These could be household robots and industrial systems that can be asked to do things such as “Please pick up and throw away all the trash on the floor.” The robot ideally would throw away the trash but not toss a child’s toy into the garbage bin.

Tesla isn’t the only company looking to build AI-enabled robotics that can learn and interact in human environments currently. Figure AI Inc. is seeking to raise $500 million to continue its work on its line of humanoid AI-powered robots named Figure 01. And humanoid robot maker Sanctuary AI is developing general-purpose robots with AI control subsystems that mimic memory, sight and hearing to augment human labor.

Image: rawpixel/Freepik