Genesis AI said it aims to build a scalable data engine and open-source ecosystem to enable robust robot deployments. | Source: Genesis AI
Genesis AI, a physical AI research lab and full-stack robotics company, today emerged from stealth with $105 million in funding. The company stated that it is using the funding to develop a universal robotics foundation model, or RFM, and a horizontal robotics platform.
“General-purpose robots powered by physical AI will define the next major chapter of human history,” said Zhou Xian, CEO of Genesis AI. “While digital AI has made extraordinary progress, physical AI — the intelligence that allows machines to perceive, understand, and interact with the real world — has lagged behind.”
“We’re here to change that. By building on the foundations laid by existing digital AI models, we’re bringing human-level intelligence into the physical world,” he continued. “Genesis’ unique approach by fueling digital AI knowledge to drive the emergence of physical AI will deliver unmatched capability, scalability, and cost-efficiency to unlock unlimited physical labor. With 75% of global companies struggling to fill jobs, physical AI is more essential than ever.”
Xian, who earned a Ph.D. in robotics at Carnegie Mellon University in 2024, founded Genesis AI with Théophile Gervet, a former research scientist at Mistral AI and Skild AI. The company includes experts in physics simulation, graphics, robotics, and large-scale artificial intelligence model training and deployment.
AI could help automate a wide range of tasks
Physical labor contributes an estimated $30 trillion to $40 trillion to the global gross domestic product (GDP), but more than 95% of it remains unautomated because of the limitations of current systems, according to the International Federation of Robotics (IFR).
Today’s robots, such as industrial arms, rely on brittle, rigid, and overfitted software stacks, asserted Genesis AI. It said these systems can be narrow in scope, costly to deploy, and challenging to scale.
The company said it aims to enable the next generation of general-purpose robots by providing robustness, flexibility, and cost efficiency. Genesis, which has offices in Paris and Palo Alto, Calif., said it plans to bring a “data-centric, full-stack approach to physical AI.”
It is building a “scalable and universal data engine that unifies high-fidelity physics simulation, multimodal generative modeling, and large-scale, real robot data collection.” The company claimed that its simulation stack, developed entirely in-house, will generate rich synthetic data at scale.
Genesis AI is also working on an efficient and scalable real-world data collection system. This dual engine of synthetic and real data will bridge historically siloed domains to create a large-scale, diverse, and high-quality dataset to train RFMs, it said.
Save now with early bird discount
Genesis AI to make components open-source
Eclipse and Khosla Ventures co-led the round. Bpifrance, HSG, and tech leaders Eric Schmidt and Xavier Niel also participated in the funding.
“Even in the most ‘automated’ industries today, the robot-to-human ratio rarely exceeds 1:30, due to the long tail of tasks requiring dexterity, cognition, mobility, and real-world reasoning that current robots simply can’t handle,” said Charly Mwangi, a partner at Eclipse.
“General-purpose robotics is the breakthrough we’ve been waiting for and stands to impact trillions in labor value across sectors,” he added. “Genesis has the vision, strategy, and world-class team to define the era of physical AI in order to unlock unlimited physical labor through general-purpose robotics.”
Genesis AI also plans to open-source components of its data engine and foundation model to empower developers, researchers, and partners to build on its breakthroughs and accelerate progress across the broader field of physical AI. The company said that with about only 2,000 AI robots deployed worldwide, its technology could help scale robots across manufacturing, healthcare, logistics, and more.
Tell Us What You Think!