Genesis AI, a startup focused on developing foundational models for powering various robots, has emerged from stealth mode after securing $105 million in seed funding. This round was co-led by Eclipse Ventures and Khosla Ventures. Founded in December by Xian Zhou and Théophile Gervais, the company aims to create a universal model that enables robots to automate repetitive tasks, ranging from lab work to household chores.
Unlike large language models trained on extensive text datasets, AI models for robots require training on physical world data, which is costly and time-consuming to obtain. To address this challenge, Genesis AI is turning to synthetic data. The company uses a proprietary physics engine capable of accurately simulating the physical world to generate synthetic data.
This synthetic data engine originated from an academic project led by Xian Zhou in collaboration with researchers from 18 universities. Several participants from this project have joined Genesis AI, forming a team of over 20 researchers specializing in robotics, machine learning, and graphics.
Genesis AI claims its proprietary simulation engine allows faster model development, offering a distinct advantage over competitors relying on NVIDIA software. The company is developing its synthetic data and foundational models in offices located in Silicon Valley and Paris, with plans to release its model to the robotics community by the end of the year.
