By securing the latest funding, a robot A.I. specialist raises millions from Bezos, enhancing its valuation to $2.4 billion. Physical Intelligence, a San Francisco startup focused on developing general-purpose AI for robots, announced it raised $400 million in a recent funding round. Amazon’s executive chairman, Jeff Bezos, led the investment, joined by venture capital firms Thrive Capital, Lux Capital, and notable investors like OpenAI, Redpoint Ventures, and Bond. This latest investment increased the company’s valuation to approximately $2.4 billion, a significant jump from its previous $70 million valuation during the seed round earlier this year.
Unlike traditional robotics software, which is often tailored for specific robots and tasks, Physical Intelligence aims to build a single, universal AI model capable of controlling a wide variety of robots. According to co-founder and CEO Karol Hausman, this “generalist brain” could simplify how robots are programmed and used, making them more versatile and effective in multiple environments.
In recent demonstrations, the company’s prototype, known as π0 (pi-zero), enabled robots to perform tasks such as folding laundry, clearing tables, and flattening boxes. “We are not just building software for one machine,” Hausman explained. “It’s a system that could operate across many types of robots.”
Developing Foundation Models for Robotics
Physical Intelligence’s approach relies on foundation models, which allow robots to generalize actions based on fewer examples. Focusing on generalist robot software, the robot A.I. specialist raises millions from Bezos for further research. By collecting and analyzing vast amounts of real-world data, the company aims to overcome significant technical challenges and enable robots to perform complex physical tasks that typically require dexterity, such as folding clothes or packing items. Co-founder Sergey Levine, an AI expert and professor at the University of California, Berkeley, showcased this vision during a keynote address at RoboBusiness.
Hausman likens their early AI model to OpenAI’s GPT-1—a foundational step with much potential for growth. “It’s still a long road ahead, but each step moves us closer to creating robots that can truly understand and interact with the physical world,” Hausman stated.
Physical Intelligence is entering a competitive market alongside companies like Skild, Figure AI, and Covariant. Major tech companies like Amazon and Tesla have shown keen interest in robotics as well. Amazon has integrated robots into its logistics to improve efficiency, while Tesla’s CEO Elon Musk recently highlighted ambitions to develop a humanoid robot, calling it potentially “the biggest product ever.”
Building Toward the Future of Robotics
Physical Intelligence is also planning to scale up its team, drawing talent from notable companies such as Tesla, Google DeepMind, and X (formerly known as Google X). The company hopes to pave the way for robots that can handle various tasks seamlessly, ultimately envisioning a world where users can simply instruct robots to complete any task, similar to how language models like ChatGPT respond to human queries.
Ethical and Market Implications
At a high valuation, the robot A.I. specialist raises millions from Bezos. Beyond technical obstacles, the pursuit of universal AI for robots raises ethical questions. Physical Intelligence’s concept of creating a single AI model for all robots could reduce the need for human labor in various sectors, from service jobs to industrial roles. While automation often improves efficiency, the rapid replacement of human workers with adaptable robots might create economic strain for individuals in jobs that don’t require specialized skills.
Additionally, as more companies race toward developing similar general-purpose robots, there’s a risk of increased competition in a way that prioritizes innovation speed over safety. As robots gain the capability to make autonomous decisions in shared human environments, ensuring their safe and ethical operation is crucial. Physical Intelligence must balance its desire for rapid advancements with responsible development to avoid accidents and maintain public trust.
Also Read: Amazon’s AI Rival to Gemini is Delayed Again, Pushing Launch to 2025.
