While the first wave of the AI revolution was dominated by software – language models, image generators, chatbots, and coding assistants – the next wave is taking shape in the physical world. Physical AI refers to artificial intelligence systems that perceive and act in three-dimensional reality: robots, autonomous vehicles, drones, and smart factories. The companies building in this space represent some of the most well-funded startups in the world right now, backed by the conviction that AI’s significant impact on physical industries will ultimately be larger than its impact on software alone.

Nvidia CEO Jensen Huang has been the most vocal executive in framing this shift, repeatedly describing ‘physical AI‘ as the next trillion-dollar opportunity in technology. Nvidia’s CUDA platform, which powers most AI model training today, is being extended specifically for robotics and autonomous systems through the company’s Isaac and Cosmos platforms designed for training AI in simulated environments before deployment in the real world.

The Companies Leading the Physical AI Wave

The field is populated by companies attacking different parts of the problem. Some are building the robots themselves, others are building the software platforms that make robots trainable, and others are creating the simulation environments in which robot AI learns before touching the real world.

• Figure AI: Building general-purpose humanoid robots for factory floor deployment, backed by Microsoft, Nvidia, Intel, and Jeff Bezos personally.
• 1X Technologies: Norwegian startup building humanoid robots with a focus on home assistance, backed by OpenAI.
• NEURA Robotics: German company that just raised $1.4 billion from Nvidia, Amazon, and Tether for industrial humanoid deployment.
• Apptronik: University of Texas spinout building the Apollo humanoid, partnering with NASA and major manufacturers.
• Waymo: Alphabet’s autonomous vehicle subsidiary, the most commercially deployed autonomous driving company in the world with paid robotaxi service in multiple US cities.
• Tesla: Developing Optimus humanoid for internal factory use while continuing to develop Full Self-Driving for its vehicles.
• Agility Robotics: Amazon-backed company whose Digit robot is already deployed in Amazon warehouses.
• Boston Dynamics: Pioneer of dynamic robot locomotion, now owned by Hyundai and deploying Atlas for industrial applications.
• Skydio: American autonomous drone company building AI-powered inspection and mapping systems for infrastructure.
• Machina Labs: Applying AI and robots to metal forming, dramatically reducing the cost and time to produce complex metal parts.
• Covariant: Building the AI ‘brain’ that connects across different robot hardware platforms, enabling robots to learn from each other’s experiences.

Why Now?

The convergence of three trends has made physical AI viable at scale for the first time. First, large AI models trained on internet data have given machines a broad understanding of the physical world that would have taken decades to program manually. Second, hardware improvements in actuators, sensors, and edge computing have made the physical components of robots cheaper and more capable simultaneously. Third, simulation technology has matured to the point where robots can accumulate millions of hours of training experience in virtual environments in the time it would take to accumulate weeks of real-world experience.

Frequently Asked Questions

How soon will humanoid robots be common in workplaces?

Industry analysts project meaningful humanoid robot deployment in industrial settings by 2027-2028, with broader commercial availability by 2030. Home deployment at scale is likely a decade away given the additional complexity of unstructured home environments.

Will physical AI take jobs?

Physical AI will automate specific tasks in manufacturing, logistics, and agriculture. The broader economic impact – whether it creates more jobs than it eliminates – depends on the pace of deployment and the adaptability of the workforce, and remains genuinely uncertain.