Physical AI represents the next phase of the artificial intelligence (AI) revolution, as software-driven intelligence evolves into embodied, autonomous machines operating in the physical world. As advances in edge computing, sensing, robotics and AI converge, intelligence is moving out of the data centre and into physical systems such as factories, warehouses, hospitals and transport networks.
The WisdomTree Physical AI, Humanoids and Drones UCITS ETF provides targeted exposure to this transformation by investing in companies developing embodied, autonomous systems and the hardware, components and platforms that enable intelligent machines to operate in real-world environments.
Artificial intelligence has historically delivered value through software applications, with many of us now relying on AI for various day-to-day tasks in the virtual world. But unlike digital AI, physical AI is about machines interacting with the real world, impacting the real economy, and addressing structural challenges such as labour shortages, productivity constraints, and operational safety across global industries.
Physical AI | Digital AI | |
|---|---|---|
Operational realm | Real-world, physical environments (for example, factories, roads and homes) | Digital and virtual environments (for example, software platforms, cloud and the internet) |
Form factor | Integrated hardware that perceives the environment and takes physical action (for example, humanoid robots, drones, autonomous vehicles) | Exists entirely as software, algorithms and data; interacts with information |
Applications | Manufacturing automation, autonomous transport, healthcare robotics, smart warehouses | Customer service, finance, data management, search and analytics |
For more information on digital AI and AI value chain investing, please see our artificial intelligence strategy page.
Several long-term forces are accelerating adoption:
Blake Heimann
Leading AI labs are scaling revenue at an unprecedented pace, but this growth depends on a vast physical infrastructure buildout. From semiconductors and servers to networking, power and data centres, a multi-year investment cycle is reshaping the AI value chain. Discover how the WisdomTree AI Infrastructure UCITS ETF provides exposure to the companies enabling AI at scale.
Baoqi Zhu
Physical AI is beginning to move from concept to commercial deployment, with applications increasingly appearing in factories, transport networks and industrial automation systems. China is playing an important role in this trend, supported by its manufacturing ecosystem, hardware supply chain and growing group of robotics companies. This blog looks at three Chinese companies, UBTech, Pony AI and Shenzhen Dobot, which illustrate different paths to physical AI adoption across humanoid robotics, autonomous mobility and smart manufacturing.
Ayush Babel, Baoqi Zhu, Blake Heimann, Elvira Kuramshina
The Nasdaq-100 recorded its strongest monthly performance in over 23 years, but the underlying story extends beyond the headline index. A range of technology themes, spanning AI, robotics, and quantum computing, are gaining momentum, with several innovative companies and thematic strategies outperforming the Nasdaq-100 in April. This broadening opportunity set highlights how growth is being increasingly driven by diverse and emerging areas across the tech landscape.
Baoqi Zhu
Recent rallies in AMD and Intel have brought CPUs back into the AI infrastructure debate. While GPUs remain the central engine for AI training and inference, the next phase of AI workloads may require much more supporting compute around GPU clusters. Agentic AI involve more planning, routing, tool calls, code execution, verification and data movement, many of which are CPU-intensive tasks. This suggests the AI semiconductor opportunity is broadening beyond GPUs and CPUs could be the next bottleneck of AI Infrastructure. 
