Physical AI Leaves Screens for Robots, Drawing Investor Capital

At VettaFi’s Q2 Market Outlook symposium, Rafael Silva, a research analyst, described physical AI as “artificial intelligence that steps off the screen and into hardware.” He said the technology is being embedded in robots, vehicles and other machines so they can sense, navigate and interact in real time outside data centers.

Silva described the transition as moving advanced neural networks from models and software into physical devices. He highlighted advances in edge computing and so-called world models that allow machines to process data locally and to model the physics of their surroundings. He used the phrase “blind and laggy” to describe earlier generations of robots that lacked that capability.

For investors, Silva indicated physical AI creates exposures that broad megacap indexes may not capture. He pointed to thematic ETFs that use in-house scoring to rank companies by revenue concentration, market and technology leadership, and capital investment instead of relying solely on market-cap weighting. The ROBO Global Robotics and Automation Index ETF (ROBO) and the ROBO Global Artificial Intelligence ETF (THNQ) were cited as examples that follow such a scoring approach.

Index selection for those funds involves input from industry experts, including Engelberger Award winners and directors of robotics and AI labs at MIT, UC San Diego and Carnegie Mellon, Silva noted. VettaFi disclosed it is the index provider for THNQ and ROBO and receives an index licensing fee. The firm also stated the ETFs are not issued, sponsored or sold by VettaFi and that VettaFi and its affiliates have no obligation or liability related to the funds’ issuance, administration, marketing or trading.

Silva said robotics is the primary beneficiary of physical AI as machines move from isolated, pre-programmed devices to adaptable systems that operate in dynamic spaces with people and other robots. In surgical robotics, for example, systems can recognize tissue and bone properties and enforce safety limits while a surgeon performs the procedure. Silva emphasized, “We don’t want robots to replace humans. What we want is people to be able to do jobs where the human qualities are needed.”

He pointed to structural factors that are increasing demand for autonomous machines, including chronic global labor shortages, aging populations and efforts to reshore manufacturing.

Silva recommended that investors focus on companies that either develop core physical AI technologies or implement those technologies in fielded systems to gain exposure to the sector’s development.