LAS VEGAS – The Consumer Electronics Show (CES) 2026 made one thing clear: artificial intelligence is no longer confined to screens. Instead, it’s moving into the physical world through robots designed for practical labor, though the path forward isn’t without hurdles.
From Spectacle to Utility: The Shift in Robotics
For years, robotics demos focused on flashy stunts—robots running marathons or performing parkour. Now, industry leaders are pivoting towards more realistic applications. Robert Playter, CEO of Boston Dynamics, noted that “useful work” in fields like mining, construction, and logistics is where automation will truly take hold. This shift acknowledges that the high costs of robotics are justified only when applied to repetitive, expensive tasks.
The move toward pragmatism isn’t just about functionality. Trust is emerging as a critical barrier. Unlike early AI systems limited to chat windows, where errors were mere annoyances, mistakes in physical AI can be dangerous. A recent incident involving a Zoox robotaxi in Las Vegas—which stalled at a crosswalk, leaving pedestrians confused—underscores the need for robots to be predictable.
The Challenge of Trust and Transparency
Experts at CES emphasized that physical AI must be “transparent in its thinking and motions” so humans can understand what to expect. This is essential for building confidence in systems like self-driving cars, where unpredictable behavior can have catastrophic consequences.
Achieving this transparency requires a surge in computing power. The demand for more advanced AI models is outstripping the pace of chip development, forcing manufacturers to accelerate innovation cycles. However, even faster chips won’t solve the underlying problem: the sheer scale of AI deployment is driving up energy consumption and costs. The infrastructure needed to support AI running “everywhere, all the time” simply doesn’t exist yet.
The Long View: Hype vs. Reality
The enthusiasm surrounding physical AI at CES echoes earlier tech waves, such as the Internet of Things (IoT) in 2010. Some hype is inevitable, but time will tell which innovations endure. The core question isn’t whether AI will become part of daily life but when and how. If it does, the robots showcased at Mandalay Bay will need to do more than just wave for cameras; they will need to perform reliable, useful work.
The future of AI is less about spectacle and more about the practical challenges of integrating intelligence into the physical world. The journey will require overcoming issues of trust, cost, and computing limitations before the full potential of these technologies can be realized.
