Artificial intelligence is increasingly moving beyond software and into physical environments.

Manufacturers are exploring robotics and intelligent automation to improve production, reduce errors and address shortages of skilled workers. The development is creating a new category of technology often described as physical AI.

Unlike traditional software applications, physical AI has to interact with the real world.

Machines must understand objects, respond to unexpected situations and perform physical tasks accurately. That makes the technology considerably more complex than generating text or analysing digital information.

Manufacturing companies are among the early adopters because factories provide structured environments where automation can deliver measurable improvements.

Companies such as GE Appliances are already exploring AI-driven automation, while startups such as Hadrian are developing technology designed to improve industrial production and reduce dependence on hard-to-find technical skills.

The opportunity is significant.

Manufacturers across the United States and other developed economies are dealing with ageing workforces and shortages of specialised technicians. Automation could help companies increase output without relying entirely on expanding their human workforce.

But the transition also raises difficult questions.

Previous waves of industrial automation created productivity gains while eliminating certain categories of employment. Physical AI could create a similar transformation, particularly in repetitive manufacturing roles.

The difference this time is the level of flexibility.

Traditional industrial robots were generally designed for specific tasks. Newer AI-powered systems are being developed to handle more variable environments and learn from changing conditions.

If that capability improves significantly, it could expand automation into areas that were previously too difficult or expensive to mechanise.

For business leaders, however, technology adoption will require more than buying machines.

Companies will need workers capable of maintaining, supervising and improving automated systems. That means investment in technical training and workforce development will become increasingly important.

The rise of physical AI could therefore create both job displacement and new categories of employment.

The strategic opportunity is especially important for countries attempting to strengthen domestic manufacturing.

If AI-powered automation can make factories more productive and competitive, it could support efforts to bring more industrial production closer to major markets.

But the economics will depend on capital costs, reliability, workforce skills and the ability of companies to scale the technology.

Physical AI is still developing, but its direction is becoming increasingly clear: the next stage of artificial intelligence may not remain inside computers.

It could increasingly become part of the machines that manufacture, move and build the physical world.