Arm CEO Rene Haas predicts that AI-powered humanoid robots could take over large portions of factory work within the next five to 10 years, revolutionizing manufacturing.
One of the main drivers for bringing humanoid robots into factories is their advantages over the robotic arms and other automated machines currently in use, Haas said. Traditional factory robots are specialized machines designed for a single task, with both hardware and software optimized for that specific function. In contrast, general-purpose humanoid robots will be able to take on a variety of tasks on the fly by rapidly changing instructions, combined with increasingly sophisticated “physical AI” to help them navigate the real world.
“Within the next five years, a large portion of factory work will be replaced by robots, and one of the reasons for that is that these physical AI robots can be reprogrammed to do different tasks,” Haas said Monday at Fortune's Brainstorm AI conference in San Francisco.
“One of the problems we've had with factory robots in the past is if it was a pick-and-place machine in a factory, the robot was only optimized for one task. The software was optimized for one task, the hardware was optimized for one task. Now, if you design a general-purpose humanoid whose software is all AI and learns by doing, you're going to completely replace a bunch of factory workers,” he said.
What will happen to these workers and the broader job market as AI and robots become more prevalent in businesses is a growing concern among many policymakers and industry players, with options being discussed include ideas ranging from worker reskilling to universal basic income.
Although Haas did not specifically address the employment issue, he suggested that widespread adoption of physical AI could reshape the dynamics of global manufacturing and help level the global playing field through greater automation of factory work. “Physical AI will be a great enabler,” he said.
Haas also pointed to Waymo's self-driving cars as an early indicator of the potential of physical AI.
He said the next generation of autonomous systems could require even less hardware. While current self-driving cars are equipped with radar and cameras to survey their surroundings, future iterations with more advanced AI models could work with fewer sensors and rely on artificial intelligence rather than exhaustive data collection to make decisions.
Semiconductor supply chain has 'many single points of failure'
Arm doesn't make or sell its own chips; it designs and licenses the architecture used in processors made by companies like Qualcomm and Apple. Chips based on Arm's designs are used in everything from smartphones and refrigerators to cars and servers, and most people have between 50 and 100 Arm chips in use in their personal lives and homes, Haas said.
This widespread usage and market share is a testament to the energy efficiency and performance that makes Arm's chip designs so popular. But it also increases risks to the semiconductor supply chain.
Asked about the vulnerability, Haas acknowledged the extreme market concentration within the industry, noting that several large companies each control key parts of the semiconductor supply chain.
“There are a lot of single points of failure in the semiconductor supply chain… You have TSMC, which is in a very obviously interesting part of the world geopolitically. You also have a very sophisticated device that needs to be put into these fabs that is coming from one company on the planet… It's called ASML.”
In recent years, the COVID-19 pandemic has exposed some of the vulnerabilities in the supply chain, with chip shortages leaving consumers unable to get new car key chains for weeks. Haas said the crisis is “simply a function of the semiconductor supply chain, which has many single points of failure.”
Haas said the industry as a whole is “learning to live with” concentration risk.
Read more about Fortune Brainstorm AI:
Cursor has developed an in-house AI help desk that handles 80% of employee support tickets, says CEO of $29 billion startup
OpenAI Chief Operating Officer Brad Lightcap says 'Code Red' forces ChatGPT maker to focus as company ramps up enterprise efforts
Amazon's robotaxi service Zoox will start charging for rides in 2026, co-founder says it will 'focus' on transporting people rather than deliveries
