Robot Reality Check: Why Humanoid Robots Won't Take Your Job

Updated: April 19 2025 07:46
At a recent Stanford HAI presentation, robotics pioneer Rodney Brooks offered a refreshing perspective that cuts through the noise, drawing on his 40+ years of experience in both academia and industry.

Brooks—Panasonic Professor of Robotics (emeritus) at MIT, founder of multiple robotics companies including Rethink Robotics and currently CTO of Robust AI—shared insights on how robotics can truly succeed in our human world. His talk balanced optimism with realism, offering a much-needed compass for navigating the current AI and robotics hype cycle.




Unlike Asimov's famous fictional laws focused on preventing robots from harming humans, Brooks' laws address practical considerations for successful robot deployment in human environments:

• First Law: "The visual appearance of a robot makes a promise about what it can do and how smart it is. It needs to deliver or slightly over-deliver on that promise, or it will not be accepted."
• Second Law: "When robots and people coexist in the same space, the robots must not take away from people's agency, particularly when the robots are failing—and they always fail."
• Third Law: "Technologies for robots need 10+ years of steady improvement from lab demos before their limitations are well enough characterized to deploy in the real world."

Brooks admitted that his third law is somewhat understated—the timeline is really closer to 30 years, but "people can't believe that." He illustrated this with the example of self-driving cars, which have been in development since at least 1979, with significant demonstrations on highways by 1990, yet still haven't achieved widespread commercial deployment in 2025.


Brooks' career has consistently focused on robots that can safely and effectively interact with humans. He showcased his journey:

• Early mobile robots at Stanford AI Lab in the 1970s
• Building robots that could navigate around humans at MIT in the 1980s
• Creating humanoid robots that could respond to faces and imitate human actions
• Developing emotional interaction capabilities through prosody (the rhythm and intonation of speech)
• Commercializing with consumer products like the Roomba, which featured a handle specifically designed to give humans agency when the robot got stuck
• Building safe, collaborative industrial robots that could work alongside humans without cages

Today, with his company Robust AI, Brooks is tackling warehouse automation—but with a distinctly human-centered approach. Rather than trying to replace human pickers (who typically walk 30,000 steps daily), his team built Carter, an AI-powered cart that follows workers, carries heavy loads, and enhances human capabilities rather than attempting to replace them.



One of the most valuable aspects of Brooks' presentation was his historical perspective on AI and robotics hype cycles. He emphasized: "For 70 years, we've been working on AI, and there's been significant progress in the last decade—and that's true for every decade... We are not living in a special, blessed time."


Drawing on his decades of experience, Brooks identified several recurring hype cycles:

• Multiple waves of reinforcement learning excitement
• Repeated "revolutions" in medicine
• Six distinct neural network hype cycles dating back to 1943

Brooks traced the history of neural networks from McCulloch and Pitts in 1943, through Marvin Minsky's 1953 PhD thesis, Frank Rosenblatt's perceptron in 1957, and Bernie Widro's implementation in 1960—which remarkably used the same basic neuron architecture we still employ today. This historical perspective highlights how technologies often undergo multiple cycles of enthusiasm and disillusionment before finding their practical applications.


He specifically cited IBM's Watson as a cautionary tale. After its 2011 Jeopardy victory, Watson was hyped as revolutionizing medicine and consumer purchasing, with claims that "the future's happening very quickly." By 2018, the project faced major layoffs and scaled-back ambitions. Brooks warned that today's generative AI and humanoid robots are following similar trajectories of inflated expectations.



According to Brooks, two technologies are currently at the peak of their hype cycles:

1. Generative AI - Already past peak hype according to Gartner
2. Humanoid Robots - Still climbing the hype curve with "worse to come"

Both share common characteristics driving their hype:

• Unrealistic startup valuations
• Billions in spending from tech giants
• Promises of plentiful, low-cost labor (white-collar for generative AI, blue-collar for humanoid robots)
• Deceptive surface capabilities that create an illusion of deeper intelligence

Brooks was particularly critical of the "humanoid theater" demonstrations often showcased in viral videos. He encouraged viewers to be skeptical, noting telltale signs like slowed-down movements, carefully controlled environments, and the absence of naturalistic human-robot interactions.


To explain why we fall for these cycles, Brooks borrowed three insights from physicist Werner Heisenberg:

1. "What we observe is not nature itself but nature exposed to our method of questioning."
2. "The reality we can put into words is never reality itself."
3. "An expert is someone who knows some of the worst mistakes that can be made in their subject and how to avoid them."

These principles help explain how our understanding of AI and robotics is shaped by our preexisting models and limitations. For example, the current focus on end-to-end learning through image labeling misses crucial aspects of human vision like color constancy, looming detection, and our ability to understand novel situations.

Similarly, our tendency to anthropomorphize robots (like five-fingered hands) ignores that human anatomy is largely the result of evolutionary history rather than optimal design. Brooks cited oyster shucking as a task that requires force feedback and tactile sensation—not just visual imitation—highlighting the gap between current robot learning approaches and true dexterity.


Brooks concluded with advice for those working in robotics and AI:

• Don't abandon promising research just because it's not currently trendy
• New faculty members shouldn't let department heads pressure them into following hype
• Expect significant experimentation in reducing training costs over the next decade
• Look at technological claims with an open yet critical mind
• Focus on the human dimensions: appearance, agency, and maturity

Perhaps Brooks' most important message was to avoid the "soccer children syndrome" where researchers all chase the same ball, ignoring the rest of the field. Innovation often comes from those working on neglected problems or approaching challenges from unconventional angles.

Throughout his career, Rodney Brooks has consistently emphasized that robotics must be developed with humans at the center of the design process. His current work with Robust AI epitomizes this philosophy with its mission statement:

To make robots work for people, not the other way around.