The roboticist who co-founded iRobot just delivered a brutal reality check to the humanoid robot industry. Rodney Brooks, who spent decades at MIT developing the technology that powers today's autonomous vacuums, says investors pouring billions into companies like Tesla and Figure are funding "pure fantasy thinking" that will never scale to mass production.
The man who helped create the Roomba just torched the entire humanoid robotics industry. Rodney Brooks, co-founder of iRobot and former MIT AI lab director, published a scathing takedown of the billions flowing into companies trying to build human-like robots.
Brooks specifically targets Tesla's Optimus program and Figure, which just raised funding at a staggering $39 billion valuation. Both companies are betting they can teach robots dexterity by showing them videos of humans performing tasks. Brooks calls this approach "pure fantasy thinking" in his latest essay.
The fundamental problem? Human hands pack about 17,000 specialized touch receptors into a space smaller than a smartphone. No current robot comes remotely close to matching this tactile sophistication. "While machine learning revolutionized speech recognition and image processing, those breakthroughs built on decades of existing technology for capturing the right data," Brooks explains. "We don't have such a tradition for touch data."
This isn't just an academic critique from the sidelines. Brooks has been building robots since the 1980s, when he pioneered behavior-based robotics at MIT. His work led directly to the algorithms that power modern autonomous systems, from Mars rovers to the millions of Roombas navigating living rooms worldwide. When he calls something impossible, investors should listen.
The safety mathematics are equally damning. Full-sized walking humanoid robots require massive energy to stay upright - energy that becomes destructive when they inevitably fall. Brooks points out that basic physics means a robot twice the size of today's prototypes would pack eight times the harmful energy upon impact. That's not a software problem you can patch.
Meanwhile, the funding frenzy continues. Figure announced its latest round just weeks ago, with backers including Microsoft and Nvidia betting on humanoids for warehouse and manufacturing work. Tesla keeps promising Optimus will revolutionize labor, with Elon Musk claiming the robots could eventually be worth more than the car business.
Brooks isn't buying it. He predicts that successful "humanoid" robots 15 years from now will look nothing like humans. They'll roll on wheels instead of walking on legs. They'll sport multiple specialized arms rather than trying to mimic human anatomy. Most importantly, they'll abandon the constraint of human form factor in favor of sensors and actuators designed for specific tasks.
The veteran roboticist sees today's humanoid investments as expensive training experiments that will never scale to mass production. Companies are burning through hundreds of millions trying to solve problems that may be fundamentally unsolvable with current technology. "We're seeing the same hype cycle that hit self-driving cars," one robotics industry insider told me, speaking anonymously. "Lots of promises, lots of funding, but the physics haven't changed."
The contrarian view comes at a crucial moment for the industry. Venture capital firms have poured over $2 billion into robotics startups this year alone, according to PitchBook data. Much of that money is chasing the humanoid dream, betting that human-shaped robots will unlock massive markets from elder care to construction.
But Brooks has seen these cycles before. He watched the first AI winter in the 1970s, the robotics boom and bust of the 1990s, and the recent autonomous vehicle reality check. His track record of predictions - including early skepticism about self-driving car timelines - has proven remarkably accurate over decades.
The essay arrives just as several high-profile humanoid demonstrations have captured public attention. Boston Dynamics continues showcasing its Atlas robot's athletic abilities, while startups promise household helpers and factory workers. But Brooks argues these demos mask fundamental limitations that no amount of machine learning can overcome.
Brooks' warning carries weight precisely because he's not an outside critic - he's the guy who actually built the robots that work. His message to investors is stark: the humanoid robot bubble will burst because it's trying to solve the wrong problems with the wrong approaches. The future belongs to specialized machines optimized for specific tasks, not general-purpose androids that look like us. For an industry built on mimicking human form, that's a fundamental challenge to everything they're building.
