Show summary Hide summary
When a legendary roboticist calls your vision “pure fantasy,” it might be time to listen. Rodney Brooks, the MIT pioneer and co-founder of iRobot, has delivered a stinging critique of Elon Musk’s Optimus ambitions, exposing a fundamental gap between what Silicon Valley promises and what physics actually allows.
The Man Behind the Critique
This isn’t just another skeptic chiming in from the sidelines. Rodney Brooks is a historical figure in modern robotics, a professor emeritus at MIT whose credentials are carved into the field itself. When he speaks, the robotics community listens. His latest assessment? The billions of dollars that venture capital has poured into humanoid robots rest on a profound misunderstanding of physical limits.
The real problem, Brooks argues, isn’t what most people think. It’s not about teaching machines to walk or talk. Those challenges pale in comparison to the true bottleneck: complex coordination and the sense of touch. These are domains where current machines remain primitively inadequate.
MIT robotics pioneer calls Musk’s Optimus vision “pure fantasy” as he exposes the impossible challenge ahead
Sara Errani transitions to player-coach role, coaches Jasmine Paolini
The Touch Problem Nobody Wants to Face
Here’s where it gets sobering. The human body operates a sensory system so dense that today’s technology cannot replicate it. A human hand alone contains 17,000 mechanoreceptors capable of processing vibrations and pressure through 15 different families of neurons. This precision allows us to manipulate fragile or complex objects without conscious thought. The creator of the famous Roomba argues that without this sensory foundation, the entire idea of a versatile robotic assistant is doomed.
Tesla and the startup Figure are taking what Brooks sees as a dangerously incomplete approach. Both companies rely on imitation learning, feeding their AI videos of humans performing manual tasks. The expert finds this methodology fundamentally insufficient because it ignores tactile data, which is absolutely essential for genuine dexterity. Unlike image or sound, there is no massive dataset for robotic “touch.”
No amount of funding can overcome this gap. Humanoid robots will not learn to be dexterous despite hundreds of millions, or even billions of dollars given by investors and major technology companies to finance their training.
A Better Path Forward
Brooks suggests the industry is heading in the wrong direction by injecting massive capital into training vision-language models. He points to a jarring disconnect between marketing promises and the reality inside robotics labs. His prescription? A fraction of current spending would prove far more effective if directed toward fundamental research in university settings.
As he told Fortune: “If the large technology companies and investors throwing their money into large-scale humanoid training spent only 20 percent of it giving it to university researchers, I think they would get closer to their goals more quickly.”
The future of useful robotics probably won’t follow the human blueprint at all. Within fifteen years, Brooks predicts, the most effective robots will be equipped with wheels for mobility and multiple arms for efficiency, far removed from the bipedal silhouette of Optimus. The term “humanoid” may persist for marketing purposes, but the physical structure will ultimately be dictated by operational profitability.
Reality Versus Hype
Yet Musk continues to maintain an ambitious timeline, promising to sell Optimus to the public by late 2027. Figure boasts a record valuation of 39 billion dollars. Brooks offers a sobering reminder: investor enthusiasm doesn’t guarantee corporate survival. His own company, iRobot, filed for bankruptcy last December after Amazon cancelled its acquisition.
Plenty of money will vanish, spent trying to squeeze any performance from today’s humanoid robots. But those robots will be long gone and mostly conveniently forgotten.
