Rodney Brooks—another of the world’s top robotics experts and co-founder of iRobot, the company behind the Roomba—has spoken out. He is worried about where humanoid robots are headed. Brooks’ decades of leadership at the Massachusetts Institute of Technology (MIT) make him appropriately skeptical about the likes of Tesla and Figure, which have garnered the most hype in humanoid robotics. He contends that billions of dollars are being wasted on the startups that will never produce a viable product.
Brooks is at his most scathing about the style of training adopted by most robotic companies. As O’Toole sees it, the government’s underlying approach is completely off-base. This approach typically focuses on teaching robots dexterity by training them through human demonstrations that are recorded in videos. According to Brooks, this strategy fails to address the complexities of human movement and interaction, which cannot simply be replicated through observation.
The roboticist cautioned that the next 15 years will usher in a new approach to designing humanoid robots. He argues that the most successful robots will be the least anthropomorphic. Instead, they’ll have wheels for off-road mobility, multiple jointed arms for versatile actions, and specialized sensors to enhance their operations. Finally, he argues that simply leaving the human form behind is necessary. This shift is necessary for robots to be able to carry out operations in varied and complex environments.
Brooks cautions against pouring dollars into humanoid robots as they stand today. He asserts that billions of dollars are being shuttled into these plans. He argues that this funding primarily goes to expensive training experiments that won’t eventually lead to large-scale production. He points to the challenges of replicating human dexterity as a major hurdle. A human hand has at least 17,000 specialized touch receptors. No robot has even come close to matching this amazing subtlety.
Additionally, Brooks underscores the existential threats posed by more massive robots. He points out that a bot twice the size of today’s models would have eight times the deadly energy. This not only raises safety issues, but raises practical concerns. The costs of deploying these kinds of machines make the case that there should be high standards of consideration and innovation in robotic design.
