The Evolving Landscape of Humanoid Robotics
This story was originally published in The Highlight, Vox’s member-exclusive magazine. To get access to member-exclusive stories every month, join the Vox Membership program today.
There’s something unsettling about witnessing a humanoid robot lying motionless on the floor. Without power, these bipedal machines cannot stand; if they’re turned off and not secured in another form, they simply sprawl out, staring up in a position that evokes a sense of helplessness.
That’s how I first encountered Atlas a few months ago. I had seen the robot perform impressive feats on YouTube—navigating obstacle courses, executing backflips, and more. Yet, there it was in a lab at MIT, motionless on the floor. The stark contrast was jarring. Humanoid robots have become increasingly capable and prevalent since Atlas first gained fame online.
Advancements at Boston Dynamics
At Boston Dynamics, the company responsible for Atlas, a newer version of the humanoid robot is learning skills like instinctively picking up and dropping objects, thanks to an innovative AI model that governs its movements. Some of these next-generation Atlas robots are on the cusp of working on factory floors and even venturing further into various roles. Due to the advancements in AI, it seems likely that general-purpose humanoids are on the horizon.
“In Shenzhen, you can already see them walking down the street every once in a while,” said Russ Tedrake, who oversees the Robot Locomotion Group at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL). “You’ll start seeing them in your life in places that are probably dull, dirty, and dangerous.”
A Shift in Focus: Understanding and Interaction
Tedrake’s team has transitioned from merely teaching robots how to walk to instructing them on understanding and interacting with their environments through advanced software and AI applications. A significant spike in venture capital investment in robotics supports this trend; funding grew from $42.6 million in 2020 to an anticipated nearly $2.8 billion in 2025. Analysts at Morgan Stanley predict that global sales of humanoid robots could reach 900,000 by 2030, escalating to more than 1 billion by 2050, primarily aimed at serving industrial and commercial needs.
The potential implications are substantial. As humans have designed the world with themselves in mind, humanoid robots may efficiently navigate and replicate our capabilities.
Janik Söllner for Vox
However, not all humanoid robots will be relegated to factory floors. Some startups are developing autonomous robots for domestic tasks. For instance, X1 Technologies has started taking preorders for its home robot, Neo, which can perform chores like washing dishes and fetching snacks. Other companies, such as Figure AI and Sunday Robotics, are also introducing consumer-friendly robots intended to assist with various household activities.
The Hurdles of Deployment
While many impressive demonstrations of these AI-driven home robots exist, actual deployment remains limited. Often, they rely on human operators, making them less autonomous and more akin to puppets. Manufacturers like Apptronik and Figure AI have only minimal robotic presence on production floors, primarily handling rudimentary tasks.
As Tedrake points out, “In order to make them better, we have to make AI better.” Most robots currently represent the physical embodiment of AI. This intersection aims to enhance capabilities in various sectors, such as agriculture, logistics, and even surgical applications.
Bridging the Gap between Hype and Reality
The technological landscape indicates that while humanoids hold potential, they aren’t ready to replace human workers widely just yet. Tedrake informs us that humanoids will require significant improvements in AI capabilities before they can effectively operate in roles alongside humans. This reality is particularly evident in the disparity between expectations and current advancements.
Janik Söllner for Vox
People often get overly excited about the idea of humanoid robots, neglecting to consider their limitations. Rodney Brooks, co-founder of iRobot, has noted, “People in general get too excited by the idea of the robot and not the reality.”
The disparity between exciting technological demonstrations and practical applications underscores the ongoing challenges. In 2023, a paper co-authored by Tedrake discussed a novel method for training robots to mimic human movement, utilizing motion-sensor gloves worn by humans. This groundbreaking research promises more intuitive robot behavior, which will be crucial for future developments.
The Future of Robotics: Better AI Needed
The seriousness of making humanoid robots effective lies not only in their abilities but also in the investment towards ensuring safety in human environments. While humanoids capture the public’s imagination, many experts suggest focusing on specialized robots capable of performing specific tasks efficiently and safely.
Nevertheless, the public’s fascination with humanoid robots persists, as society has long envisioned machines resembling human capabilities. The projection for humanoid robots is optimistic but may require decades to come to full fruition. As Tedrake wisely notes, “Let’s not do open heart surgery right away with these things. This is more like folding laundry.”
For regular updates and information about the evolving world of robotics, you can read more Here.
Image Credit: www.vox.com
