Elon Musk recently announced on Twitter that "Tesla will have genuinely useful humanoid robots in low production for Tesla internal use next year." This suggests that robots with human-like physical characteristics, capable of performing "genuinely useful" tasks, may be closer to reality than ever.
Despite decades of effort, useful humanoid robots have mostly remained the stuff of science fiction, never quite materializing in everyday life. Are we finally on the brink of a breakthrough? It's worth asking if we even need humanoid robots in the first place.
Tesla’s Optimus robot is one of several emerging humanoid designs, alongside Boston Dynamics’ Atlas, Figure AI's Figure 01, Sanctuary AI’s Phoenix, and others. These robots typically have a bipedal form, allowing them to walk—and sometimes even jump—while their arms and hands handle objects with varying degrees of dexterity.
Behind their eyes, artificial intelligence drives navigation, object recognition, and task execution. These robots are often envisioned for industrial settings, where they can perform repetitive, dirty, dull, or dangerous tasks. They could also work collaboratively with humans, like carrying a ladder together.
Beyond factories, humanoid robots are being considered for the service industry, potentially replacing today’s utilitarian "meet and greet" or "tour guide" robots.
There’s even potential in social care, where robots like Japan's Riken Robear have been developed to lift and move patients, though they resemble bears more than humans. Another sector seeing growth is humanoid sex robots, which stir ethical debates that other areas of humanoid robot use do not seem to provoke.
Still, delivering practical humanoid robots remains a significant challenge. Why?
One major hurdle is achieving flexible, human-like bipedal locomotion over varied terrain. It took humanity millions of years to evolve this ability, so current robot progress is impressive, but robots still struggle with adapting to dynamic environments.
Similarly, replicating the human hand’s ability to dexterously manipulate objects of all shapes, sizes, and fragilities has been stubbornly difficult, though companies like the UK’s Shadow Robot have made significant strides in robotic hand design.
Human bodies are covered with soft, flexible skin that constantly senses and responds to the world around them. Robots, in contrast, have limited tactile sensitivity, usually confined to a few contact points like fingertips.
To move from automating specific tasks to performing general functions in unpredictable environments, robots need much more advanced artificial intelligence, sensing, and mechanical capabilities. And if a robot is to resemble a human, people expect it to communicate like one, possibly even showing emotional responsiveness.
This is where things get tricky. Human brains are wired to pick up on subtle non-verbal cues, and if a humanoid robot doesn’t get these micro-expressions right, it can come off as unnerving or even creepy. This "uncanny valley" effect is a significant barrier to human-robot interaction, alongside the challenges of making these robots operate effectively in unpredictable, real-world conditions—far more complex than controlled lab environments.
But is building robots in our image the best approach? This brings us to a broader question: Why are Musk and others so fixated on humanoid robots? Does a robot need to look like us? One argument is that human-shaped robots would easily fit into environments already designed for humans—our buildings, tools, and cities are built to accommodate our physical form.
Yet, our built environments often require a certain level of strength, dexterity, and sensory ability, which disadvantages many people, including those with disabilities. Would the rise of humanoid robots reinforce this divide, introducing machines that surpass human abilities in strength or precision?
Perhaps a more inclusive vision would focus on integrating robotics into our environment—embedding smart systems into buildings, furniture, and vehicles to make the world more accessible for everyone, regardless of physical ability.
Strikingly, current robot designs rarely reflect the diversity of human bodies. Maybe our obsession with creating humanoid robots stems from deeper motivations, like the age-old desire to create god-like versions of ourselves—a theme often explored in dystopian science fiction and eagerly adopted by the tech industry.
Or perhaps humanoid robots are simply a "moonshot" goal—an ambitious, easily understood vision that’s incredibly difficult to achieve. Even if we’re not entirely sure why we need robots that look like us, the innovations that emerge from the effort are likely to be transformative.
The Fourth Age:
Smart Robots, Conscious Computers, and the Future of Humanity
"Timely, highly informative, and certainly optimistic." ― Booklist